There are three ways of trying to win the young. There is persuasion, there is compulsion, and there is attraction. You can preach at them: that is a hook without a worm. You can say, You must volunteer, and that is of the devil. You can tell them, You are needed. That appeal hardly ever fails.

10 to 25: The Science of Motivating Young People: A Groundbreaking Approach to Leading the Next Generation—And Making Your Own Life Easier, by David Yeager1

How to Talk So Kids Will Listen & Listen So Kids Will Talk, by Adele Faber & Elaine Mazlish

I’ve always thought it was crazy that people will go to school for 20 hours a week for four years to be good at their job, but when it comes to being a parent—a “job” they’ll be working at for many hours a week for 18 years, and continuing part-time for the rest of their life—they’ll, like go to a 1-hour class one night a week for a few months before their kid is born and call it a day. What if we put the same effort into parenting that we put into being good at our careers? I myself read these books as a coach—it doesn’t look like I’ll have kids anytime soon—but maybe that’s something we should all think about.

How to Talk So Kids Will Listen (”HTTSKWL”) was a bit outdated but still worth it for a few useful nuggets in a short book; it’s more aimed at parenting younger kids. 10 to 25 was more insightful than I expected and definitely a useful resource for high school/college coaches. Here’s some book reviews2:

How to not hit your kids

HTTSKWL was first published around 1980, and it shows. There’s a number of stories that involve parents either hitting their kids or at least thinking about hitting their kids, like this one:

I was so gung ho to get this new approach going in my family that I came home from a meeting, tripped over my daughter’s skates in the hall, and sweetly told her, “Skates belong in the closet.” I thought I was wonderful. When she looked up at me blankly, and then went back to reading her book, I hit her.

At no point is it implied that any of this is abnormal. But things have changed—for the most part—45 years later. Some of the parents in the book’s anecdotes are so bad that I really can’t see my median friend being this bad at parenting. The following is probably not even the worst example (aside from the actual hitting of course), but it’s bad enough to serve:

[BILLY and his FATHER having a conversation about dad’s toolbox]

BILLY: Gee, Dad, I didn’t mean to. Sometimes I just forget.

FATHER: Well, maybe this will help you remember. Not only are you never going to get a chance to use my tools again, but for lying to me on top of it, you’ll stay home when we all go to the movies tomorrow!

Or:

“Bruce, you forgot your lunch! If your head weren’t screwed on to your shoulders, I bet you’d forget that, too.”

So to some extent, I think the fact this book has been so successful is a reflection of where society was 40, 30, 20 years ago, and the best parenting book for the current generation of middle-class parents is going to be some other book that presumes a higher level of average parenting skill.

And of course the book itself never suggests hitting your kids is OK. They promise “methods that affirm the dignity and humanity of both parents and children”, and for the most part they deliver. In other words, it’s not without its useful tips! They suggest the radical idea of...actually trying to problem-solve together, with your kids:

Step I. Talk about the child’s feelings and needs.

Step II. Talk about your feelings and needs.

Step III. Brainstorm together to find a mutually agreeable solution.

Step IV. Write down all ideas—without evaluating.

Step V. Decide which suggestions you like, which you don’t like, and which you plan to follow through on.

Obviously there’s some age where kids are too young for this...but that age may be younger than you think! This is one place where I actually can see the modern median parent not doing such a great job. It takes time to be a good parent and it’s so easy to fall into the trap of just using your power as an adult to unilaterally decide on a solution.

10 to 25 likewise supports the strategy of collaborative troubleshooting:

Berquist’s [liver transplant] clinic also invites the transplant patients together at summer camps, so they feel like a part of a community... Berquist’s team takes a very respectful approach in their communications with patients, one that I call collaborative troubleshooting.

A few times he uses the metaphor of “coauthoring a treaty” as your strategy for working through difficult situations:

Gary didn’t feel that confident going into the conversation [with his daughter]. To his surprise, it went shockingly well. They collaborated on some guidelines for screen time and in-person friend time that worked for Charlotte.

Overall, I think a fair bit of what both books boil down to is the revolutionary idea of treating young people like they’re actually humans. HTTSKWL has a chapter on “Encouraging Autonomy”; 10 to 25 asks us to “presume agency” in adolescents. Obviously there is some nuance to this—10 to 25 is telling us something like “adolescents are actually humans...who are just especially attuned to feeling low status/unwanted”. But basically you won’t go too wrong by treating young adults a bit more like how you’d an adult you respect as an equal—deciding things collaboratively instead of defaulting to using your power as an adult to make decisions, being careful about giving suggestions vs just letting someone vent, etc.

But on the other hand, another part of the reason to collaborate is to help young people develop into the person they can become—10 to 25 quotes says a coach “wanted [players] to have their own coaches in their heads that could keep helping them improve”; later the book warns us off “the compulsion to tell”, while HTTSKWL advises “Don’t rush to answer questions.” So, partly you should collaborate because it gives young people dignity, and partly you should collaborate because it lets them develop their own problem-solving skills.

I’d say both books are basically saying the same thing. 10 to 25 is:

• a little more modern,

• a little more worried about using scientific research to convince you, and provides a little more of an overarching theory of why kids are the way they are

• a little more focused on helping young people be “high performers” and not on just surviving as a parent,

• and aimed at working with young people who are a little bit older.

HTTSKWL focuses a bit more on providing lots of anecdotes and suggestions in the world of day-to-day parenting issues (bully at school, came home late for dinner, acting out in the supermarket, etc). But the underlying themes are often pretty similar—hitting the sweet spot of providing dignity and freedom to kids without being a total pushover.

Here are a few more tips from HTTSKWL that I found useful:

• “Give them what they want in fantasy” — instead of saying “No” to a kid when they ask for something you feel you can’t give them, tell them how much you wish you could give them what they ask for:
  

  “Giving in fantasy is my favorite skill...Kristen, my four-year-old, only likes peanut butter–flavored granola bars...Kristen was screaming and sobbing [when her sister ate the last one]. I assured her that we’d buy more as soon as possible, but she was not impressed...On and on it went until I finally remembered...“Oh yeah, give in fantasy.”

  In my most dramatic voice I said, “Kristen, do you know what I wish?” She stopped crying to say, “What?” ... “I wish we had a granola bar as big as that picnic table.” Kristen’s eyes grew wide. She said, “No, Mom. That would be too big.” “Well, then...I wish we had a granola bar as big as that rock over there.”
  

• Various versions of not over-asking questions:
  One common parental inquiry that seems to be experienced as a pressure is “Did you have fun today?” What a demand to make upon a child! Not only did he have to go to the party (school, play, camp, dance) but the expectation is that he should enjoy himself. If he didn’t, he has his own disappointment to cope with plus that of his parents. He feels he’s let them down by not having a good time…
  
  Some children can tell you why they’re frightened, angry, or unhappy. For many, however, the question “Why?” only adds to their problem...It’s much more helpful for an unhappy youngster to hear, “I see something is making you sad,” rather than to be interrogated with “What happened?”...Children don’t need to have their feelings agreed with; they need to have them acknowledged.

• Being careful about praising/how you praise (see also: Punished by Rewards book review):
  I learned...that helpful praise actually comes in two parts:

  1. The adult describes with appreciation what he or she sees or feels.

  2. The child, after hearing the description, is then able to praise himself.

  I recall the first time I tried putting that theory into practice. My four-year-old came home from nursery school, shoved a page of penciled scribble under my nose, and asked, “Is it good?”

  My first reaction was an automatic “Very good.” Then I remembered. No, I’ve got to describe. I wondered, How do you describe scribble?

  I said, “Well, I see you went circle, circle, circle . . . wiggle, wiggle, wiggle . . . dot, dot, dot, dot, dot, dot, dot, and slash, slash!

  “Yeah!” he nodded enthusiastically.

  I said, “How did you ever think to do this?”

  He thought awhile. “Because I’m an artist,” he said.
  

• Respect feelings as rational—”Denial of feelings teaches kids not to know what their feelings are—not to trust them”. E.g.:
  ME: It’s cold. Keep your sweater on.

  CHILD: No, I’m hot.

  ME: I said, “Keep your sweater on!”

  ...

  CHILD: That TV show was boring.

  ME: No, it wasn’t. It was very interesting.

  CHILD: It was stupid.

  ME: It was educational.
  
  It may be hard to believe, but kids will sometimes not like the things that you like! You gotta be able to let them do that.

The book has a fair number of short cartoons illustrating the relevant parenting concepts; I always give bonus points to a book that understands the power of using images/diagrams:

Resolving the mentor’s dilemma

Now let’s talk 10 to 25, which starts by introducing us to the mentor’s dilemma.

A player I coached once said something like,”Luke will point out something dumb you did, but he’ll do it in the nicest possible way.” I think they mostly meant it as a compliment, but I still wish I could’ve done an even better job of not making them feel dumb. That’s basically what the mentor’s dilemma is—for people to grow and improve, they have to notice their mistakes and work to fix them. A mentor can help a young person notice those mistakes...but it can feel bad to have someone point out your mistakes, to the point where it may strain the relationship. How do you choose between being a young person’s support, or giving them the useful constructive criticism they need to develop? As the author puts it:

Both Alex and the manager were trapped by the mentor’s dilemma. This refers to the fact that it’s very hard to simultaneously criticize someone’s work and motivate them, because criticism can crush a young person’s confidence. It’s a dilemma because leaders feel like they’re stuck between two bad choices. They could either put up with poor performance (but be nice) or demand high performance (but be cruel).

Science tells us that when young people are being critiqued by an authority figure, they’re asking themselves a deeper, more existential question: Does this person who has power over my life think I’m incompetent?...Before they can hear the criticism for what it is—helpful advice on how to better meet their goals—they have to feel safe.

One main solution to the mentor’s dilemma is something they call wise feedback:

We had instructors be critical with their feedback but accompany that criticism with a clear and transparent statement about the reason they were giving that feedback—namely that they believed the student could meet a high standard if they got the right support. So-called wise feedback is wise (or attuned) to the predicament of young people who don’t want to be held to an impossible standard and who also don’t want to be talked down to...The wise-feedback study pointed to a way to solve the mentor’s dilemma: by conferring status and respect on young people.

They say that “young people don’t like the compliment sandwich”, their main argument being that people will see through a weak attempt to tack on at-best-semi-related compliments (”Bland praise for something unimportant and unrelated to personal or professional success”) to soften the blow of a super-relevant criticism.

The wise feedback note, on the other hand, can “break through the barrier of mistrust” that occurs when someone on the wrong side of a power dynamic assumes the worst about the criticism they receive from above.

The theme in the quote above of “high standard and high support” recurs consistently throughout the book.

If there’s one concept you should remember from 10 to 25, it’s that—it’s possible to set high standards, if you also offer mentees a high level of support. One, but not the other, means you’re either a hated drill sergeant or a coddling mother hen; but both together is the sweet spot.

A lot of what the rest of the book breaks down to is convincing you high standards are actually possible (rejecting the incompetence model) and showing you how to create real, effective, lasting support (transparency, belonging).

The incompetence model

10 to 25 wants to convince you that the idea young people are basically mentally deficient until they’re 25+ is harmful to your relationships. We’ve all heard about how “young people’s brains are incapable of correctly weighing the future consequences of their behaviors—such as how an impulsive action now could lead to poorer health later.” But is that really right?

The book asks us to replace this “neurobiological-incompetence model” with a model of young people as highly effective, but also highly attuned to issues of status and respect.

Neuroscientists have shown that during puberty the brain becomes attuned to social status and respect—”earned prestige”...

The UCLA report explained why much of the behavior of young people that worries or annoys us does not come from ten-to-twenty-five-year-olds being inherently incompetent. Rather it comes from this age group’s attempt to learn how to be socially successful in the world...young people want status and respect from peers and mentors, earned by making meaningful contributions.

Instead of thinking of the adolescent brain as fundamentally flawed due to puberty’s hormones, we can think of it as ready to change in response to the prospect of status and respect.

That’s why, the book tells us, the “Think. Don’t Smoke.” tobacco PSA campaign was found to be ineffective (as are many well-meaning social interventions...) — it tells kids what to do, and it has “an insulting implication that the problem is young people’s inability to think clearly”. On the other hand, the more-successful “Truth“ campaign “depicted young people as the only ones who could think clearly...It was showing the world they were a rebellious and autonomous person worthy of an adultlike status.”

HTTSKWL makes a similar point in one chapter:

The hardest part is the shift we have to make in attitude. We have to stop thinking of the child as a “problem” that needs correction...It requires a great act of faith to believe that if we take the time to sit down and share our real feelings with a young person, and listen to his feelings, together we’ll come up with solutions that will be right for both of us.

This way of putting, from 10 to 25, felt especially apt for new players coming together on, say, a frisbee team:

we secure our sense of belonging through experiences of looking competent in front of the people whose opinions we care about.

This simultaneous need-for and lack-of belonging is perhaps a problem caused by the way modern society has made adolescence longer. Young people could be fully adult at age 18 (or even much younger) in preindustrial cultures; but nowadays there’s a “mismatch between young people’s neurobiological needs for status and respect and the level of status or respect afforded to them by their current circumstances”. Being in college until 22 and then at the bottom of the corporate totem pole even after graduating doesn’t mesh completely smoothly with young people’s innate need to garner the community’s respect.

And the incompetence model isn’t just, some half-truth the authors want you to believe for the sake of being a good mentor; there’s some science to back it up:

In newer laboratory experiments, Galván explained, young people often do better at goal-directed behavior than adults, when incentive structures are right.

[One study, for example, found that] their brains performed competently when their mothers talked to them neutrally.

The above-mentioned Adriana Galván also authored one of the original studies that showed less prefrontal cortex activation in adolescents. But she’s since changed her mind on the question: she “stands by that study’s results but no longer thinks the field has interpreted them correctly”— kids just aren’t always motivated by the same things adults are because “...they’ve got different motivational priorities.”

Transparency and belonging

Aside from the “wise feedback note”, author David Yeager suggests using a “transparency statement” to build a shared sense of the relationship between mentor and mentee:

The mentors had the right mindset. In both cases, however, they miscommunicated with mentees, who felt threatened. Then the miscommunication spiraled into hurt feelings and distrust. To resolve this problem, each could have been more transparent about their meaning...[the solution is] a transparency statement: a simple and clear declaration of your intentions at the start of any potentially threatening interaction.

I think this is an area where I could continue to strive to do better. Going back to my previous example, I think I’m pretty good at giving feedback in a way that’s calm and nonjudgmental; but I could still do better at explicitly making sure we’re on the same page about the hows and whys of feedback before I give it.

And it’s about being consistent, too—you may have to repeat the message “far more often than [you] initially thought [you] should have to”. It’s not the kind of thing you can say once on the first day of class and then forget.

He suggests3 using belonging stories to help young people feel included and normal when they’re at risk of feeling like a bad fit or dropping out of a culture they’re new to:

This could be something a coach writes their player about their own experiences as a young player. But even better is to have a senior team leader writing a note to a freshman a few years younger than them.

As mentioned in the previous section, young people are genetically attuned to search for status and respect, so belonging can be built through making people feel needed: “Our parents may offer us unconditional acceptance, but our peers usually want us to bring something to the table. Thus, our need to be accepted becomes tied to our need to feel competent. “

However, they warn against making a kid feel too much pressure: “The most serious mistake that I often see is invoking purpose in a way that causes anxiety. It can backfire to tell a young person that the whole world is counting on them...” To be honest, I wish they had said more on this topic. As a coach, how can I best say “the team needs you” in a way that makes players feel a sense of belonging instead of a sense of pressure?

A few more bits and pieces

Both books ask parents to not be afraid to do a do-over:

Parents think, I went to the seminar, and I read the book, and I still ended up blowing up at my kids. I guess I’m a bad parent. They fall back into [their previous] mindset...Lorena wants to free parents from that cycle, using a simple but profound concept: the do-over.

Basically, don’t be afraid to say “hey, I messed up...it wasn’t great of me to [X]. Can I work together with you on this again now that I’ve had some time to notice my mistakes?”. No one’s perfect, adults included. You definitely won’t get every interaction right on the first try, especially when emotions are high.

One little idea I liked from 10 to 25, that I’ve noticed in my own life, is that you should never ask a question you don’t legitimately want to know the answer to. Don’t ask “what were you thinking!?” if you don’t legitimately want to know what someone was thinking. Don’t ask “do you really think that was a good idea?” if you’d get mad when your kid replies “yes”. I first noticed this in my social life a few years ago, and, IMO, it’s a great tip for relationships of any kind. People notice the difference between an actual question and a rhetorical weapon being used against them that happens to be superficially question-shaped.

10 to 25 calls the right kind of question an “authentic question with uptake”. “Authentic” means you actually don’t know the answer yet and legitimately want to; “with uptake” means basically that it’s relevant to the specific conversation that came before and the answer will actually influence the conversation going forward. Elsewhere in the book they use the term humble inquiry: “the fine art of drawing someone out, of asking questions to which you do not already know the answer, of building a relationship based on curiosity and interest in another person.”

Happy parenting/mentoring/coaching!

I’d love to make this article as clear as it can be—with the reasons I choose to vote for someone as strong as can be—so if you’ve got good reasons for voting (or not voting) for someone that I don’t mention below, please let me know and I’ll add them to the article (and possibly change my decision if it’s not too late). Come back before voting to see if anything’s changed. Update: I’ve added small edits to a few sections; search for the word “edit” to access them quickly.

Resources

RI Government voter information page — See your sample ballot and confirm your in-person voting location.

PVD Streets Coalition endorsed candidates

BlueVoterGuide.org Rhode Island Voter Guide — Good for seeing endorsements (from organizations, doesn’t have all endorsements from individuals), not much beyond that.

Rhode Island Current Voter Guide — Has a Q&A with the biggest contenders — but only for the bigger races (Statewide, plus Providence Mayor)

OnTheIssues.Org — attempts to show you on one page how each candidate breaks down across the issues (here’s their page for the Governor’s race), but IMO it’s not complete enough to truly be useful.

PVD Eye, Providence Voter Guide — Good for showing you who’s running but (aside from links to candidate websites) won’t really help you decide who to vote for. (also, PVD Eye, City Council candidates answer questions)

Various debates are linked in the relevant sections below.

Where I’m coming from

This guide won’t be super useful to you if our opinions are vastly different, so I’ll start by briefly sharing my general outlook. I would generally say my goals are very progressive (I believe in equality for all sexes/genders/etc, I believe we can and should do more to help the poor, I believe we can and should spend less money on war, etc) but as a very nerdy, numbers-loving guy I sometimes believe that the policies that could best get us there are sometimes somewhat libertarian-flavored (it would be easier to build housing if we didn’t have a multi-hundred-page zoning code restricting what can be built! To give the biggest example.)

I make a point of being open to everyone, so I check the websites of all candidates, but I am pretty anti-Trump so I’m going to quickly count out most Republican candidates unless there’s some pretty extreme reason why I think they’re great.

I’m posting this before the primary election because I live in a left-leaning city in a left-leaning state, so the vast majority of races will be determined by who wins the Democratic Primary. But for completeness, I’ve included names (and links where I could find them) to Independent or Republican candidates as well.

Senator (US Congress)

D primary:

Jack Reed (incumbent)

Connor Burbridge

Luis Daniel Munoz

Others:
Raymond McKay (R)

Michael Bahry (I)

Jon Lindley (I)

Thoughts:

• Couldn’t find a website for Lindley though he’s listed on Ballotpedia.

• Allen Waters is also listed as an (R) option on Ballotpedia but is not on my sample ballot.

• Bahry posts AI-generated campaign material to his Twitter page that gets...less than 100 views, has a campaign calendar on his website that has literally nothing on it...

  

  ...and is very religion-forward on his website. (Not in any clearly bad way; but it still feels a little weird to me personally.) Not a viable candidate.
  

• Munoz is on the ballot but I can’t find a single news story mentioning him this election cycle and he’s not even listed on this prediction site, so if you’re looking for a non-establishment candidate it seems like you should go with Burbridge, not Munoz.

• So it’s really between Burbridge and Reed. Which I basically knew already but wanted to verify before continuing.

• Jack Reed is old (76) and establishment but it seems like he’s doing at least an OK job. The Reddit comments are surprisingly...fine, given how progressive and anti-gerontocracy Reddit comment sections often are:

  • “Say what you want about his age, but he’s very quietly one of the most respected Senators. Head of the armed services committee for years and helped secure long term contracts for EB in Quonset. He’s one of the good guys”

  • “Reed is suoer approachable. Show up to a ton hall or other event. Dude is serious but there is only so much you can do being in the minority. .had the chance to talk to him at the airport as we were both flying out. Proud to have him as my Senator. “

  • (About Burbridge) “He has no political experience. He thinks he’s AOC but he’s not inspiring; his speeches are choppy, and he hasn’t learned to memorize them so it makes them even more awkward (came to our recent No Kings and it was … not great). He doesn’t really understand the role of a Senator, or one from RI, as most of his messaging is focused on reclaiming the working class from Rs (makes more sense in a swing state or national). He’s raised less money than state senate candidates at this point in time. He’s sadly an empty idea progressive. I’m as liberal as they come but he’s up in the clouds on dreams and has zero rizz to back it up.”

• In his interview with Steve Ahlquist, Burbridge said this about how he’s different from Jack Reed:

  people ask, “How are you different from Senator Reed?” To me, it’s night and day. From my perspective, Senator Reed has the politics of the 1990s. He’s still fighting the Cold War. Activists will tell you that when they try to talk to him, he says things like, “We need to invest more and more into the military because what about Russia?”
  
  The heart of what we are putting forward, this idea of working-class politics, is something you’ll never hear Jack Reed talk about. Universal healthcare, taxing the greedy billionaires who have rigged the system against the working class, and having a government that is responsive to the needs of working people and is willing to throw its weight around to make the lives of regular hardworking Americans better.

In my mind, it’s kind of hard to imagine it matters that much whether Reed or Burbridge is our senator. The RI senator will continue to be (D), the Senate will continue to get nothing done, etc etc. I don’t really want to vote for a 76 year old b/c I think it’s time we let some younger people run the world, but nothing I’m hearing about Burbridge impresses me, either. I guess I’ll vote for Reed but I very much look forward to younger candidates and a competitive primary next cycle.

Edit: After posting I saw this post from Steve Ahlquist where Burbidge talks about how Reed has continually voted for/okayed larger military budgets. I think I had heard about this before but struggled to find a good source when I was writing the post. I do think I care about this enough that I’m going to switch my vote to Burbridge given that I was already somewhat on the fence between the two. After all, I did write above that “I believe we can and should spend less money on war”. I consider this more of a protest vote than a strong support of Burbridge though.

Representative (US Congress)

Gabriel Amo (incumbent)

Kellie Keenan (R)

Wilfred Curioso (I) is also listed on Ballotpedia but I couldn’t find a website for him.

• I would like to give Keenan some credit for having, for the most part, what seems like reasonable policy “goals” on her website.

• That being said, I think:

  • (a) her website is perhaps a little too lacking in detail to take 100% seriously, a bullet point that’s a few words long sounds nice but could mean anything in reality.

  • (b) I also see barely any news about her and she has no political experience

  • (c) I’m not really going to vote for any Republican who isn’t willing to be vocally anti-Trump

So I guess I’m voting Amo pretty much by default. One Reddit thread complained about him voting in favor of tax loopholes for crypto investors, and another thread noted his “platitudes and self-aggrandizement” and being “safe and middle of the road” instead of decisive. He also still doesn’t have an Issues page on his campaign website (as I noted he didn’t in 2024). Nothing too interesting comes up on a news search for his name, best I can tell it seems like he does mostly middle-of-the-road Democrat things. So, a vote for Amo but not an especially enthusiast one based on what I’ve seen. Based on what happened in 2024, seems like he’ll get 65-70% of the vote, at least.

Governor

D primary:

Daniel McKee

Helena Buonanno Foulkes

Others:

Aaron Guckian (R)

Elaine Pelino (R)

Ken Block (I)

Julian J Smith (I)

(Upcoming debate: August 18, WPRI news)

• Guckian has reasonable bullet points on his website. There’s actually a fair amount of recent news including his name, compared to the Republican candidates in other races. He came relatively close to winning the Lieutenant Governor race in 2022. He’ll put up a fair showing in the general election but it seems like he’s an extreme long shot to actually win.

• Pelino is more of a fringe/MAGA/non-politician candidate (one of her first-year priorities is “fighting for our Second Amendment rights”)

• Ken Block has a recent news release that internal polling has him ahead in the governor’s race, which just seems...absurd (meanwhile his Upcoming Events page is completely empty). I at least value his ability to say that there are some issues he simply doesn’t care about: “No Governor can realistically promise you that they will protect your 2A rights. Any gun control measures would pass with a veto proof majority. I have zero interest in engaging on gun issues. Not for. Not against. Nothing.” There is actually one Block yard sign that I’ve noticed in my neighborhood.

• I couldn’t find a website for Julian J Smith but his Instagram is, uhh, something.

• So, basically as expected, it’s between McKee and Foulkes. Kalshi has Foulkes as an 87-15 favorite to win over McKee. I didn’t realize she was that big of a favorite!

• Here’s a good Reddit thread where commenters share their opinions on why McKee’s approval rating is so low. The Washington bridge debacle obviously being a big reason. Other comments call him variously “feckless” and “spineless” and point out that he was appointed (not elected) governor initially when Gina Raimondo took a job in the Biden administration, then narrowly won a 4-way primary to keep his job—so this will be the first time he faces a 1-on-1 election. It’s also mentioned that he isn’t a very captivating speaker, and specifically people didn’t like the way he handled the Brown shooting incident press conferences.

  • Personally, I’m not completely sure he did anything worse-than-average when it comes to the bridge. He didn’t cause it to fail. It’s getting fixed. People complain that he didn’t fire the Department of Transportation director but, IDK, I don’t love the idea that firing people who make mistakes is the way to fix things. What kind of growth mindset is that? So, I’d love to hear a more nuanced explanation of what he could’ve done better.

• I feel like...I mostly hear nothing about him. Is that because he’s not doing much, or because it’s the nature of being a governor? (i.e. the state assembly is what passes laws, not him.)

• Foulkes has some very detailed plans on her website—the “housing” section of her Issues page is a link to an 8-page PDF.

• On the other hand, The accomplishments page on McKee’s website is one three-sentence paragraph, followed by 6 pictures with a few words of text over them describing some accomplishment, with no hyperlinks or other way to get more information. Pretty disappointing; nor is there anything on there about his future plans as governor.

I’ll vote for Foulkes. McKee has done nothing to impress me (even if, as mentioned above, it’s not as obvious to me as it is to others that he’s bungled the bridge situation). And I’m a sucker for someone who actually has detailed, SPECIFIC plans about what they want to do in the position they’re running for. And at least she’s not in her 70s. The yard signs I’ve been seeing lately are almost exclusively for Foulkes, so I expect she actually is the big favorite at this point that Kalshi says she is.

Edit: Some recent links:

• McKee asks for a pause on Flock cameras. The Reddit commenters call him out for only doing this b/c he’s losing a competitive race. But good on him for doing something the people want…

• Reddit discusses Foulkes. There are definitely people who don’t like either main candidate in this race (“I’ll take the imbecile over the corporate healthcare executive monster. Please think ppl.” is one comment that is fairly well upvoted). The main arguments I hear against Foulkes (both on Reddit and in person):

  • She was a former CVS executive which means both (a) she probably more pro-corporate than the Progressive Dems would want, and (b), she was at CVS during the opioid crisis, and did she really do enough there1? I think this article is a well-researched look at that question. My take: I’m not sure she did anything amazing there, but it’s hard to say she was, like, doing anything evil.

  • She’s never held public office, so a voter for her has to trust that her corporate experience will translate.

Lieutenant Governor

D primary:

Sabina Matos

Sue AnderBois (interview with Steve Ahlquist)

Ross McCurdy

Cindy Coyne

Xay Khamsyvoravong

Others:

John Loughlin

July 31 WPRI debate

• What does a lieutenant governor even do? In a thread discussing Matos from 2023, both of the top comments call the Lt. Gov. role largely ceremonial and useless:

  • “RI LT Governor is weird in that they have no power in government other than if the governor cannot serve anymore. I think many voters on all parts of the political spectrum think the position is a waste of money. Remember in 2010 a third party candidate got 40% of the vote promising to abolish the office”

  • In her interview with Steve Ahlquist, Sue AnderBois says about the job: “The job allows you to say, “This is an important thing, a hairy topic that we need to dig in on. We need to bring people to the table. We need to know how to work the legislature. We need to know how to work with the administration. We need to know how to bring in the community.” The job has room to solve some things”
    …
    [Steve]:We know there are only a few things that a lieutenant government has to do.

    [Sue]: There are three task forces—the committees for long-term care, small business, and emergency management.

  • McCurdy has a page on his website explaining what the job is...although perhaps that’s not surprising given it’s apparently his third election in a row running for the position.
    
    So, yeah, basically it’s a bully pulpit—aside from a few set duties, Lieutenant Governors can kind of work on what they want to, and do that job by the power of persuasion more than by actually having set authority given to them by the state.

• Why are there so many Democrats running for Lieutenant Governor when there’s already a Democratic incumbent (Matos) in the role? I guess for a similar reason to McKee, although she had a stronger showing than him in winning re-election in 2022 (47% of the vote in a three-way primary). She later ran for US Congress in 2023 and lost to Gabe Amo.

• Matos, AnderBois, and Xay are the only Democrats that have any endorsements listed on BlueVoterGuide. McCurdy’s record page focuses on his time as a science teacher, it’s seems that he hasn’t had any political offices before. I don’t love that Xay’s website is apparently lacking any discussion of what he wants to do as Lt Gov (as is the website of John Loughlin, the Republican candidate)

• I think I’ll vote for AnderBois. My thought process being:

  • The role is not very important.

  • She was previously the city councilor for my ward. I thought she did a good job and I know her values are aligned w/ mine generally.

  • Her “priorities” page is probably the most thorough, including a 17-page PDF of her plan for improving the economic structure of our energy grid

  • I think basically all of the candidates are fine, Matos will give you more of the Lt Gov quality we’ve had the past four years; Xay seems like a solid progressive option who was probably #2 on my list; McCurdy is more of a non-politician option who doesn’t seem to have much of a chance but I do think his website is well done at least; Coyne seems like more of a right-leaning Democrat.

I’ve seen the most yard signs for Xay so perhaps he’s the favorite, but it’s hard to say since Matos is the incumbent, and AnderBois has fought to position herself as the most progressive option.

Secretary of State

Gregg Amore is running completely un-opposed, no other Democrats nor any Republicans or Independents.

• I don’t have any strong opinions on him. It seems like he generally gets favorable coverage when he’s mentioned by Steve Ahlquist. He usually comes in and drops off a box of cookies when I’m working the polls on election day (and poll workers are getting a $25 raise this year! I’m rich!). I never love to see someone running completely unopposed b/c I think healthy debate is good for making progress as a society, but I’m happy enough to vote for Amore here. Our elections have run smoothly enough the last few years, as far as I can tell.

Attorney General

D primary:

Jason Knight

Joseph Solomon

Keith Hoffman

Kimberly Ahern

Others:

Alan Gordon (R)

Southside Cultural Center Debate transcript via Steve Ahlquist

Police Chiefs Association “Discussion with the candidates”

WPRI debate

The one poll listed on Wikipedia has 72%(!!) of voters undecided.

• PLEASE read Alan Gordon’s “Investigations and Initiatives” page, which starts with “Rhode Islanders deserve to know whether historical records concerning Nikola Tesla, John G. Trump, architect Stanford White, and the design of the Rhode Island State House contain information of scientific or public importance that has never been fully examined or disclosed.”

• Hoffman, Ahern, and Knight have endorsements listed on BVG (in order of more endorsements to less). (But it doesn’t list individual lawmaker endorsements, of which Solomon has a few listed on Wikipedia)

• Bonus points to Ahern and Knight for having the best Issues pages.

• Backgrounds:

  • Solomon is a lawyer and state lawmaker

  • Knight: “Jason was a prosecutor and special assistant attorney general in the Rhode Island Attorney General’s office, and now runs his own criminal defense practice. Jason has 20 years of experience in Rhode Island’s courtrooms” and state lawmaker

  • Hoffman: “senior leader in the Rhode Island Attorney General’s Office”

  • Ahern: “served as a Special Assistant Attorney General under three Rhode Island attorneys general” and “senior counselor to Rhode Island’s past two governors” and “Governor McKee appointed Kim as the first-ever Chairperson of the Rhode Island Cannabis Control Commission.”
    

• About that...Steve Ahlquist thinks the CCC is not doing a good job:
  “Upon taking the helm as CCC Chair in 2023, Ahern told Steph Machado at the Boston Globe and Channel 12’s Tim White and Ted Nesi that the CCC would issue 24 new retail licenses (including the 12 social equity and co-op licenses) so that stores could open in 2024 and that the CCC would establish the rules and regulations process within a few months. Three years later, none of this has happened. Perhaps Commissioner Ahern had a change of heart, since in September 2025, she told the NBC10 I-Team, “You’ll see a slow, I think, and thoughtful and deliberate expansion of the additional retail stores” in Rhode Island.”

• I find it hard to say there’s much clear difference between these candidates. The 9th question in one of the debates linked above is “What is one way that you would distinguish yourself on a policy issue from the other people here on this stage?” My notes on their answers is:

  • Ahern doesn’t say much of use and just talks about the “type of AG she will be”

  • Knight says he’s the progressive option: “When it comes to fighting back against the Trump Administration’s attacks on reproductive justice and fighting for our trans communities, I definitely distinguished myself by fighting, as far as I have been able, every day for the last 10 years as a legislator.” (etc)

  • Solomon highlights his breadth of experience “I’m the only person here who’s been a prosecutor, a bail commissioner, a legislator, and a criminal defense attorney.”

  • Hoffman highlights his work on the Public Protection bureau and his support for an assault weapons/high-capacity magazines ban.

• In the Police Chiefs discussion, they ask about “top 3 priorities”. Here are answers:

  • Ahern says “Crime, funding, and recruitment.” (i.e. funding and recruitment for police officers - “Obviously, everyone in this room could use more funding.”)

  • Hoffman says violent crime, public safety/gun safety and fighting the Trump Administration.

  • Knight gives an answer that it’s hard to determine a top three, but includes “violent crime as number one, for sure... drugs, human, and occasionally weapons trafficking...traffic safety...[police] recruiting”

  • Solomon: Doing more for repeat offenders, mental health, police funding.
    
    

• None of the candidates were willing to take a strong stand against the Flock cameras that have suddenly become a hot political issue. Here’s a quote from each of them from the linked forum:

  • “I support the use of Flock cameras.”

  • “You’re going to be hard-pressed to find a prosecutor who doesn’t want one more tool in the toolbox as Flock provides.”

  • “I completely support the use of Flock-like cameras and other surveillance systems, but...”

  • “It’s hard to talk to the regular person about how effective they are in helping with the job of law enforcement...I think it’s a great tool.”
    
    I’m not telling you who said which quote because it doesn’t seem like it matters. They all do go on to make comments about how we have to go about it the right way, etc etc, but none of them were willing to flat-out say “no, we don’t want Flock cameras”.

OK I think I’m running out of energy here. Final thoughts:

• I’m down on Ahern b/c of the CCC stories I’m hearing (friends drive to another state to buy weed! so this is a personal issue!) and b/c she lists more police funding as one of her biggest priorities. That said, to the extent I’ve seen any yard signs for the AG race, the majority have been for Ahern so I expect she has a solid chance even if I’ve personally found reasons not to vote for her.

• The other three are about even, with maybe Solomon in 3rd behind the other two. (His experience seems perhaps a tad less impressive than the other two, plus there are vague allusions in the debates linked above to progressive policies he didn’t used to support but has since come around on.)

Personally I sorta feel like I could flip a coin on Hoffman vs. Knight. Let’s say Hoffman I guess, but without much conviction. (What are my reasons? IDK they mostly seem like similar options...I guess b/c he’s ~20 years younger and his current experience in the AG’s office could mean a somewhat smoother transition)

General Treasurer

D primary:

James Diossa

Others:

Micholas Credle

• Credle doesn’t seem to have a campaign website. Here’s a Facebook page; I don’t have Facebook so I can’t tell you much about what’s on it. Here’s a Providence Journal story. He doesn’t really seem like a serious candidate.
  

• What does a General Treasurer even do? How do you know if they’re doing a good job or not? Diossa’s page on ri.gov says:

  Under his leadership, Treasury has launched several initiatives to support Rhode Islanders, including RISavers, which expands retirement access to the 40% of private-sector workers without employer-sponsored options, and the RI AnchorHome pilot program, which offers low-interest home loans. Treasury has also expanded efforts to return unclaimed property, reuniting residents with millions of dollars. In addition, Treasurer Diossa has strengthened financial education statewide, convened municipal leaders on public finance, and launched a medical debt relief initiative that eliminated $12 million in debt.

• No one has posted anything about him on Reddit since before his 2022 election as far as I can tell. It seems the vibe at the time was that he was a bit under-qualified (for the position of Treasurer at least), but there hasn’t been any bad news since then at least.
  

I don’t have any strong feelings for Diossa but nor do I have any reason to vote against him, and Credle doesn’t seem like a serious candidate. I’ll go with Diossa and I’m sure he’ll win in a landslide.

Senator (state level)

D primary:

Meghan Kallman

Vanessa Lopez

I like my current State Senator Meghan Kallman a lot. From everything I’ve heard about her it seems like her values align with mine quite nicely (sorry for not linking specific examples, here’s her news page). The only thing I don’t like is that she seemingly hasn’t updated her campaign website since the last election.

Lopez, as I understand it, is basically a Republican running in the Democratic primary in order to have a chance. Admittedly I don’t see any objective proof that she’s “actually a Republican”, but he Issues page includes things like “repeal the high earners tax”, “My commitment to public safety has earned the endorsement of the Rhode Island Fraternal Order of Police”, “Uphold the Constitution - not selectively, but fully.”, and “I believe diversity of thought makes our communities stronger” (not that I disagree! It’s just that it sounds like a dog whistle when she says it on her campaign page...)

I’ll vote for Kallman, enthusiastically.

Representative in the House (state level)

(District 4)

Rebecca Kislak

• As far as I can tell, Kislak is running unopposed in both the primary and general election.

As with Kallman, Kislak is pretty closely aligned with my values (here’s her news page) and I’ll happily vote for her.

Mayor

D primary:

Brett Smiley

David Morales

RI Current Q&A

Reddit link to podcast debate (with lengthy text summary in the comments)

Community Library Mayoral Forum

WPRI debate

• I’ll try to summarize their policy difference, as per their respective (linked) Issues pages:

  • Housing (1st on Morales’s issues page, 5th on Smiley’s): Both support increasing the supply of housing. Morales supports rent stabilization proposals, Smiley says they don’t work. Morales says he does “not accept donations from corporate landlords and real estate interests.” Smiley says PVD has increased its housing production an impressive amount under his administration.

  • Safety (1st on Smiley’s issues page; 7th on Morales’s): Smiley likes talking about how he got rid of people riding ATVs on the streets (he mentions this both in his RI Current Q&A and near the top of his Issues page). He likes pointing out how Morales has previously made pro-defunding-the-police comments. But Smiley also has some “left wing” takes on safety, like a recently-established EMS bicycle unit, “the best way to prevent crime, improve quality-of-life, and support our local economy is through immediate and significant investments in opportunities for youth”, etc.
    
    Morales focuses on similar safety-through-community-building topics, and points out that “the civilian oversight body that investigates allegations of police misconduct (PERA) is a critical agency. Yet, under the Smiley Administration, PERA’s policy recommendations have largely been ignored and the department’s budget has been slashed by nearly 30%.”

  • Protecting Immigrant communities (2nd on Smiley’s page, 5th on Morales’s). Smiley has signed orders to ban police from cooperating with ICE and banned ICE from staging on city property. Morales says he will “establish a dedicated task force...to share accurate information on ICE sightings and develop plans to protect our neighbors.”

  • Schools (3rd on Smiley’s page, 3rd on Morales’s): I don’t really see anything here clearly setting them apart in an obvious way. Both want to invest in our youth...

I generally think of politics as something that’s not personal to me. I don’t generally know any politicians and I believe I’ve never donated to a campaign. But then the mayor of Providence recently called a friend of mine a “crazy person”. I don’t think Jackie and I share the same opinion on every political issue, but I think it’s incredibly obvious that Jackie really does care and, to make an analogy to some of the news I see online about YouTube star Ms. Rachel, Jackie’s only crazy if crazy means believing things like “we shouldn’t let children die”, and believing it enough to actually do something about it, even when it’s uncomfortable and inconvenient.

Overall thoughts:

• Morales is pretty much Providence’s Mamdani-clone, down to the “roughly 5 years in state legislature before running for mayor”. (Though he’s a solid few years younger than Mamdani). To be honest I do worry a bit about his qualifications/life experience, but the fact I seem to be seeing mostly good news about NYC gives me a bit more confidence that Morales likewise could handle the job.

• I personally don’t think rent control is the “theoretically optimal” solution to our housing crisis. But I don’t think either candidate is as gung-ho to reform the zoning code as I am. And if that’s not happening, I like the idea of at least trying to help the working class.

• On Safety, Smiley mentioned our record low homicide rate but I doubt it’s due to his brilliant leadership given that the US homicide rate is also at a record low. (Just today the NYT did a story about Baltimore’s historic-low homicide rate.) Is he really providing any marginal benefit over whatever effects and trends are prevailing throughout the country? I’m not super-anti-police but I HATE abuse of power so I don’t love that Smiley has defunded the oversight board.

  

  Can we just have good public safety and strong police oversight, please? I don’t believe the only way to have a safe city is to let the police get away with it when they’re being thugs.

• Last time, Smiley ran on promising effective city services. Smiley actually has a paragraph on his page now bragging about how his administration handled this winter’s record-breaking snow. Which is frustrating because the city actually never plowed my street—it was eventually cleared by a neighbor with a private plow truck. To me, this makes him feel a bit more like “typical politician”, where nothing ever goes wrong. I’m not saying Morales would’ve done any better! I’m just saying the way he talks about it annoys me (though his response in the WPRI debate was much more reasonable).

Overall, I think I’ll vote Morales. I have some worries about his levels of experience, and I have reservations about rent stabilization, but I can’t helping but wanting to try to get to a better place than we are now and I’m hopeful Morales can follow in Mamdani’s footsteps. And I can’t shake the feeling that Smiley is too much “typical politician”, too much “can’t help but care the most about his upper-middle-class-to-upper-class constituency”. Also the whole “calling my friend a crazy person” thing.

It looks like a close race currently (Prediction Market, NYT poll aggregator), with Smiley perhaps a bit ahead. Yard-sign based analysis leads me to an identical conclusion, I see many signs for both.

But also shout out to both of these two for having incredibly detailed campaign websites compared to a lot of the others profiled in this post! (What a little healthy competition does to a guy...)

Edit: One story I came across since posting that had missed from earlier in this year was that Smiley built a $4-million home a half-hour from the city. Mixed, slightly negative, feelings about this for me - I generally don’t personally begrudge people for their success, but I do think it really reinforces the conception that he’s a little too landlord-class/a little too fancy to really be down with the rest of us.

City Council (Ward 3)

D primary:

Yulyana Torres

Jason Kashdan

Steven London

(Shout out to Yulyana for getting the “yulyana.com” domain name!)

Steve Ahlquist transcript of Ward 3 candidate forum

Edit: See also, Yulyana Torres interview with Steve Ahlquist. A friend texted me some feedback on this race, but none it was things I was able to find hard sources for (perhaps common in such local races) so I won’t add those details here unless I can find official news.

There is no incumbent, as our previous city councilmember is leaving the position to run for Lieutenant Governor (see above!).

All three candidates seem nice, and there aren’t too many HUGE differences between them (in the linked candidate forum, one of them says “I honestly believe that any of us would be incredible in this role.”). All three candidates are, far as I can tell, running for a political position for the first time (though with various other types of community-service-based experience).

The one thing that set a candidate apart for me was Jason Kashdan mentioning in the linked forum that he supports a split-rate tax / land value tax, a favorite fringe policy of mine:

There are options we have that haven’t been tried here in our city yet, with taxes. Something I’m excited to bring on board is the split-value tax, or land value tax...That encourages development. It can encourage development with the character of the neighborhood.

Guessing based on vibes, I expect Yulyana to win, Kashdan to come 2nd and London 3rd. Those vibes are mainly:

• Yulyana is on the primary ballot as the party-endorsed candidate and has the endorsement of the outgoing councillor

• I’ve seen the most yard signs for her. (I’ve seen at least one yard sign for Kashdan, but none for London)

I think I was originally leaning towards voting Yulyana—mainly b/c she feels a bit more homegrown than Kashdan which is about all I can go on when there’s not much setting them apart policy-wise. But...I can’t help but be excited for the chance to vote for someone who supports LVT, so I’m gonna vote Kashdan, though I’ll be perfectly happy with it when Yulyana wins.

Ballot questions

I can’t see them on a sample ballot b/c the state website hasn’t uploaded sample ballots for the general election yet (these will be on the November ballot, not the September primary), but Ballotpedia says there will be 5 questions on the issuing of bonds to pay for projects, with some details in the image below:

If this is giving you deja vu, it’s because some of these measures are reminiscent of the measures that passed in 2024:

It’s even the exact same amount of money—$120 million—that we approved towards housing in 2024 (but with different intended uses).

Note that in her Housing Plan (PDF warning), Helena Buonanno Foulkes cites an analysis that the $120 million approved in 2024 was expected to generate “fewer than 600 homes”. So...if you’re voting for Foulkes does that mean you should vote against providing another $120 million to housing, given that she herself complains about how it went last time?

As I pointed out in 2024, Rhode Island has apparently approved every single bond measure since 2006.

At least with the higher education funding measure, it’s easy to understand exactly what the money is being spent on:

(As opposed to the housing measure where all the detail we get is: “A “yes” vote supports issuing $120 million in bonds to fund the construction and development of affordable housing projects...at least $25 million must be allocated to increasing the production of housing specifically intended for homeownership. Additionally, up to $10 million of the $120 million could be used to support a public housing development program.” But what are they actually, like, spending money on?)

I think the one I’m least interested in personally is the Historical Center:

IDK, do we really need a new state history center? I highly doubt I would ever go there but maybe that’s just me being biased about the types of cultural things I’m interested in.

I think I’ll vote a strong yes on the higher education facilities and watershed protection (because they’re more specific, and seem more like problems best solved by direct government funding), and a weaker yes on the infrastructure development and housing questions (because they’re less specific, because Foulkes says the last housing bond underperformed, and b/c I tend to believe changing our laws is a better way to encourage solutions to the housing crisis & spur the economy than using funding).

I’ll vote no on the Historical Center just b/c it’s not something that interests me much and I’ve seen no complaints elsewhere that our historical culture is underfunded; but given recent trends we should probably expect all five to be approved.

It’s not listed on Ballotpedia yet as far as I can tell, but I know from the news that the city of Providence will have a ballot question on divesting their municipal funds from foreign countries complicit in ethnic cleansing, genocide, etc. It’s written in a country-agnostic way but is very clearly intended to “voice our support for a free Palestine” (quoting from the linked article).

Mayor Brett Smiley says “the City’s legal department has determined that this ordinance is both unenforceable and likely in violation of existing state law.” However, the article goes on to note that that law has been criticized for being potentially unconstitutional: “The ACLU testified that Rhode Island’s law “potentially chills the legitimate free speech activities of contractors.””

I’m going to vote yes. I don’t doubt it would be very complicated—maybe impossible—to actually make it work given the way investment funds work. And I know a lot of companies do some pretty immoral stuff even if it doesn’t usually reach the level of genocide—so even if we divest from some companies we’ll probably still be often investing in some questionable behavior. But I think we should at least try to engage in a morally-better version of capitalism and not just give up before we even make an attempt. I don’t find Smiley’s “don’t even bother trying to have morals, the system won’t let you” argument against voting yes very inspiring.

I think alternative corporate structures (like B-corporations -- the legal entity, not the certification) could be a powerful tool in a society that embraced them, and choosing not to invest in especially immoral companies feels like a strategy in a similar vein.

Note: I think there’s going to be a few weeks in a row of posts that aren’t strictly about frisbee strategy coming up (though some, including this one, will be at least tangentially about frisbee). If frisbee strategy is mainly what you’re here for, I appreciate your patience as I work through some of these other topics.

Another summer league season complete, let’s take a look at how everyone’s predictions did.

As a refresher, here are the preseason predictions:

And here are the averages:

(Note: “Humans” = LT, Sammy, and Sam. “Oliver” went in the “Machines” category here b/c his method is purely data based.)

And the average of all six predictors:

Results

Here’s how everyone did:

Yellow was involved in the two single worst guesses, with me putting them 15th and Sam 16th (They had a bunch of new-to-the-league players who I guess were better than Sam & I expected). Maroon was a weak point for the computers, while Silver was a weak point for just about everyone besides Sammy. No one was off by more than 2(!) for Navy (nor for Orange in 9th!), and not a single predictor was off by more than 3 for Black.

Here are the results for the averaged predictions1 :

I’m pretty impressed with how the average of everyone was off by two spots or less for 12 of the 16 teams!

It’s also interesting to see how adding a “worse” prediction (Humans) to a “better” prediction (Machines) produces a prediction that’s even better than the “better” prediction. As mentioned in the predictions post, this makes some sense since the machines and the humans have somewhat different areas of knowledge.

The Humans were too high on Pink and Gold, for example, but the Machines were too low on them and it somewhat canceled out. The humans were more wrong on average than the machines, but, more often than not, they were wrong in opposite directions.

Here’s how the six predictors (and the averages) look statistically, across a few different methods of measuring:

As mentioned in previous editions of this post, the Spearman coefficient is a more nuanced mathematical measure of the correlation between the predictions and results. The “correct pairs” column measures each possible pair of two teams, and checks whether, if A was ranked higher than B in the predictions, did A finish higher than B in reality.

Everyone had a Spearman coefficient well over 0, which means everyone’s predictions have at least some positive predictive power.

A few final thoughts:

• Congrats to Sammy for being the best of the human guessers! (And so far the only one to beat Oliver’s model in any year, something I still haven’t done)

• While my predictions weren’t amazing, here’s a few things I’m at least somewhat happy about:

  • For the second year in a row, the team I put at #1 finished outside the top four…but for the 2nd year in a row, they also made it to the finals in postseason play.
    (In fact, both of this year’s finalists, Maroon and Gold, were in the top 4 of the human rankings, while they were ranked 6th and 9th by the machines.)

  • I correctly predicted that the average of all 6 predictors would be the best prediction method (if “There’s a good chance the full average outperforms any of us” counts as predicting it).

  • I correctly predicted that I’d do better than last year (“I said last year (2025) that I was less confident in my predictions than [I was] in 2024. This year my meta-prediction is that I feel more confident than I did in 2025.”)

• For the 4th year in a row2, my predictions and Oliver's have by fluctuated a similar amount, so it still seems like there’s some amount of overall “predictable-ness” that’s inherent to each year’s league (related to maybe…number of new players, or number of new captains?):

  

I think I can beat the machines someday…there are some places where it feels like they “get lucky” a little bit, and they might not always have that luck. But this year wasn’t the year.

If you know me in real life & you’d like take part in providing predictions next year, just let me know!

Anyway, here are some books I read between now and...the end of 2024. As always, feel free to skim to find the things that seem interesting to you personally, or skip it entirely if it’s not for you.

I got a little too far behind on writing reviews, and I think I just need to admit I’m a little too far behind to review everything. So here’s a few books I read in that time period but you won’t find a longer review of below:

The Barcelona Way: Unlocking the DNA of a Winning Culture, by Damian Hughes

Supercommunicators: How to Unlock the Secret Language of Connection, by Charles Duhigg

The Land Trap: A New History of the World’s Oldest Asset, by Mike Bird

The Panama Papers: Breaking the Story of How the Rich & Powerful Hide Their Money, by Frederik Obermaier & Bastian Obermayer

The Hidden Life of Trees: What They Feel, How They Communicate — Discoveries from a Secret World, by Peter Wohlleben

The Explorer’s Gene: Why We Seek Big Challenges, New Flavors, and the Blank Spots on the Map, by Alex Hutchinson

The Unaccountability Machine: Why Big Systems Make Terrible Decisions — and How The World Lost its Mind, by Dan Davies

On the Edge: The Art of Risking Everything, by Nate Silver

Tribal: How the Cultural Instincts That Divide Us Can Help Bring Us Together, by Michael Morris

Spaced Out: How the NBA’s Three-Point Revolution Changed Everything You Thought You Knew About Basketball, by Mike Prada

The Midrange Theory: Basketball’s Evolution in the Age of Analytics, by Seth Partnow

Burn: New Research Blows the Lid Off How We Really Burn Calories, Stay Healthy, and Lose Weight, by Herman Pontzer

Nature Wants Us to Be Fat: The Surprising Science Behind Why We Gain Weight and How We Can Prevent—And Reverse—It, by Richard Johnson

There’s some good stuff in there too, sadly. Maybe I’ll come back around to some of it, someday. I wanted to write a review of The Panama Papers in the context of this meme. Burn was pretty interesting. Nature Wants Us to Be Fat was too, but edged a little closer to feeling like crackpot science than Burn did. The Unaccountability Machine was interesting but I was somewhat underwhelmed by his suggested fixes. The Barcelona Way and Supercommunicators are on the list to re-read before I start my next coaching job (whenever that may be). Spaced Out and Midrange Theory are both great but don’t have too much you won’t also find in a Thinking Basketball video on YouTube (but the books came first, to be fair). A long review of The Land Trap can be found here (land use/urbanism is an interest of mine, see some previous book review posts).

On to the reviews:

NONFICTION

Kingdom of Characters: The Language Revolution That Made China Modern, by Jing Tsu

Although this book got some pretty good reviews, I never really got into it. I think someone could make a really brilliant book about Chinese characters for English-language readers. But this book, for the most part, felt like I was reading a history book—the “characters” in the title are the people inventing things and going to the conferences where decisions were made about the Chinese script, not the symbols of the Chinese language.

Almost made it on my list of “books I couldn’t finish”, but I slowly kept coming back to it when I wasn’t reading anything else and eventually finished it after 4+ months.

Growth: A History and a Reckoning, by Daniel Susskind

I was disappointed by Growth. The subject matter could make for a book that I find really interesting. But I felt like...it didn’t really have any insights that you couldn’t get from some dude at a bar who’s read a couple economics-related articles online. Growth is good, but it also has downsides. We should care about GDP less, but actively trying to shrink the economy is a step too far. He doesn’t make any strong recommendations, which I can respect, but it also means he doesn’t have anything to say that’s all that thought-provoking.

I also felt like Susskind didn’t actually do that good a job explaining the downsides of growth. To partially explain why I say that: I’m someone who lives in a first world country but my expenses are 1/3 to 1/4 of the US median income. But I’m perfectly happy—in fact, I probably have a more fulfilling life than the average American. So...are those 10,000 dollars/year that other people are spending on their cars (for example) actually buying them $10,000 worth of happiness? From my perspective, the answer is obviously not. The claim that “degrowth would undoubtedly make things worse” just feels like it’s been proven untrue by my own personal experience. People don’t need all the money they have to be able to afford happiness, they’re just...stuck in an inadequate equilibrium. I’m not going to be fully convinced by everyone claiming degrowth is bad unless they take the time to explain why other people living the way I do would be such a horrible thing.

In the “history” part of the book, he makes the point that economies didn’t really grow, for basically thousands of years—from the start of recorded history, up to about 1800 or so:

The idea of pursuing growth would have been unimaginable to most classical economists...

Clark, who calculated English living standards back to 1209, cleverly solved this problem [of comparing across vastly different times] by re-expressing workers’ wages in terms of the amount of wheat they could buy, an approach similar to Allen’s subsistence ratios. These “wheat wages” for nineteenth-century English laborers he found to be roughly the same as those four thousand years ago in ancient Babylonia and Assyria—a sign that the Long Stagnation stretched far back in time.

...“Primitive man,” Clark concludes, “ate well compared with one of the richest societies in the world in 1800.

I’m going to totally skip over the details here but basically Susskind says it’s ideas and our new relationship to ideas—brought about by the Scientific (and Industrial) Revolution—that led to increasing growth:

In his view, it was a profound change in culture during the eighteenth century that supplies the necessary connection...

...What took place is far deeper: the discovery of a way to find new ideas and its widespread adoption. “To paraphrase Alfred North Whitehead’s famous statement,” writes Mokyr, “the Industrial Revolution did not mark the beginning of invention, but it marked the beginning of the ‘method of invention’—turning it from an occasional event to a regular and routine part of the economy.”

...how we raise productivity is therefore the same as asking how we generate more ideas.

One of his ideas for ideas is to make better use of our human capital. I liked this pro-growth-mindset study result:

When kids have similar math scores in early childhood, for instance—a useful predictor of their future inventive ability—race, gender, and parental income still play a big role in determining whether they become inventors.

Come Together: The Science (and Art!) of Creating Lasting Sexual Connections, by Emily Nagoski

...Peggy replies, “Well, I rather enjoy sex, but if I was having the sex you described, I wouldn’t want it either.”

I’d say Come Together was maybe...less unexpected than I’d hoped. The ideas weren’t too surprising, which is not to say that they’re not useful.

She makes a few main points. First, we should worry less about what’s “normal” and instead value what works for us:

Some variation on “Am I normal?” is the most common question I’m asked....when you replace “Am I normal?” with “Am I confident and joyful?” your whole outlook on your sexuality shifts.

And not unrelatedly, we should center the idea of pleasure—if it makes us feel good, it is good (obviously stressing consent whenever other people are involved):

Assume everything you were taught about sex for the first two decades of your life is incorrect, and start again, with pleasure as your only “measure” of sexual well-being.

To summarize some of the main ideas in her own words:

I’m happy to give away the ending right here at the start. The three characteristics of partnerships that sustain a strong sexual connection are:

1. They are friends—or, to put it more precisely, they trust and admire each other.

2. They prioritize sex—that is, they decide that it matters for their relationship.

3. Instead of accepting other people’s opinions about how they’re supposed to do sex in their partnership, they prioritize what’s genuinely true for them and what works in their unique relationship.

She often uses the metaphor of “accelerator and brakes” for understanding sexual interest—people want to be turned on, but they also want to be not turned off.

The brain mechanism that governs sexual response has two parts, the accelerator and the brakes. The accelerator...sends the “turn on” signal many of us are familiar with...All kinds of things can activate the accelerator, from the sound of your partner’s voice to the sight of a sexy body part or watching them do something completely nonsexual that makes you feel admiring of them and lucky to be their partner.

The brakes notice all the good reasons not to be turned on right now—everything you see, hear, smell, touch, taste, think, believe, or imagine, plus your internal body sensations that your brain codes as a potential threat—and it sends the “turn off” signal...relationship conflict to stress and exhaustion to trauma or body image issues to sex negative cultural messages to worry that the kids will interrupt to irritating grit on the sheets.

And she suggests “Aim[ing] for the “Room Next Door” to LUST” (in the metaphorical “house” of one’s different emotional states). In other words, too many people go wrong by skipping right to lust/sex with their long-term partner. Instead, aim for the room “next door”—e.g. situations with a shared sense of playfulness, or loving warmth, or feeling home and safe. From there, lust can become a natural next step.

It was sort of a weird book, I was often left with the impression that I was reading a self help/couples’ therapy book about communicating with your significant other, etc, then all of a sudden the next sentence would be about [redacted explicit sexual scenario that could be a lyric in WAP]. I didn’t save any examples, but you can imagine.

Random quote that taught me something:

If you weren’t raised in the white Western world, you might have been raised in a culture with a different number of gender roles. Traditional Samoan culture has three but also four. These things are complicated and don’t map neatly onto Western, white binary concepts. Traditional Navajo culture has four (or five—again, these things really don’t translate). Traditional Bugis culture, in Indonesia, has five gender categories, including gender transcendent, androgynous shamans. Pre-colonial Yorùbá culture had people with different reproductive roles, of course, but no distinctions of gender.

:07 Seconds or Less: My Season on the Bench with the Runnin’ and Gunnin’ PHOENIX SUNS, by Jack McCallum

“Shawn’s going to be a monster tonight,” says Griffin. “He has the I’m-a-bad-motherfucker factor going for him.”

Back when I was in high school, I heard about this book coming out and really wanted to read it, but my library never ended getting a copy. But I finally got around to reading it 15+ years later. In terms of learning how the Suns’ offense worked, 7 Seconds or Less didn’t really get into the nitty gritty details. It’s more of the story of the Suns’ season—the wins and losses, the personalities, etc. Here’s a few parts I noted down.

The funniest part of the book was this sentence:

Bryan Colangelo is wearing, as is his custom, a suit and custom dress shirt with a high collar. (No tie.) Dick Van Arsdale, a former Suns player and now a team executive, calls it the “Flying Nun look.”

For anyone who isn’t aware, there’s a story behind this. Well, actually, there’s a story ahead of this. In 2018 (~10 years after this book came out), an investigative report by The Ringer—The Curious Case of Bryan Colangelo and the Secret Twitter Account—revealed the Colangelo, who had since become the GM of the Philadelphia 76ers, was outed as being the owner of various anonymous Twitter accounts, which had discussed team strategies, criticized players, etc. Colangelo was forced to resign from his position.

But perhaps the most lasting part of The Ringer’s article was this non-basketball anecdote:

Several posts wandered from organizational matters, debating the size of Colangelo’s shirt collars. When user @Philly_Asshole commented on the executive’s well-documented sartorial preference, “This dude just love collars,” Enoughunkownsources responded, “That is a normal collar. Move on, find a new slant.”

The phrase “find a new slant” has since become a lasting meme in the NBA-sphere, for example with the following picture posted to Reddit with the caption “That is a normal collar. Find a new slant:

All that is to say, it was a funny blast from the past in 2025 to read in a book from 2006 about an NBA GM’s collar-obsession that played a bit role in a 2018 scandal. Apparently Bryan is really into vitamins, too:

“They got a lot of diseases up here, B?” D’Antoni asks, his eye on Colangelo’s heroically sized vitamin bag.

I liked this quote about shooter Eddie House:

House and his teammates loved it when Noel Gillespie found the scouting report of an opposing team that said this about House: Won’t shoot it unless he has it in his hands.

And this quote, of a Suns opponent talking about their style:

Out in Los Angeles, that noted basketball scholar, Kwame Brown, has weighed in with his assessment of the Suns’ offense. “They’re not a fundamental team,” Brown said. “They just go out and they just run a bunch of screen-and-rolls and have such good shooters.” That is quite a telling comment. In Brown’s world, “fundamental” equates not to movement and spontaneity but to isolation plays and set offense. It’s not his fault, really. It’s the basketball world in which he’s grown up.

McCallum himself obviously notices the tension here. But now, nearly twenty years later, it feels all the more ridiculous. At this point, there’s hardly anything to NBA offense but pick and rolls and shooters—it’s so standard that Thinking Basketball on YouTube did a video in December 2024 titled “The NBA team NOT running pick & rolls (!)” — now not running pick-and-rolls is a clickbait headline.

And I liked this quote about Steve Nash and motion—feels like something that applies to playing frisbee, too:

One interesting question about Nash’s game is the extent to which athleticism plays a part. Half the basketball world thinks he’s an incredible athlete, the other half thinks he’s the classic slow-footed, overachieving Caucasian.

...Nash is not that quick. On his own team, Marion, Stoudemire, Bell, and House are all quicker, never mind Barbosa, who’s twice as quick. But Nash has great feet. “Once I get going, I can do a lot of things. But I’m painfully bad at explosiveness.” What Nash has done, then, has mastered ways to always be moving slightly. The Suns’ offense is predicated on that principle, too, even in the half-court. Nash gives it up on the run and gets it back (via pass or dribble-handoff) on the run.

Anyways, read this book if you think you’d like reading the story of a famous, innovative basketball team’s season; don’t read it if you’re hoping to learn a lot about basketball theory.

Think - A Compelling Introduction to Philosophy, by Simon Blackburn

I struggled to get into this one. I did make it to the end, but without making too many highlights. One of my few highlights was this following anecdote about Bayes Theorem:

When the call-girl Mandy Rice-Davies was told that some member of the aristocracy denied having had an affair with her, she replied, “Well, he would, wouldn’t he?” She was in effect reminding people that the antecedent probability of this particular piece of testimony was high regardless of which hypothesis is true, and this undermined its value as evidence.

If you’re looking for an introduction to philosophy book, I’d personally recommend Philosophy: A Complete Introduction, which was one of the books I reviewed in this post.

Metazoa: Animal Minds and the Birth of Consciousness, by Peter Godfrey-Smith

Peter Godfrey-Smith wrote another book about octopuses that I really liked, called Other Minds. When I saw he had another book now, I was tempted to try it. Metazoa is kind of about the evolution of minds. Here’s how he puts it in the first chapter:

In some ways this book is about the evolution of subjectivity—what subjectivity is and how it came to be. Subjects are the home of experience, where experience lives.

I will also talk sometimes just about the mind, as I think that is what we come to understand through this story: the evolution of the mind and how it fits into the world. I’ll move between terminologies without laying down the law. Our present understanding is not good enough to insist on one language or another.

The project I am trying to advance can be described in a number of different ways, but it is difficult no matter how we look at it. This project is to show that somehow a universe of processes that are not themselves mental, or conscious, can organize themselves in a way that gives rise to felt experience. Somehow, a part of the world’s often-mindless activity folded itself into minds.

He asks us to accept that there’s a spectrum of consciousness/”there-ness”/subjectivity, and to not see it as an all-or-nothing question:

Though gradual transitions look natural on the biological side, it has seemed to others that consciousness, or felt experience, cannot be like that. The lights in an animal (ant, jellyfish, squid) are either on or not: yes or no. I’ve been struck in recent discussions by how many philosophers want to insist on this; people say there is no possible way to see the existence of experience itself as a graded, more-versus-less matter.

This picture is one to resist. If the features I’ve been describing are the basis for experience, and these shade off into partial cases, then subjective experience does too. Subjective experience is not something that comes into being at some stage and from that stage is just there; rather, its thereness is a matter of degree.

The book sort of goes through evolutionary history showing how different small pieces of the biology that leads to “thereness” developed, highlighting different animal(s) each chapter.

To be honest I’m writing this review a bit too long after finishing the book, so I won’t get too deep into the details here. One passage that did stick in my mind is that he actually says he thinks octopuses are not that clever (in a way, at least...he kind of says they have a physical cleverness more than a mental cleverness—they’re good at using their bodies to explore):

People often now talk about octopuses as “smart,” and in some ways they are. But that is not the term that comes readily to my mind...Octopuses are exploratory animals who direct the complexity of their bodies on whatever confronts them. They fiddle about and try things and turn the problem over and over—physically, not mentally. Octopuses have an extraordinary sensorium and an anarchic bodily embrace of novelty, but they are not, for the most part, ruminative and “clever” sorts of animals.

Anyway, here’s some other animal facts I liked:

The “mirror test” is a test used to work out whether an animal can recognize itself in a mirror as itself, for example by grooming or cleaning a spot on its body that can only be seen in the mirror. Very few animals pass this test. The only animal reported to pass a version of it, so far, that is not a mammal or a bird (and only very few of those pass it) is a cleaner fish.

Amid the fidgeting of all the limbs [of a banded shrimp], it took me a while to work out what was there...Looking at photos, the usual toolkit seems to be something like six claws, four legs, six feelers, and two other comb-like bits of paraphernalia. That makes eighteen limbs and protrusions, at least. A body like a Swiss Army knife.

This emptied-out-toolbox appearance is fairly typical of arthropods. If we are building a hermit crab, why not put some things that look like spatulas on the top of the face? Why not? That is the arthropod way of evolving: when in doubt, add some legs. Add some spatulas to your head.

Their [flatworm] bodies achieve a lot with a little. There is no “through gut”—what goes in one end goes out the same end—and no circulatory system. A cyclops-like eye cluster sometimes sits in the middle of the back, though eyes, of a very simple kind, seem to pop up just about anywhere on these animals, including on tiny pinches of the tissue-like body, raised a little above the plane of the rest.

Flatworm sex lives are more complicated than one might expect. They are generally hermaphrodites, and some engage in “penis fencing,” trying to dart sperm into the body of another, while the partner does the same thing.

A bee, he says, has a brain that is a cubic millimeter in size. That seems tiny. But, he adds, a single bee neuron can have the complexity of a full-grown oak tree, with respect to its branching. Each neuron can connect to ten thousand others.

Superbloom: How Technologies of Connection Tear Us Apart, by Nicholas Carr

If you don’t live by your own code, you’ll live by another’s. [get it?]

I was tempted to write this book review as its own post, to clickbait y’all into thinking it was a post about frisbee given there’s a pro ultimate team of the same name. But I didn’t end up having the motivation for that; so here’s the short version. To be honest I’m writing this book review a bit too long after reading the book, and his thoughts about social media have sort of blended together with everything else I hear about the downsides of social media when I’m simply existing online. I think his argument was something like this:

First, while early “social media theorists” thought access to information would be a democratizing force, it turned out that there’s just so much information out there now that we can’t deal with it:

Charles Cooley, in arguing that greater media efficiency widens our influences and broadens our minds, took it for granted that new communication technologies would make information more accessible without altering its expression or interpretation. As the mechanisms of communication gained speed and capacity, people would draw from a larger, more various set of works; they would think deeply about what they had read; and they would learn from the good and discard the bad. What he failed to anticipate was how media efficiency, when pushed to an extreme, so accelerates the flow of information that people no longer have the luxury of careful reading, methodical evaluation, and contemplative inquiry…

And that while there’s this idea that communication is super important and we can all understand each other if we just communicate, that’s not necessarily true in the flood that is social media:

Sherry Turkle, an MIT social psychologist who has been studying how people communicate through computers for decades, has described social media as an “anti-empathy machine.”...Although we’re quick to assume a direct correlation between communication and empathy—the more information we exchange, the more empathy we’ll feel—empathy depends far more on the quality of our communication than its volume. It’s another thing we sacrifice when we talk too much.

While the original science of communication theory was built around the idea of being smart with redundancy and error-fixing to optimize for the limited bandwidth of our communication infrastructure, in a world where bandwidth is essentially limitless, the best way to get a message across is to simply repeat it endlessly:

Claude Shannon had assumed that the goal of the network engineer was to remove redundancy from communication without sacrificing its comprehensibility. The assumption reflected the capacity constraints that had always characterized communication networks. You had to prune what was communicated in order to squeeze everything through a narrow pipe. Remove the capacity constraints, and an inversion takes place. The goal is no longer to reduce redundancy but to increase it—to send the same signal over and over again in order to increase its influential power.

And hopefully by now we’re all familiar with the way that people existing in social groups are less open to new information than we’d like to believe:

[Researchers] found that exposure to an opposing view triggers a sort of immune response within a group. Members band together even more tightly to defend their position against what they perceive as an invasive idea. People “experience stepping outside their echo chamber as an attack upon their identity,” Bail reports. That makes “differences between ‘us’ and ‘them’ seem even bigger.”

I don’t remember feeling super enlightened by this book, but that’s b/c I spend a lot of time online where I naturally end up hearing about all the ways social media sucks. If you’re not that online and want a book about social media, this could be a solid option.

Here were a couple other random funny/interesting things I highlighted:

Fake photographs have been around as long as photographs have been around... In a photographic portrait made to memorialize [Abraham Lincoln] after his assassination, the artist Thomas Hicks transposed Lincoln’s head onto a more muscular man’s body to make the fallen president look heroic. (The body Hicks chose, perversely enough, was that of the proslavery zealot John C. Calhoun.)

And:

It would be hard to overstate how deeply people’s lives were changed by access to the mail. In the latter half of the nineteenth century, a time of growing migration and mobility, not to mention the dislocations of war, “epistolary intimacy” became “a basic mode of social and familial interaction,” writes historian David Henkin in his book The Postal Age. “Americans began to turn expectantly to the post, calibrating their perceptions of connectedness to the regular schedules of the U.S. mail.” Letter writers were not just able to converse with friends, lovers, and family members across great distances; they often found they could say things through the post that they found difficult or impossible to say in person.

On Muscle: The Stuff That Moves Us and Why It Matters, by Bonnie Tsui

A nice little book that’s partly about the science of muscles and partly about the author’s personal experience with strength and sport throughout her life. Personally I might’ve liked to learn a little more on the science side, but that’s a small quibble.

I liked this historical story about how the Greeks made art with muscles before they really knew what muscles were:

Like their Chinese counterparts, ancient Greek physicians in the time of Hippocrates, circa the fifth century BCE, did not generally distinguish muscles from flesh overall. But Greek artists of that period did sculpt their human figures with bulging ripples, even in places where no muscles existed. The appearance of an articulated body was deemed attractive, even before the Greeks understood what muscles actually were.

I liked this quote about the lack of historical “mythmaking” in women’s sports; it’ll be interesting to see how culture changes in the future with more time to hopefully develop those mythical figures:

The journalist Kate Fagan, who has spent her career studying the place of women’s sports in society, has suggested that “the history of men’s sports is uninterrupted mythmaking, the kind through which momentum is created.”...The lack of mythmaking for women athletes until recently has contributed to the fiction that women athletes are not worthy of the same attention...“What we don’t consider often enough is how stealthily men’s sports intertwines with history.”...Little by little, this is changing.

I think I was sort of already aware that we don’t (exactly?) generate more muscle cells when we get “bigger muscles”:

Skeletal muscle fibers are unique cells in the human body. They are long and skinny, run parallel to each other, and have multiple nuclei. We’re born with a set number of those fibers. They grow bigger in response to exercise, not by dividing, but by recruiting muscle satellite cells—stem cells specific to muscle that are dormant until activated in response to injury—to contribute their own nuclei to muscle growth and regeneration.

But even given that, muscles are quite adaptable in how they grow:

Though we’re all born with a particular number of muscle fibers...we are not completely predestined. A study of identical twins, one of whom ran marathons and did triathlons, and one who stopped due to an injury in high school, revealed that by the time the two brothers had reached their fifties, the endurance athlete was 94 percent slow-twitch muscle fiber, while the sedentary truck-driving twin was just 40 percent.

Look for us to one day use bear genes to prevent muscle loss:

Scientists are also studying the ways other species in the animal kingdom have built-in protections against muscle atrophy. I’m fascinated to learn that serum from the blood of hibernating black bears not only prevents human muscle cells from atrophy when added to those cells in lab culture, but also boosts mass in those same cells.

So, yeah: pretty good book. Her personal story had a certain beauty that I haven’t tried to capture here. I think the main thing keeping me from thinking it was a great book was there wasn’t that much I felt like I was learning for the first time.

The Stronger Sex: What Science Tells Us About the Power of the Female Body, by Starre Vartan

I made kind of a lot of highlights in this book and maybe I should’ve written an even longer review of it. My beloved ultimate frisbee even gets a small shoutout near the end of the book:

Additionally, a small but growing number of teams and sports allow men and women to play or compete together on the same teams, such as relay events in cycling, running, and swimming; racquet sports like badminton, tennis, and racquetball; as well as timber sports, korfball, and ultimate frisbee.

I thought the author generally did a good job of balancing biological fact with a desire to be politically liberal/”pro-woman”/”body positive”, though there were a few moments where I felt like she was putting a bit of a slant on it. Here’s the facts that seem incontrovertible:

Male bodies have a very real advantage when it comes to explosive strength due to glycolytic capacity,

This is why, for those quick, heavy lifts, male arm muscles can be up to twice as strong.

the finding was (and remains today) that “women on average have more fatigue-resistant muscles than men,” as Hunter puts it....although men outperform women by about 10 percent in short-duration activities, that gap drops to just 4 percent in long-distance events like cycling,

So men have more explosive potential on average, and generally speaking stronger upper bodies. Women have advantages in endurance. These differences don’t really appear until after puberty starts:

According to a foundational 1998 muscular strength test by researchers at the University of Puerto Rico, girls are as strong, fast, and able to play any sport as boys are up until about age thirteen or fourteen

Here’s how the author lays out her own main points in the first chapter:

As I detail in The Stronger Sex, women’s muscles retain strength over time better than men’s, women’s fat and metabolism offer a huge advantage for any pursuit that requires endurance, menstruation might be a pain, but it has lots of unique powers, sensitivity is a key to strength, and the very biology of women’s nervous systems provides significant advantages when it comes to pacing in physical challenges, durability, and pain tolerance.

That pain tolerance point was one where I felt like she was maybe showing her bias. It’s a popular myth that women have better pain tolerance, but, as I currently understand it, the science doesn’t agree with that. She herself says later in the book “The available data shows either no difference in pain tolerance between male and female athletes or a female advantage”, and the footnote accompanying that sentence leads to a paper that says “The data...revealed no statistically significant difference in pain tolerance when NCAA Division IA male and female athletes are compared” (Also she cites two papers, one from 1981 and one from 2000 (which she erroneously cites as being from 2022)...why not cite up-to-date research?)

I liked this quote (see also the discussion of mythmaking in the On Muscle review above) about how women have been underlooked due to modern researcher’s gender bias:

This modern research [that ancient women hunted] supports the fossil evidence. In the past, skeletons found with hunting equipment were typically assumed to be those of male bodies. Even when remains of women were unearthed alongside spears or chipped stone points, they were often considered, as one researcher put it in 1971, “difficult to explain.”

I generally agree with her theory that one reason women’s sports are underdeveloped is the way we treat girls from the time they’re quite young:

Practice makes perfect, and even young boys are taught the physical skills needed to succeed in sports more often and at younger ages than girls...

Girls aren’t pushed to do physical labor; in fact, it is actively taken away from them and given to boys and men. This has lifelong repercussions. “Tonnage, or training volume, as measured in sports science, is how much weight you have lifted over the course of your training history. It’s a crude measure, but it’s a pretty good indicator—if you’ve lifted tens of thousands of kilograms, you’re probably quite strong,” said Nimphius.

“If we go back and look at the difference in tonnage for young women and men over a longer time, you’re going to see some big differences. If every time you go to the grocery store, your dad was like, ‘Sophia, bring the groceries inside,’ that move is the same as a farmer’s carry in the gym, right? So then you go to college, and now you’re in a weight room. The tonnage you can pick up will be higher because of these small, seemingly innocuous jobs.” Women start weight training later than men, and girls start most sports later than boys.

Here’s another quote that makes a similar point, but this time in the context of injuries:

Dance shares similar high-risk movements and knee-joint strain as Aussie Rules Football, but sex disparities in ACL injury rates aren’t seen in dance. This is likely because both sexes begin dancing at very young ages (well before puberty) and receive similar training—a fact that points away from physical differences in male and female bodies and toward training and experience as the risk factor for injury. [I agree with the general point, though to be honest, I’m not sure I believe that dance is actually as stressful on the body as sports? -LT]

The problem is...we’re just not working out the way our ancestors did, basically no matter how hard we try:

In a 2017 study at the University of Cambridge, researchers performed CT scans of arm and leg bones from skeletons of central European women dated 7,300 to 1,150 years ago (early Neolithic era and into the medieval period). They compared them to the bones of modern-day women, including both sedentary women and athletes (for example, soccer players and women on Cambridge’s championship rowing team). The researchers found that the ancient women’s arms were up to 16 percent stronger for their size than those of the rowers (who trained twice a day and rowed seventy-five miles a week), and 30 percent stronger than those of nonathletes. Interestingly, their legs were about the same strength.

Even the long-consistent research finding that men have better mental spatial reckoning may come down to amount of practice and cultural roles:

One oft-discussed area is the difference in spatial ability between men and women, which has been found over and over again through decades of studies...Most researchers have attributed greater male abilities in this area to testosterone’s influence on the brain...One small Italian study showed, in a group of seventy-one women and eighty-one men, that “women improved performance and reached men’s scores in the MRT [mental rotation test] when they were led to believe they were better than men.”

The book isn’t only about sports; she actually spends quite a bit of time talking about menstrual cycles, in ways that only sometimes relate back to sports. One of the points she makes is that we simply need more research:

For example, the most current and commonly used research on first periods is a 2008 study of just ten girls’ hormone levels.

I’m not sure I totally understood all the details here but she also says that modern science has learned the role of the egg in reproduction is not as “passive” as is commonly understood:

in text after text, the language used for eggs [wrongly] described them as queenly, princess-like, and regal, wishing and hoping for a sperm...

The newest evidence indicates that not only does an egg decide which sperm to admit; the egg may attract or repel different sperm even before they reach the egg. In 2020, scientists at Stockholm University, in collaboration with colleagues at the University of Manchester, discovered that eggs release a chemical that can attract sperm as they travel. They also found that different eggs attract different varieties of sperm.

Here’s one final highlight I found interesting:

In this study of 105 people, those who exercised during pregnancy had shorter labors (by three hours) and were more likely to have a vaginal birth...This makes sense. It’s not as if our ancestors were sitting around all day while pregnant.

Cloud Warriors: Deadly Storms, Climate Chaos—and the Pioneers Creating a Revolution in Weather Forecasting, by Thomas Weber

I was personally kinda disappointed by this book, but for very specific personal reasons. I’m an aerodynamicist, and aerodynamics is about as close as you can get to weather forecasting without actually studying weather forecasting. The same laws of physics determine both how weather patterns move through the atmosphere and how air flows around a car. I actually took a class in college where I wrote a piece of weather-prediction software.

So, I was hoping for some certain level of technical detail about why we’re better at predicting weather than we used to be, and for the most part I was left wanting. The first highlight I made in the book was this sentence: “Today’s five-day weather forecast demonstrates accuracy at least as good as a one-day forecast from 1980, researchers have concluded. “ But for the most part, he doesn’t really explain why that is.

One of the “revolutions” he covers is the social science aspect of weather, i.e. finding ways to better communicate weather alerts/information:

Klockow McClain is part of a new vanguard in weather forecasting: the social scientists. It’s one of the most consequential developments in the battle to keep people safe from weather.

While I’m sure it’s important, this wasn’t super interesting to me.

Another “revolution” was that of hyperlocal weather forecasts. But that’s basically how I’ve consumed weather information for as long as I’ve been an adult. My current favorite site is Windy but I was previously a big fan of Dark Sky until they were sold to Apple and took their website offline. My weather tip: don’t just look at the daily prediction! Look at the hour-by-hour prediction, and even better, look at the actual radar and you can get a way better feel for whether a particular storm is actually going to hit you or miss 15 miles to the north.

That’s not the only thing the author meant by “hyperlocal” weather, but for a big part of that section it did feel like they were describing exactly how I already use weather apps and calling it a “revolution”:

“Let’s say a competitor predicts a high of thirty-five, mostly cloudy, for tomorrow. And AccuWeather predicts thirty-four in the morning, with temperatures falling sharply through the twenties into the teens in the afternoon, with all water and slush freezing up.” If it turns out the high was 35 degrees and not 34, Myers says, technically the competitor won. But, he asks, which forecast was more useful? “We told the story.”

But anyway, at least we got some good weather facts:

As much as wind-speed numbers dominate news coverage, talk to any hurricane expert and they’ll tell you: It’s the water, not the wind, that kills. That isn’t to minimize the threat from those tornado-strength gales, but any given hurricane can easily dump ten or more inches of rain. If that much precipitation fell as snow, it would pile up ten feet or higher. Big storms get much wetter. Hurricane Harvey in 2017 deluged some locations in Texas with over sixty inches of rain...

Sleep Groove, by Olivia Walch

Every year or so there’s a study that’s like, “We made light brighter in this school and/or work environment, and everyone was happier and smarter and more productive.” And then nobody does anything about it.

I’m not great at sleeping. I tend to read a new book about sleep every once in a while (see here for example), and Sleep Groove was this year’s edition, after I saw in on a Best Of 2025 list when looking for things to read.

In short, she basically makes two points:

1. “Sleep rhythmicity needs to be prioritized the same way we prioritize sleep duration” (i.e., this is the “sleep groove” — sleeping/waking at a consistent time)

2. Helping your body stick to that groove is basically all about the light you’re getting. (You want MORE during the day, especially in earlier hours, and less at night)

One thing I was somewhat surprised by was learning that, actually, sleep deprivation isn’t as crazy bad for you as is sometimes claimed.

Does extreme sleep deprivation damage something permanently inside us? Does it shatter our minds for good?

Well, maybe, but the evidence isn’t particularly overwhelming. If anything, efforts to track down the answers to this question have mostly led to answers like, “Eh, probably not?” [Read the book to get the full story of the few random people who’ve stayed up for like a week straight]

She also points out the “contradiction” in how, although light late at night is her main culprit for bad sleep, actually, pre-industrial societies don’t sleep that much more than we do:

In one study looking at sleep in pre-industrial societies, the authors found subjects sleeping an average of 5.7 to 7.1 hours per night, on the lower side of what most modern societies report.

She says that caring about “blue light” has real scientific basis...but that sometimes what is meant (scientifically) by blue light isn’t exactly the same as the layman’s conception (of “light that is bluish”):

And now we get to the punchline: ipRGCs—your subconscious visual system—love 480 nm light. That’s a tiny slice of the mountain range [read: the light spectrum]. If you made an LED light source that was just a little spike around 480 nm by itself, it would look blue-green to your eyes.

So you could have a spectrum that looks “blue” to your eye with a ton of 480 nm light, which means it would make your ipRGCs go wild. They’d fall over themselves for a blue containing lots of 480 nm. Yet a different blue, which could look identical to your eyes, might have essentially no 480 nm light...

...That said, if you see a headline that says, “Blue Light DEBUNKED,” it almost certainly means “the experience of consciously perceiving a color as ‘blue’ is immaterial to the signal received by your circadian clock,” and not that 480 nm light doesn’t matter.

Anyway, most of the rest of the book is about light and how we get too much of it at night. And some people are more affected by it than others. In one of those “genes interacting with environment” situations, we didn’t notice this sensitivity when we were all living out in nature, but now that we have artificial lighting it turns out some people are especially effected by it:

Some people only needed 10 lux to have their melatonin suppressed by 50%. That’s not very bright at all (think: your phone)

There ended up being a fifty-fold spread in the lux it took for people to have their melatonin suppressed by half. Said another way, two people could live in the same house, with the same lighting in their bedroom, and one of them could be fifty times more sensitive to that light than the other.

Also, if you didn’t know this already, the sun is like WAY brighter than indoor lights. Here’s another little screenshot of the ebook with a couple helpful diagrams:

Anyway, I liked this book, I thought the breezy, Millenial-style tone actually was done well and the handmade diagrams were a solid addition (love an author who understands the power of good diagrams). I was generally already pretty aware of this stuff around light so I’m not sure how much the book has changed my habits but it at least made me a bit more cognizant of the light I’m getting at night.

Superagency: What Could Possibly Go Right with our AI Future, by Reid Hoffman and Greg Beato

Reid Hoffman is the founder of LinkedIn and one of the big early donors to OpenAI. He and his writer wrote this book about how awesome AI is going to be. I found it on a list of best science books of 2025 but I was pretty disappointed.

I have nowhere near the negative feelings towards AI that I’ve seen elsewhere on the Internet. I could legitimately see AI doctors and therapists being a good thing (AI therapists are used as an example in the book)! Yes, there are risks, but there are also downsides to our current system where it can take weeks to find a therapist and then they still might be someone whose personality doesn’t mesh with you at all. It took 6 weeks for my doctor to have 10 minutes to discuss my MRI results with me! I absolutely believe there could be use for AIs who are slightly less expert but vastly more available.

But the book mostly annoyed my with its overly rose-colored view of AI. Looking back at my highlights (I’m writing this a little too long after reading...) I maybe don’t have enough quotes easily accessible to prove that feeling, so take it as more opinion than fact. I highlighted, for example, this comment about how the government might benefit from using AI:

So imagine a world where IRS Prime is a thing, and you get your tax refunds in two days or less...

But...ummm...doesn’t everyone already know that taxes are only annoying because tax prep companies lobby the government to keep it that way?? (See here, for example). We don’t need AI to fix that!!! Unless you’re inventing an AI that’s got a way to dis-incorporate TurboTax? Or take this quote on adopting AI in medicine:

A nation that lags in adopting AI-driven drug discovery and personalized medicine techniques may soon find itself facing a significant gap in health care outcomes...

But...ummm...isn’t the US already facing a significant gap in health care outcomes (relative to its income, etc)? Is the AI bringing us Medicare for All?

I realize these are maybe not the best, most-detailed examples, nor the most central to the actual meat of the book, but I think they do a solid job of summarizing the pro-free-market, pro-AI vibe of the book. Another feeling I’m left with after reading that I didn’t take good enough notes to strongly back up with quotes: many of the pro-AI arguments were made via analogy to other technologies that have “gone right” in the past (GPS, Big Data, development of the car industry). They were fine analogies as far as they go, but there are real reasons for us to be worried AI is fundamentally different.

So, yeah, while the book is probably technically right about many things, I can’t help thinking there are better, more thoughtful, less biased-feeling books about AI out there. Though I suppose I got exactly the bias I was promised with a subtitle of “What Could Possibly Go Right with our AI Future”...

What’s Our Problem?: a self-help book for societies, by Tim Urban

I had long ago been a reader of Tim Urban’s blog Wait But Why. He mostly stopped posting and then I mostly forgot about his site, until I remembered he had been working on a book that I never read. What’s Our Problem? is that book.

I was...pretty disappointed by the book. The author’s answer to the book’s title, is, in short, “cancel culture”. I’m sure the author would say I’ve simplified it too far, and that’s true, but when it comes down to it, complaining about the American Left is easily the biggest part of the book. Going by the numbers in my ebook, The first ~3 chapters (28% of the book) introduce his general ideas, then there’s ~1 chapter (6% of the book) applying those ideas to the American Right, followed by ~3 chapters —clocking in at 50%(!) of the book—criticizing the American Left. (For corroboration, see some of the comments in this Reddit thread, like this one: “My biggest criticism is that he spends like 5 times as long criticizing certain left-wing habits as he does criticizing the right-wing side. I think the arguments he makes are solid, but such an imbalance could give the appearance of much political bias to a book which argues against being biased.”)

And I would actually label myself a reader who’s especially open to the points he’s making. I’ve read the books by Helen Pluckrose and Jonathan Haidt that he references extensively. But from the perspective of 2026, to me at least, his book feels like it aged very badly and very fast. The Left, as far as I can tell, has mostly gotten over the craziest moments of “social justice fundamentalism”/”cancel culture”, while the right is still being just as anti-science, anti-democratic, etc as they were 5-10 years ago.

His big idea in the book is that we should view politics not just as a 1-D, Left to Right spectrum, but as a 2-D spectrum, which also contains a Ladder that runs from Low to High. Low is basically short for low-minded, a “politics as a sport where I root for my team no matter what” approach, where it’s all about winning arguments (whether you’re right or not) and winning power, facts and fair play be damned. And High is basically short for high-minded, taking a collaborative, scientific, thoughtful approach to politics where you’re open to changing your mind when you hear new information and you care more about the success of our country than about your side always being right.

Urban argues that Social Justice culture is evidence that the Left has followed the Right down into “Low”, win-at-all-costs, punish-the-nonbelievers politics. And I don’t think he’s totally wrong, but again, it really feels to me reading in 2026 that the left has (for the most part) gotten over the strongest parts of the social justice craze that occurred from 2017-2023 or so.

And the irony is, to me that feels like exactly the kind of process he was writing about in his definition of the High rungs of the Ladder—some new ideas started gaining popularity in the mainstream, they became a craze, but then eventually critics pointed out some good objections, and people were willing to admit they were making good points, and eventually we settled down to something of an equilibrium. Social Justice culture makes some very good points, and the mainstream Left has now adopted some of those concepts while throwing out some of the wildest ideas from the heights of the SJ craze.

That happened over the course of a few years, which seems natural since the brain of a society changes its mind slower than the brain of an individual...but in the meantime Tim Urban got so worried that he wrote a whole book about how extreme social justice thinking was going to conquer society (Chapter 7 is titled “How to Conquer a Society”). The Left underwent exactly that “High” rung process of being open but also critical to new ideas, absorbing its insightful lessons while (eventually) rejecting its excesses...they just didn’t do it fast enough for Tim’s liking.

But is he even free of bias himself? In that “Conquer a Society” chapter, he uncritically shares the story of Yelp adding a “Business Accused of Racist Behavior” warning to their site:

In a late 2020 blog post, Yelp—the restaurant reviews site critical to any modern restaurant’s business—announced that it flagged hundreds of restaurants for “racist behavior.” In future cases, a large “Business Accused of Racist Behavior” warning would appear front and center on the restaurant’s Yelp page. In a statement to the site Eater, a Yelp spokesperson clarified their criteria: “We’re taking a firmer stance against racism with the Business Accused of Racist Behavior Alert that will be placed when we learn of reports of egregious, racially-charged actions such as the use of racist language, symbols or sentiment that clearly discredits the Black Lives Matter movement.”

A super quick internet search tells me that it was “only ever used two times”. And that linked article was published in early 2022, while Urban’s book was published in 2023. When literally the first thing I’m interested in fact-checking is proven to be totally over-blown with one minute’s work...well I have a bit less trust than I used to that he’s actually unbiased. In a book that’s all about encouraging us not to be biased for “our team”, I feel like you just gotta do better.

He says near the end of the book:

U.S. history is a roller-coaster of oscillating power between left and right. It would not surprise me if a decade from now, the lion’s share of cancel culture stories were the work of the political Right and it was the Left electing the demagogue president.

I’ll grant him that the Right seems to have embraced various versions of cancel culture (people getting fired for things they said about Charlie Kirk, the Trump Administration denying funds to states that have policies they don’t like, etc). We’re yet to see (and admittedly we’ve got some time to go before his “decade from now”) if the Left will also start electing Trump-level demagogues. If we do, I’ll happily admit I was wrong in my criticisms of his book...but I have my doubts.

Proof: The Uncertain Science of Certainty, by Adam Kucharski

I think I was expecting more of a book about the science of mathematical proof, whereas Proof actually ended up being more about “real-world” proof (Which maybe should’ve been expected based on the subtitle! But I think when I first saw the book on Amazon, it had an alternate subtitle: “The Art and Science of Certainty”!). But I still thought it was a pretty solid book.

He spends a solid amount of time talking about Abraham Lincoln and his devotion to being able to “demonstrate” proof in his arguments, which...really just makes one nostalgic for the days where a president might actually care about something like that. Ugh:

Digging through dictionaries and reference books, Lincoln couldn’t find a description [of what “demonstrate” meant] that really made sense. ‘I consulted Webster’s Dictionary. That told of “certain proof”, “proof beyond the possibility of doubt”; but I could form no idea what sort of proof that was...At last I said, “Lincoln, you can never make a lawyer if you do not understand what ‘demonstrate’ means” and I left my situation in Springfield, went home to my father’s house, and stayed there till I could give any proposition in the six books of Euclid at sight.’

Did you know Lady Justice didn’t originally wear a blindfold?:

the history of Lady Justice mirrors the complexity of the idea she represents. Take her blindfold, which first appeared in a fifteenth-century cartoon. It was designed to be satirical, with a jester wrapping a cloth over her eyes to hinder her judgement. The accompanying text warned of ‘folly, blindness, error, and stupidity’. Only later was the blindfold reinvented as a symbol of impartiality.

He talks about the concept of “proof” in our trial systems (how many guilty people should we let go free to ensure we don’t send an innocent person to prison?):

Condorcet also calculated the size of jury required to ensure that the chance of a false conviction fell within his acceptable risk. If each juror had a 90 per cent probability of making the correct decision, he concluded that a jury should have at least thirty members, with twenty-three ‘guilty’ votes as the threshold for a conviction.

Then we learn about the historical development of the famous “5% p-value” from statistics, which was developed somewhat arbitrarily, at least relative to how much esteem it was held in for quite a while:

Fisher seemed happy to blur his discussion of p-values and the 5 per cent significance level. In his 1925 book, he initially suggested it was ‘convenient to take this point as a limit’, but in the next sentence he stated that findings meeting this requirement should be ‘thus formally regarded as significant’.

I really enjoyed the description of this medical trial, what a great way to do science while also valuing the lives of patients:

To reduce this dilemma [i.e. the ethical dilemma of denying a potentially lifesaving cure to the control group in a study], they settled on an adaptive approach known as a ‘play the winner’ trial. Suppose we take two balls – one labelled ‘control’ and one ‘ECMO’ – and put them in a bag. One ball is picked randomly and the first patient is put in that group. In the real trial, this happened to be an ‘ECMO’ ball, which was returned to the bag after it was picked. If the patient survives (as they did in reality), an additional matching ball (i.e. ECMO in this case) is added to the bag. Otherwise, the other ball type is added. This way, patients are still allocated randomly, but the allocation adapts depending on how effective the treatment is.

And of course, given that it’s a science-y book published in the last few years, there’s a chapter near the end about AI. I didn’t really highlight anything here, but at least I didn’t have any obvious complaints about it (see Superagency, above). Nor did I highlight much in the section on Bayes theorem—it was well-done, but none of it was really new to me. I did highlight this sentence though, just because I liked the way it put things: “In other words, Bayes forces us to think about the underlying processes influencing the event we’re interested in. This can be useful, because it helps us explore how various uncertainties propagate in each situation. And it is often in uncertain situations where science can struggle most.”

FICTION

The Blade Itself, by Joe Abercrombie

This is the first book in the series The First Law, which is often considered one of the best fantasy series released in the last 30 years (source: IDK, some Reddit threads).

I thought The Blade Itself was...fine. Pretty good. I finished it a year and a half ago now, and haven’t bothered to start the rest of the series yet. Admittedly I have seen a Reddit comment or two saying the first book is the worst in the series and it’s after the first book when it really becomes great. (Also admittedly: I wasn’t that impressed by Game of Thrones the first time I read it, and didn’t read the next book in A Song of Ice and Fire until a year-plus later, either. And then I got addicted and it became one of my favorite series.) So I very well might still read more of it someday!

Master of the Revels: a return to Neal Stephenson’s D.O.D.O., by Nicole Galland

A sequel to The Rise and Fall of D.O.D.O., except this one doesn’t have Neal Stephenson as a co-author. Overall...it was enjoyable enough. I was never anywhere near feeling like I wouldn’t finish it. But I wouldn’t say I loved it, either. At one point, I was about 80% of the way through, and I realized, hmm, we’re almost done. The plot isn’t going to have time to go anywhere new. We’re just going to get a climax to the current plot points. So I was a little underwhelmed—my intuition was that more was going to happen. A perfectly OK book if you’ve read the first one.

The Running Grave, by J.K. Rowling

I was looking for a page-turner, and this one worked well enough. It definitely wasn’t the most comfortable book to read—the tension of knowing something bad might soon happen to one of the main characters made me want to stop reading the book almost as much as it made me want to keep reading. But I did stick with it, and if you’re looking for a long detective story, this is a solid option. Maybe I should read more detective novels, I do like solving puzzles...

Wind and Truth, by Brandon Sanderson

The fifth book in a five-book narrative arc of a (planned-to-be) ten-book fantasy series...whatever that means. I really liked the first book or two in this series. The rest of them have felt like a bit of a slog (this one was 1700 pages, according to my Kindle). Calling it a “slog” is perhaps too harsh—it’s a relatively easy read. But it’s long and it sometimes feels like not that much happens for how long it is.

The series is also so long that I’ve forgotten a fair amount of the characters’ backstories. I didn’t have much trouble remembering what I needed to about the main characters, but there were definitely a few side characters (and references to old plot points) that had me saying, wait...what? It’s a little too much when, for example, every single book in the Harry Potter series reminds me in Chapter 1 that Harry is a wizard who goes to Hogwarts...but this time I felt like Wind and Truth could’ve used a bit of that energy given how long it was since I’d read some of the earlier books.

This book, being the climax of the series, had more excitement than the last two. Even though I’m not in love with the series, I do feel invested in the characters at this point, and even though the book’s a little corny, it still got me to tear up at a few climatic moments. Who would’ve guessed that the moral development of a sentient sword could bring a tear to the eye?

Polostan, by Neal Stephenson

I’ve read basically every book by Neal Stephenson after first discovering him when I was in college and reading some of his most well-known books at the time like Snow Crash, Cryptonomicon, and Anathem. Polostan felt like a departure from his last few books (i.e. “Reamde”, “Fall; or, Dodge in Hell” and “Termination Shock”). It’s maybe the book that feels the least Neal Stephenson-y, and basically just feels like a straightforward World War II novel. The book is very obviously a set-up for future books in the series, so I don’t really want to judge it plot-wise without knowing where the future books go, but I made it through Polostan pretty smoothly and liked it enough that I’ll probably see where the next release takes things.

The Green Bone Saga (trilogy), by Fonda Lee

Picked it up because I saw a couple separate comments on Reddit (over the course of a few months) saying that it was one of the best fantasy series of the last decade or so. Liked it enough to make it through all three books pretty easily but didn’t, like, love it. (I’m writing this a while after finishing so not sure I remember everything that felt imperfect about the book. I do remember I was surprised there wasn’t more of a series-wide climax to the third book...things ended a lot more...peacefully(?) than I’d expected. But books should do that sometimes!)

I believe this is actually the first book I’ve ever read in the genre of “urban fantasy” (i.e., it has magical elements but in a world with guns, cars, TVs and skyscrapers). Having lived in Asia it was fun to see the Asian elements and try to figure out which real-world cultures were represented by various in-book countries.

Blood Over Bright Haven, by M.L. Wang

Yet another fantasy book that was fine. Maybe I need to find something new that interests me instead of continuing to be slightly underwhelmed by fantasy novels. The main thing I remember about this book is that I figured out the “twist” basically the first moment that it was hinted at.

I do remember thinking that the book’s magic system, which is basically magic brought into the world via computer programming, was fun and unique. And I enjoyed the way the two main characters from very different backgrounds came together and came to understand each other.

Too Like the Lightning, Seven Surrenders, The Will to Battle, Perhaps the Stars

aka Terra Ignota series, by Ada Palmer

I read the first two books in Ada Palmer’s Terra Ignota series back in 2018, shortly after they came out. I wasn’t super into them at the time, even though a blogger I really liked had written a very positive review of them (well, sorta...it’s here). But I decided to start the series again, partly b/c I wanted to give them another try and partly b/c I could finish the series now that all the books are out. (I felt like I remembered most of book one, but there was a lot in the second book that I had basically no memory of having read before.)

The series definitely felt interesting and innovative to me in a way that some of the books above didn’t. They’ve got a unique blend of semi-near-future setting and a plot with characters who take inspiration from our past (mainly the Renaissance period—the author is a historian/professor who has also published academic nonfiction about the Renaissance. The last book also has some allusions to The Odyssey which was funny b/c I was reading it the same week I watched the new Odyssey movie).

Recommended if a combination of sci-fi, politics, philosophy, and plot points hinging on human potential, gender politics, and restorative justice (kinda) sounds interesting to you. The last book was long but in the end it was worth it for me. As good novels sometimes do, I feel like it re-inspired me to try to live my best life.

All Person arose from an attempt to do a few things differently from how your standard force forehand/force backhand defense works. The main impetuses behind AP are:

1. How can we design a defense the gets the most out of players who might not have super elite athletic abilities? (TK interview, 16:15)

2. How can we design a defense that makes hucking harder? A traditional force forehand leaves open an open side throwing lane that’s:

   1. straight down the field—so the disc gets vertically down the field quickly

   2. unobstructed by downfield defenders2—so there’s no need to put extra height/angle on your throws to get them over downfield defenders.

   As a result, open side hucks (it was theorized) are completed at a high rate (Shayla Harris interview, 1:06:00).
   
   Recent content, like this Hive Ultimate video by Noah and this Ultiworld video by Joe Marmerstein, have given us some evidence that this theory is actually true, and same third, down-the-sideline hucks are completed at a higher rate than other hucks, with Joe, for example, finding that sideline same-third hucks were completed at a 64% rate vs. 55% for non-same-third hucks3.

   

   
   Aside from the presence of traffic, there’s the simple geometrical fact that a throw straight down the field can gain (vertical) yardage more quickly:

   

   
   

3. How can we design a defense that gets more out of mark defenders in general? If a thrower facing a Traditional Force Forehand defense wants to throw a pass down the open side, the mark is just...not really doing anything. Is that really the best we can do? Can we find a way for the mark to “be in the way” more often without also giving up lots of throws to the breakside?

Bravo coach Mikey Lun has described AP defense as having three principles4:

1. Make things choppy—disrupt the natural flow of the offense (with the positioning of the mark defender being a main way this is achieved)

2. Disconnect the downfield from the backfield — partly done by the active mark/“force to help” (see below), partly done by downfield defenders denying the under throws

3. Increase transit time — the longer the disc is in the air, the more things can go wrong (more time for D to catch up to floating throws, more time to reposition). This applies to hucks, but based on their explanation it also applies to forcing the offense to throw more passes swinging horizontally across the field in general

How it works

Marking Principles

A lot of the uniqueness of AP defense comes from the way the marks position themselves, and from the handler defense in general. Mikey Lun described their marking principles as follows:

1. Force it back — force the disc back in the direction it came from

2. Force to help — Force towards a help defender in the lane (this basically always means towards the other handler defender who’s sagging off the reset handler). This is where part of AP’s ability to “disconnect” comes from: if you’re always forcing to the side where the other handler is, on average there’s going to be more traffic in the way of downfield throws to cutters.

3. Force it wide — aka “force middle”. Force towards the wide side of the field.

The tricky thing here5 is that, even though there are three items on that list, there are really only like 2 or 2.5 independent principles here in practice (sometimes just one):

• If the offense just threw a reset swing pass off the sideline, then “force it back” and “force to help” are both the same direction, so you might see the mark start out forcing sideline for a few beats and then switch to forcing middle, especially once the thrower of the reset swings off the sideline.

  • But if the thrower of the reset never leaves the sideline, the mark might just continue to force sideline, as help and back both continue pointing in that direction.

• If the offense just threw a swing pass the brings the disc closer to the sideline, then “force it back” “force to help” and “force middle” are basically all pointing the marker towards forcing middle.

Here’s a clip that showcases how those examples work in practice (source):

First we saw a pass that took the disc closer to the sideline, and the mark is a mostly consistent force middle:

On the next touch, the marker starts forcing sideline—“back”+“help”. But then as the “help” moves across the field to the wide side, the marker makes a well-timed switch to force “help+wide”. It doesn’t really work out this time6 because Doublewide is looking for handler play anyway, but the timing is impressive:

To me this is a good example of how to use dynamic marking to gunk things up for the offense. Done right, it starts to feel like there is no “open” side - there’s just the breakside on one side, and the side with traffic on the other.

Here’s another example where Cody Spicer forces towards the sideline (“back + help”) when the help/traffic is there, then switches his force to “help + wide” once the other handler clears out of the sideline lane:

Off-ball handler defense principles

The other big part of AP is what the handler defenders are doing when they’re not marking. Here, a few main principles show up on film:

• If the disc is near the middle of the field, look to poach into the downfield throwing lane and encourage a swing pass towards the sideline.

  

  

• Once the disc is near the sideline, reset defenders play tight to try to shut down all of the offense’s options.

  

  This sequence against Revolver in 2024 shows the handlers starting poachy with the disc in the center then looking to play lockdown, nearly faceguarding defense when the disc nears the sideline:
  

• In general, look for opportunities to switch on backfield defense to shut down handler-to-handler play. Mainly that’ll be on “upline”-type cuts and on cuts behind the disc towards the breakside7.

Here’s an example against Doublewide where the switching and middle-of-field poaching leads to Doublewide passing it back and forth between handlers ineffectively:

Here’s a switch on a “pepe” cut to the breakside behind the thrower (source):

The tendency to both (a) play tight handler defense near the sideline and (b) switch on pepe cuts8 can lead to some awkward situations where a defender starts by playing incredibly tight defense, but then tries to drop off that same person once they catch the short sideline reset:

Defenders will often look to shut down the next pass before they fully commit to marking, whether that’s switching on handler movement or running way downfield to stop a continuation:

I think this is generally a strength of AP, as I’ve written before that “I’m not sure frisbee theory has completely figured out all the ways to take advantage of the thrower’s inability to move”. The marker looking to mark last, only after making sure they’ve helped shut down easy continuations, definitely feels like an intelligent principle.

Downfield

In the downfield space, defenders are mainly playing straight person-to-person defense, perhaps working harder to shut down unders than they are to shut down deep cuts.

While an earlier Reddit thread mentions orbiting as an important mechanic, it doesn’t stand out that much on film. I think there’s two reasons for this:

• Forcing middle(ish) means that the downfield defenders have less orbiting to do, as they can always stay on the “middle side” of their assignment if their assignment is, say, running from one sideline to another

• Because they’re generally playing tight defense, orbiting at most ends up meaning moving a couple feet from an offensive player’s one shoulder to their other shoulder:

Maybe this just feels less important to me than it really is, because it feels like such natural instinct to do it anyway. But in general, yes, the downfield defense will need to adjust slightly as the disc swings.

The sideline conundrum

There’s one big thing about Force Middle defense that never really made sense to me. In basically every other sport, it’s well-known defensive strategy that you want to force the offense towards the sideline. The sideline acts as a “second defender”, cutting off options from the offense.

Frisbee is the same—the models built by Shown Space based off of UFA data show that frisbee teams are also more likely to score when they’ve got the disc in the middle of the field, compared to having the disc the same distance from the endzone but nearer to the sideline. So (I’ve always thought9), why would it make sense to play Force Middle defense, when you’re just endlessly giving the offense the larger half of the field to work with?

Mikey Lun makes some comments in the Shayla Harris episode that are critical to understanding AP defense, when he says:

even though it looks like we’re playing [Force Middle], we really like the disc to be on the sideline.

To me, this is a key to understanding AP defense. It’s perhaps best viewed through the lens of a “trigger-based” defense, as seen through the lens of Ian French’s article: Pressing, Traps and Triggers.

When the disc is in the middle of the field against an AP defense, the handler defense is poachy, encouraging swinging the disc towards the sideline for minimal gain.

But when the disc is near the sideline, Bravo typically switches to tight handler defense10. So, they’re forcing middle when the disc is on the sideline, but they’re defending especially hard in those moments to give you no one to throw to. (And of course, when the disc is near the sideline, a “force middle” becomes more like a “force flat”, since you don’t need to shade to the side as much to prevent downfield break throws, as the sideline is doing that for you already).

When it’s done right, it can look like the offense both struggling to make forward progress and struggling to just get the disc off the sideline.

Portland Rhino Slam player and longtime frisbee analyst Joe Marmerstein makes a similar observation in an old Reddit thread11:

First of all, the idea that FM wants the disc in the middle of the field is just wrong. In my opinion the basic strength of FM is in pushing the disc to the sidelines, and then having the “flat side” scenario where you have a flat mark making upfield throws difficult and can play very aggressive on the reset.

Thanks also to Joe for suggesting this clip of Rhino struggling in a similar way to the Machine clip immediately above:

I do think it’s interesting how Kins mentions that when the disc is on the sideline, the critical area for the downfield defense is kinda the same regardless of whether the mark is forcing sideline or wide side—10 yards downfield of the disc and a bit off the sideline. But if that’s the case, I’m almost back to thinking they should just force sideline on both sides of the field since then at least the marker is taking away as much of the field as possible.

Dissolving the upline space

[Added 2026-07-25]

A couple commenters below mentioned that they liked how AP defense denied the offense traditional upline cuts. This wasn’t included in the article originally; I suppose it didn’t stand out to me at first because it’s harder to notice the *lack* of something on film.

AP trades the upline cuts for the straight-up-the-field, “inside backhand” lead pass:

Re-watching Brown vs. Colorado in 2024 college semis, there are surely a few points where a player’s instinctive upline cut came to nothing:

But in return Brown often got offense out of the inside upfield look:

This is not an all-or-nothing situation—there are times when Brown completed a traditional strike cut even with the mark trying to take it away, and there are times when Brown turned the disc over attempting the inside upfield throw.

It’s hard to say whether this is an enduring advantage of AP defense, or if it’s just a function of many players not being fully accustomed to attacking the spaces AP makes available.

In favor of “it’s objectively a harder throw/easier to play defense” I’d probably say:

1. This inside backhand look usually requires curving the disc between two receivers; while the traditional upline has both defenders on the same trailing side when it’s open.

2. Arguably the total space in which the handler defender needs to stay connected to their person is smaller—the upline strike cut space in a Force Forehand can extend far from the reset space, while in AP defense, the upfield handler cut in an AP defense is still on the “same” side of the mark

   

   :

   As a result, it’s a bit easier in AP to play hard, tight handler defense without getting beat for huge upfield gains.

I’m personally agnostic as to whether it’s easier to defend this situation in AP. I think teams could get some of the same “keep the disc on the sideline” effect out of a Force Forehand defense if they taught their players to position better & sag a little more to deny uplines. And if more teams were playing AP, more players would get better at making (and throwing for) these cuts (more on this “off-meta advantage” below).

But without a doubt, at many levels of our current frisbee culture, having this ability to easily deny upline cuts is a useful tool.

Analysis

Now that’s we’ve looked at how AP works, here are some varied big-picture thoughts on how AP stacks up to other defenses:

The off-meta advantage

In the Shayla Harris episode, they talk about how teams want to use continuation passes—catch the pass and keep it flowing in the same direction, riding the wave of momentum.

I think this is partly human instinct—think about the field full of five-year-old soccer players where every single one of them is following the ball wherever it goes. But it’s also partly just the dominant strategy in frisbee culture—e.g., everyone learns to look for the continuation pass to the breakside before the mark seals the space.

AP defense does better than most at shutting those looks down. It’s hard for offenses to take advantage of the spaces that are left open, because it goes against the instincts they’ve developed in years of playing frisbee.

When the TK Pod Practice episode was posted to Reddit a while back, I made a comment to this effect:

Watch this sequence from the 2025 regional final:

The disc swings in the backfield from left to right, and the “ping pong” back towards the left is open immediately, but the immediate throw is looked off and the disc stops for a few seconds. The disc ends up going there anyway, but it happens ~5 seconds later and the offense has no flow.

You’d think a team in Bravo’s region would be ready for AP defense and thus ready to look for that ping-pong flow, but we don’t see that consistently. As a result, it’s a bit hard to separate “AP defense is good, period” from “AP defense is good b/c teams haven’t optimized around beating it”.

The downside of huck turnovers

In his appearance on Pod Practice, TK and the guys discuss giving up deep shots. Because the downfield defenders focus on stopping under throws, this is an area where, if anything, the defense is encouraging the offense to throw the deep shots, and expecting the defense will be able to influence enough of them to come out ahead in the end.

But aside from the problem of what happens when the offense completes those deep shots, I see another potential issue: turnovers on hucks are less valuable (for the defense earning them) than turnovers closer to the endzone that the defense will be scoring in.

In other words, there’s a reason the phrase “huck & play D” exists. Using data from Shown Space, we might say roughly that a turnover on your own endline (i.e. 70 yards from scoring) is worth about 0.2 points less than a turnover near midfield:

Using the image above as a guide, we can roughly say that 4 turnovers near midfield are worth 4*0.7 = 2.8 points for the defense, while it takes ~5.5 goalline turnovers (a 37% increase!) before you’ve won the same 2.8 points of value.

Obviously there are lots of huck turnovers against non-AP defenses, and lots of turnovers near the “line of scrimmage” for AP defenses, so the discrepancy will never be that obvious. But the general point remains: if your defense preferences huck turnovers, it’ll need to be more efficient than defenses that encourage turnovers near the line of scrimmage to make up for the lower Expected Value of having the disc that far from your goal.

Watch, for example the first point of this 2021 Bravo game against Machine, filmed by Flatball Club. Machine hucks it on their first pass, it’s a turnover, but then (after a long discussion) Bravo likewise turns the disc over, without making much progress, and Machine scores in a couple throws12:

Likewise in this sequence from a 2024 game against Revolver, Bravo forces an early huck turnover, is forced to call a timeout when their options on offense after the turn aren’t looking good…and then end up turning this disc over anyway, leading to a quick Revolver score.

Weaknesses of AP

Here are what I see as some potential weaknesses of AP defense:

• Good hucks — AP defense is, if anything, baiting hucks. Sometimes the other team will complete enough of them that they’re playing good offense. Without having data on this it’s hard to know for sure whether AP is winning this battle or not.

• Relatedly, the marginally lower value of huck turnovers — discussed above.

• Confusion on which way to force — you’ll sometimes see an AP defense give up easy yards that other defenses might not because mark and downfield defender aren’t on the same page about which way to force. I think this not a pure downside, but just a “chaos giveth and chaos taketh away” situation—having a more dynamic mark has upsides but it also has downsides. Here’s one example against Machine in 2023:

  

  If 2024 University of Colorado counts as AP defense, here’s another example where the marker switches their force direction just in time for Leo Gordon to throw an easy score to what used to be the break side a moment before.

• Point Five Frisbee — (Not that any other defense is better against fast disc movement.) Continuations are available if the offense can get them off before the mark gets into place. keeping the tempo up can lead to positioning issues and confusion for the defense and opportunities for the offense.
  Here Machine uses a combination of tempo13, give and go, and ping-pong thinking (receiver catches sliding towards the bottom of the screen, then pops up looking immediately to the top of screen):

  Or here, Doublewide gets a few continuations in a row by being ready to throw immediately:

• Too many defenders behind the disc. Teaching tight handler defense near the sideline works if the downfield defenders are also locking their assignments down, but it can lead to situations where the defense lets the offense have an easy “small numbers in big space” advantage if the downfield offense can get open. This especially happens in situations where two offensive players end up behind the disc, leaving a 4-on-4 downfield of the disc. Whether intentionally or not, Machine took advantage of this a couple times in their 2023 semifinal:

  

• Emptying the backfield — Whether by plan or not, Doublewide (2025) often ended up with no one in the backfield against Bravo’s AP defense (here’s one example). It’s risky, but without anyone near the reset space to get in the way, the “force middle” strategy ends up feeling like you’re just giving the offense the wide open side of the field. Reset cuts can come off the “front of the stack”, when needed. It wasn’t a consistent enough pattern to get more than this one bullet point, but it did catch my eye a bit.

• Good small ball — if a team doesn’t mind the small ball, and can hit the lanes, especially on upfield handler cuts, then forcing middle starts to feel like you’re just giving the offense extra room to work with. If it counts as AP defense, Colorado’s 2024 semifinal loss to Brown University is one big showcase in this downside. (See also: Slowing at the saddle point)

Is AP actually good?

In the five post-COVID years that TK has been a coach on the Bravo roster, Bravo has:

• won US Nationals once

• lost in semis once

• and lost in quarters/prequarters three times. (Never didn’t make the bracket)

These results are certainly better than a lot of successful teams (compares favorably to DiG or Revolver, for example) but arguably a bit worse than a few others (Rhino Slam, PoNY, Machine).

It’s basically impossible to say what this means about AP’s effectiveness, given that:

• a) it’s hard to evaluate talent compared to a sport like basketball, so we can’t really say if Bravo overperforms their talent level due to an innovative scheme

• b) Unlike in basketball, every single point must end with one team scoring, so we can’t reliably use scoring stats to determine the quality of a defense. For example, perhaps AP defense helps Bravo generate blocks more than their opposition, but their weak (in this hypothetical) offense means that they’re unable to score themselves, leading to them eventually getting scored on.

There is per-point turnover data available through the USAU Stats app that could theoretically help us answer point (b), but so far my attempts to contact USAU to ask for access to it have gone un-replied-to.

My personal take: there are a few areas where All Person defense aligns well with the universal defense principles I’ve written about, so I absolutely think it’s an improvement on the straight, tight person defense that many teams are still doing at the low-to-mid levels. But I don’t see it as the be-all-end-all of frisbee defenses, given a few drawbacks:

• I can’t help but think continually forcing teams to the side of the field where they have more space is suboptimal, and teams will only get better at knowing how to take advantage of this.

• The relative lack of downfield switching likewise doesn’t feel like the future

• Giving up relatively more deep shots means that the turnovers you earn are less valuable (b/c your team is less likely to score starting farther from the goal)

(Also separate from the question of AP/not-AP, I believe good defense comes from building smart players to put on that defense. I’d rather play any structure with smarter players than a marginally “better” structure but played by players with less frisbee IQ.)

To be honest, the defense that’s closest to what I believe is optimal strategy is what San Francisco Revolver was doing in 2025 (see: Notes: Revolver vs. Machine). They had a somewhat similar poachy approach in the handler space, while downfield they combined hard, switchy defense on the unders with over the top help. Maybe they’re just lacking a catchy name for their scheme to get more people talking about it. (Next post: The Some Flow Guide to Whatever Revolver’s Doing on Defense14).

In other words, I can’t help thinking to myself what if the downfield defenders played All Person defense, too? AP does a lot of interesting things with marking and with handler defense, but then gives up (mostly) on the idea of switching when it comes to the downfield defense.

It’s interesting to hear in TK’s interview that he talks about how AP came from a desire to hang with the best of the best even when the team couldn’t quite keep up athletically. But later on, when talking about games where Bravo got beat deep a lot, he suggests he’d rather let Zeke [Thoreson] “use his amazing physical tools” than using the “band-aid” of helping in the deep space. But in my opinion, you’re always going to end up playing against offenses that are just as athletic as you, and you’re always going to be at a disadvantage b/c the offense gets to move first. So I can’t help but think that even the most athletic defenders need to be able to play the smartest team defense to be truly elite.

This play stood out to me watching the tape as a strong example of the suboptimal downfield defense. After Joe White catches the pass, there’s two defenders on the top sideline. It seems like a perfect opportunity for some bracketing, but both are playing on the under side. And then…one of them gets beat under anyway, while the other simultaneously gets beat deep:

I can’t help but think it’s possible to do more in the downfield space.

Sources:

• Reddit threads

  • AP Defense (2026 April)

  • I want to learn AP defense (2026 March)

  • FM vs Directional Force (April 2025)

  • Colorado’s AP by TK with Cody Spicer (Video Episode) | Ep 84 (2025 April)

  • Is Force Middle over used now? (2024 April)

  • All Person (AP) Defense (2024 March)

  • Force Middle (December 2022)

  • Sources to learn force middle (August 2022)

• Podcasts

  • Shayla Harris on Pod Practice

  • Cody Spicer on Pod Practice

  • Tim Kefalas on Pod Practice

• Games

  • 2026 Pro Elite Challenge vs. Rhino (just uploaded! But it feels like they’re not playing AP in this game, at least for the most part)

  • 2025 Regional Final vs Doublewide

  • 2024 US Open Pool play vs Revolver

  • 2023 National Semifinal vs Machine (Vimeo)

  • 2023 US Open Semifinal vs Truck Stop (Vimeo)

  • 2022 National Final vs Truck Stop (Vimeo)

  • 2022 Nationals pool play vs Truck Stop

  • 2022 US Open Semifinal vs DiG (Vimeo)

  • 2021 Pro Champs vs Machine

  • There are many more Bravo games on the YT channel of Richard Yuan but all the search results I saw are filmed from field height

  • Also many more games in Ultiworld’s (paywalled) library of course

  • Full list via Ultiverse

• Other

  • How Bravo Slammed the Brakes on Truck (Ultiworld, paywalled)

  • Tim Kefalas’s blog (not actually that useful. He mentions on Pod Practice podcast that he deleted some old blog posts, trade secrets I suppose) (Here is his Ultimate-Reference page)

Bravo’s results at recent Nationals (via ultimate-reference):

Introduction at the end so the article doesn’t start with me talking about myself

I wanted to write up a breakdown of All Person defense for a few reasons. First, there had been a few previous Reddit threads (see Sources, above) asking how it works. Second, a Reddit user put together this post, using a combination of podcast transcripts, their own film study, and AI help. One of the replies called it an “extremely good post”, but, to be honest, I wasn’t that impressed by it (which is not to say it wasn’t better than a lot of what was out there before)15.

Third, Tim Kefalas mentions in his appearance on Pod Practice that he’s deleted old blog post(s?) to keep other teams from learning his trade secrets. And the only Ultiworld article analyzing Johnny Bravo’s AP defense is paywalled (again, see Sources above). But I take a different approach to life, and I like to see information be free. Trade secrets help one team, but open information pushes our entire culture forward (see Finite and Infinite Games).

So all that made me want to do a good writeup.

Now for the usual caveats:

I’m not a Johnny Bravo insider, so I may have gotten some of this wrong. But I’ve done my best to share what I see on the film from my own perspective, and what I hear the insiders talking when they’ve discussed AP on podcasts.

Bravo won’t play AP on every single point. Some of the points in the 2023 Machine game feature some under←→deep downfield bracketing, the 2024 game against Revolver has them playing something that looks like a matchup zone for significant periods, etc. Perhaps some of these tweaks are the next evolution of AP—resolving some of my complaints above about the lack of downfield teamwork—and it just hasn’t fully shown up on the tape yet.

As always, suggestions and corrections are very welcome!

In a new YouTube video, author David Epstein claims1 that experienced throwers don’t focus their eyes so much on the receivers themselves, but rather focus on the spaces between players:

If you did an eyetracking study of me and watched me play quarterback, you’d see me following the person I want to throw to. And if they were covered, I’d switch and follow another player. But if you did an eyetracking study of Tom Brady, you’d see him looking at spaces between players that give information about the relationship of the different parts and what’s going to happen next. [Starting at about 10:30]

So to him, it’s not even about “look at the receiver, but with soft eyes”, but rather that the best throwers don’t even necessarily have their focus centered on the player they’re throwing to (at least until the decision is made). How many of our throwers are learning to do this as young frisbee players?

Sometime soon after Ian’s article came out, my YouTube algorithm suggested a video by Michael MacKelvie on a phenomenon that movement researchers call “quiet eye”:

Quiet eye is, at least on a surface explanation, kinda the opposite of what Ian suggested: researchers have found that elite athletes spend longer focused intensely on the object they’re trying to catch (or shoot at, etc). One of the top YouTube comments provides a pithy summary:

This was something of a surprise when it was discovered, because, as MacKelvie explains, there were reasons to believe the opposite should be true:

It’s important to understand the consensus belief at the time. And that is: elite athletes directly perceive and observe their environment quicker than the rest of us...and because of that, the duration of their gaze should be shorter.

But it turned out that’s not true, and that elite athletes spent more time focused intensely on the golf ball they were about to putt, the hockey puck they were trying to save from going in the net, or the basketball rim they were about to shoot a free throw towards:

So, focusing is an important skill, but having a certain unfocus is a skill too. And while “quiet” and “soft” are two names with similar vibes, the skills they refer to are sort of on opposite ends of the spectrum.

Side note: Another video that came out in response to MacKelvie’s, with a somewhat more conversational (less data-driven) tone says that “I firmly believe that eyesight, vision, and perception in professional sports is the most wildly under-discussed subject out there today, at least among mainstream fans.” The video’s author, one Anthony Lawson, shares a story about NFL receiver Justin Jefferson who has the (supposedly pretty unique) ability to “change his eye dominance“—i.e., which eye is mainly used when both are open—depending on where he is on the field (e.g. needing to turn his head to look over his right shoulder vs. left).

He then shares that another all-time NFL receiver, Larry Fitzgerald, is the grandson of an optometrist who helped train Larry’s vision when he was a boy. I don’t know what these anecdotes mean, but they’re provocative.

For a third eye skill: I’ve written a few times about how elite athletes literally turn their head in different directions to take in more information while playing. Soft eyes is using your peripheral vision; this is actually turning your head around. There are a number of examples of this in Elite players looking away from the disc, and there are a few more examples in my post Oodles of OODAs, namely these videos of Levke Walczak and Draymond Green playing defense:

This skill goes by the name “scanning” in the sports world, as in the title of this video, The Art of Scanning: Football’s Most Underrated Superpower2:

This video adds some interesting context as well: the importance of timing your scanning. There are critical moments when you need to be focused on the ball (disc), and others when “nothing is happening” at the ball, when you benefit from looking away to scan the field. To give an example, when the disc is in flight, its future path has already been determined (wind gusts notwithstanding), so a downfield cutter should have time to make quick scan of the field instead of watching the disc throughout the entirety of a looping swing pass. Even when you’re the one receiving a pass, many skilled soccer players will look away from the ball mid-pass for a moment for one last “critical scan”.

Great soccer players, says the video, might scan 6-8 times in the 10 seconds before they receive a pass. The challenge is that it won’t always be easy to differentiate those moments and know when it’s “safe” to look away and scan. I’m just not sure there’s any solution faster than playing frisbee for years to build up your intuition.

So, how does it all come together?

I pointed out how, on the surface, it seems like “quiet eye” and “soft eyes” might be at odds, but it’s pretty clear that actually everyone is correct here. We’re just all talking about slightly different skills, that are used by athletes in different contexts.

The “quiet eye” is specifically used in the “final critical phase” of a throwing/catching/hitting movement. It’s not mentioned explicitly in MacKelvie’s video, but like with scanning, I imagine there’s an element of decision-making involved in achieving elite performance: top athletes are able to choose the right moment to commit to “quiet eye”-mode. Scanning/soft eyes help you decide whether a throw is a good idea, then a skilled quiet eye helps you lock in on your receiver in that last 0.5-1.0 seconds where you’re actually throwing.

The rest of the time, athletes are using scanning and soft eyes to gather as much data they can about the on-field environment. This helps you see what’s happening on the field, so you have the context to make good decisions when you finally decide on a course of action and switch to “quiet eye mode”. Soft eyes help with what’s already in front of you, while scanning helps with the times when you’ll need to turn your head if you want to take in enough information to act on.

Eyes or brain?

As MacKelvie touches on in his video—and as some of the YouTube commenters mention below his video—it’s hard to say how much these “eye skills” are actually “brain skills”. (Ian French touches on this as well, pointing out how soft eyes are easier when you’re comfortable/not under pressure.) Not to mention that the way you use your body can play a role too, like learning to “face the space” to make it easy as possible to take in as much as possible.

When I first wrote about these concepts in Oodles of OODAs, I framed it as more of a brain skill—as the mental model of understanding that your goal is to be constantly taking in and responding to new information, over and over again. The way you use your eyes is, in that article, more of an afterthought. It’s basically saying: your goal is to repeatedly take in information and respond to it, so yeah of course you’re gonna be using your eyes to do that.

So where am I now? I think my current take on this is that it’s probably somewhere in the middle. You’ll never develop elite “eye skills” simply by trying to train your eye skills directly. You need to play frisbee enough, and for long enough, that you learn the comfort (that lets you feel relaxed on the field and thus able to relax your focus) and build the intuitions to know where (and when) to look on the field to gain useful information. But...

...Finding ways to train these eye skills probably will help you get there a bit faster, too. Play games that involve an element of noticing something away from the main focus of the action. (Here’s an idea Levke Walczak offered in a presentation she did for the European Ultimate Federation). Build up team mottos around the idea of scanning the field (”Pictures!“ is a phrase mentioned in the scanning video linked above). Find ways to force both of your eyes to work (playing catch with one eye closed—switching eyes every so often—is an idea I’ve experimented with lately). Try one of the myriad drills available online to better train your quiet eye.

Anyway, it seemed funny to me how the various media on how we use our eyes in sport tended to focus on one particular skill at a time (The video on scanning doesn’t mention soft eyes...The article on soft eyes doesn’t mention quiet eyes). But really there’s a number of different eye skills being used at different moments—not to mention all the work we’re doing with our brain to process that information—and these are all independent skills that we can develop.

The first person to mention this idea to me was Jonathan “Sam” Goldsmith of Disc-i (thank you to him also for finding some of the links further down in this article). Since then I’ve also seen it mentioned elsewhere, like in this YouTube video that the SmarterEveryDay channel made about disc golf aerodynamics (timestamp starts at 46:40 but the clearest statement that launch positive nose negative is desired is at 47:50).

But it seems this concept isn’t super well known in frisbee, so here’s a short writeup to help change that.

Definitions

Launch angle is basically “the direction you throw the disc”. If I were to try to give a strict definition, I think I’d say it’s the direction of motion of the disc’s center of gravity immediately upon leaving your hand (i.e. before gravity or lift forces have acted to change the disc’s direction of motion). Or, say, the direction a ball would go if you were throwing a ball and not a disc.

If the disc has a negative launch angle, you’re throwing it at the ground. It’s not going very far before landing (unless you’re throwing an airbounce). If the disc has a positive launch angle, you’re throwing it (at least somewhat) up into the air.

Nose angle/angle of attack is basically “the direction the front of the disc is pointing”, relative to the launch angle. If you have positive launch angle (throwing the disc somewhat up into the air) but the nose is pointing exactly the same direction you threw the disc, the nose angle is zero (even though it’s pointed up relative to the ground).

So this is zero launch angle and zero angle of attack:

A true airbounce has an extremely large (positive) nose angle, since the launch angle is negative—pointed towards the ground—but the nose is pointed up. Here’s the airbounce with its negative launch angle and positive nose:

This is positive launch angle and zero nose angle (since the nose is pointing the same exact direction as the launch angle):

And this is what we want on hucks, a positive launch angle, but with the nose pointed slightly down relative to that:

An underappreciated insight

As far as I’m aware, this concept isn’t well known in Ultimate.

This article from Excel Ultimate, the first thing that came up when I searched for “nose angle ultimate frisbee”, says that:

To get arc on a huck, you need a touch of nose-up. Without it, gravity drags the disc down too early.

To be fair, they’re not quite wrong! Remember that nose angle is relative to launch angle, so when you have positive launch angle and the nose pointed down slightly relative to that, the nose is still pointed up slightly, relative to the ground. But by conflating nose angle and launch angle, Excel’s explanation isn’t as insightful as it could be.

Nor does this Ultiworld article mention the launch positive nose negative concept. And that’s even though it’s perhaps the most thorough treatment of the backhand you can find anywhere and even has a section titled “Angle of Release” that starts with a paragraph containing the sentence “We will borrow some terms from disc golf, where disc flight is typically described with more detail and precision”.

I don’t want to research every ultimate frisbee backhand article that’s ever been written to prove this concept is under-discussed so I’ll leave it at those two examples for now.2

This 2025 article from TechDisc does a good job explaining this knowledge in a disc golf context, suggesting (for their particular throwing simulation) an 8 to 10 degree launch angle combined with a -3 to -5 degree nose angle led to optimal distance. (This forum post from back in 2016 was already estimating 10 to 15 degree launch angle combined with -4 degree nose angle, so this insight isn’t super new by any means.)

Given that we’re playing frisbee and interested in different things (I usually want the disc to float a bit more for my receiver and don’t mind if that means I lose a tad off my maximum distance) I wouldn’t take those numbers as exactly what you want to do with a frisbee, but just a general ballpark of how this concept works.

Why does this work?

The “positive launch angle” part is the easiest to explain. If you’re throwing a huck, you’re basically guaranteed to be releasing the disc about 3-5 feet off the ground (sorry, Wemby, this article isn’t for you). So if you don’t have a positive launch angle, your disc will never get higher than that same 3-5 feet off the ground. This is bad for two reasons.

• First, because it’s at risk of hitting other players before it reaches the target if it’s only 5 feet off the ground, and,

• Second, because if you want the disc to float down nicely, it needs some height to float down from.

So, to achieve those two goals, you have to throw the disc up into the air a bit. Without thinking about or researching it too much, I’d guess the best huckers are getting 10-20 feet of maximum height on their hucks.

So then, why point the nose down slightly relative to your release angle? Well, with a positive release angle you’re going to be catching some air under the disc. If the nose is pointed too far up relative to the ground, the disc will slow down3 and can even end up significantly falling back towards the thrower before landing, as I’m sure we’ve all seen happen4. (It is also possibly counteracting the tendency for the aerodynamic forces to lift the nose near the end of flight)

Giving the nose a negative angle of attack allows you to reap the benefits of getting the disc 10-20 feet up in the air without making it so floaty that your throws become ineffective. (Getting the nose down is especially useful when throwing into an oncoming wind, as the wind pushes the disc up even more than usual.)

Another way I like to think of it is, the disc always falls towards its lowest edge, whether that’s IO/OI or nose up/nose down. Keeping the nose down helps guarantee that your hucks—even when you give them some height—will have a bit more tendency to “fall forward” instead of falling back towards you.

Final thoughts/caveats

Remember this advice is specifically for long throws. The logic that makes this type of throw effective isn’t relevant on short stuff. And I’d perhaps even say it’s specifically for hucks, since for pulls (some of them at least), the angles you use may be determined by other factors: perhaps you want it to come down blady so it’s hard to catch; or you may be happy to let it float as long as possible—even if it loses a few yards on the way down—if that gives your defense extra time to get in position.

But for hucks we often want that combination of distance, height, a bit of floatiness to give your receiver time to adjust, and precision accuracy that makes “launch angle positive, angle of attack negative” an effective tip.

I’m sure some readers know about this already and some more do it already because it’s what works without having explicit knowledge of the concept. But if you struggle to get height on your hucks without also turning them into floatfests, this may be the tip you’ve been lacking.

What I meant by that, in short, is: LeBron is better than me at basically every single possible facet of basketball skill. So why do I see so many pro frisbee players who act like they’re more scared of catching pulls than I am? Haven’t they gotten used to blady pulls by now? How do you expect to win a championship if you don’t have a championship-level pull catcher?

Then watching the college nationals recap Ultiworld podcast, I saw this comment from Marshall Lian:

So, yeah, shout out to Griffin Gee of UMass! Caught pulls like a champion and then his team won the championship. Here were a few of my favorite examples of his pull fielding (plus one subpar example):

I swear that on this pull from the semifinal vs. Oregon you can sort of hear the crowd react to the catch of the falling blade pull. If he doesn’t catch this IMO there’s basically no way the center handler still ends up with the disc that far downfield:

I would’ve liked to see him get to the backline earlier on this next pull to make it easier to determine whether it was going to land out. That quibble aside, it’s a gutsy catch that potentially saves his team from getting stuck in the back of the endzone:

Here’s a few more where the pull catch helped the offense get started just a few beats earlier:

Finally, here’s one he bailed out from at the last second, but I still give him some props for having the presence of mind to play it safe when he didn’t feel comfortable (also chalk this up as another misread to the high side?):

The nightmare scenario

Late in the same game, CUT showed exactly why handling pulls can be so critical: a misjudged pull rolls out the side of the endzone and CUT is forced to start their possession from the cone. They throw a first-throw turnover when no one gets open, essentially killing the last hope they had at staying in the game:

(Also: note the nice stunt and recover from the far-side handler defender, which helps keep the thrower from seeing either of the other handlers as good options to throw to)

This play was mentioned on the same Ultiworld podcast as well.

Bonus strategy nuance: Note how CUT had no player “backstopping” the pull catcher, while UMass had another player backing up Griffin on the pull where he bailed. Especially important on turf fields (and windy days on grass) when those pulls can really roll far after landing. It seems like most UFA teams have figured this out by now (given their different rules around pulling makes it even more critical) but clearly it hasn’t become standard at every level yet.

Be(come) a championship-level pull catcher

Catching pulls is good. Even compared to trapping the pull quickly after a bounce, catching it in the air and getting started an extra 1-2 seconds earlier gives your offense just a few more free yards to start with. I expect the psychological factor of showing yourselves and your opponents that you aren’t scared is also a boost. Yes, a pull will get dropped every once in a long while, but my hypothesis is that it’s not that hard to get so good at catching pulls that the per-possession gains in Expected Value make it worth it.

If you need more convincing, read this article from last year: Why collecting pulls matters. Most relevant quote:

One of the implications of this philosophy is that you should catch every pull you possibly can. In my opinion, it’s much easier to get good at catching pulls—and then attacking immediately for an easy 15+ yards—than it is to get good at doing all the other things it takes to gain those same 15 yards on a defense that’s already set up. (In other words, Be good at frisbee, not just frisbee-like things—it’s not going to show up in the stat sheet that you caught a pull most other people would’ve let hit the ground. But it will show up in your team winning marginally more games.)

So if you want to win a championship, don’t overlook the benefit of having a championship-level pull catcher.

I’m surprised we don’t see more teams and players investing the time to get great at catching pulls.

On the player side, it gives you one extra edge to convince the team leadership to put you on the O-line. If leadership has to choose between two players who are otherwise equal but one of them can say: “put me on the O-line because I will catch every pull and get our offense started gaining yards before the defense gets downfield”... who wouldn’t want to have that player? It’s a cheat code to unlock more opportunities for yourself.

On the coaching side, it’s one more marginal gain to put your team a step ahead of the competition. Why not ask one player to invest some extra time getting super comfortable catching blady pulls? Aren’t you looking for every edge you can find?

I find it difficult to believe that the average team is anywhere near as good at catching pulls as they could be. Players will go out for hour-long throwing sessions multiple times a week, but how often is anyone ever spending even a half hour a week practicing their blady pull catches? I’d bet it’s not that hard to get to a point where basically no pull scares you.

If you don’t always have someone to practice with, one training drill I’ve experimented with is simply throwing a vertical disc straight up in the air as far as I possibly can and then catching when it comes down. Find ways to make it extra challenging when you get good (one hand only; or, don’t move either of your hands from your side until the last possible moment, etc). It’s not the exact same feel as catching a blady pull but it’s better than nothing.

And if you’re just getting started on this journey, let me point out the negative/positive feedback loops at play here in case they help convince you: when you don’t catch pulls, you don’t get any better at catching pulls, and it just continues to be too scary. But the more pulls you catch, the more you feel comfortable catching pulls, meaning you then get even more practice, leading to even more comfort.

One final thought: it seems silly to me that we basically only see handlers fielding pulls. Why do the players playing in the backfield have to be the ones tasked with catching blady pulls? Those two skills have nothing to do with each other! Let a “cutter” catch the pull if they’re good at it, then they can jog down the sideline to get downfield (it’s not like those extra players in your side stack are really doing anything, anyway).

PS I’m writing this post in the library, and what do I see when I look up from my computer screen but the spine of a book written by someone named “Joy Fielding“. Be like her and go find pull fielding joy!

[Update (2026-07-27)]:

In the original version of this article I didn’t really try too hard to prove the point that teams are letting pulls hit the ground that could totally be caught. But since then, I’ve been on the lookout. Check out the 2023 US National Semifinal (Johnny Bravo vs. Machine). I don’t know how to timestamp-link Vimeo videos, but my notes have 7:04, 9:04, 28:12, 1:23:35, and 1:56:45 as examples of pulls that very much could’ve been caught.

I think “traditional” frisbee offenses often make this challenging moment even harder than it needs to be by initiating the offense with high-degree-of-difficulty (higher than necessary, at least) passes.

For example, Brown University’s 2022 college championship game loss to UNC started with two consecutive turnovers on unnecessarily difficult throws (clip below). They lost that game, but came back two years later to win with a retooled offense that made things look easy.

I’d rather see offenses start with something easy as possible against a set defense. That could be “dominator”-style handler movement to get the disc to players in motion who can more easily attack the breakside with throws. It could be starting with short downfield cuts from the front of the stack instead of long downfield cuts from the back of the stack. It also means being happy to punish handler poaches with short gains on swings instead of wasting time staring downfield trying to throw past the poach.

We need to recognize (as some teams already do) that offense against a set D is hard, and that the path to efficiency is doing easy things when the D has the advantage. Let’s look at some examples, and some further thoughts about the mindsets at play here.

Dominoes

Basketball coaches (as I’ve mentioned before) use the concept of dominoes:...once you’ve got a small advantage—once you’ve knocked over the first domino—can you keep the dominoes falling until you reach the end of the chain, i.e. a high quality shot?

But there’s one thing I think could be improved about that metaphor. To me, the default mental image for a “row of dominoes” is something like this:

But what we really should be thinking about is a domino chain that looks like this:

The real power of a domino chain is not that one domino can knock over the next one, but that a small domino with momentum can knock over progressively bigger dominoes.

Our goal shouldn’t be to initiate the possession by gaining an immediate big advantage and then maintaining that advantage at a constant size. Possessions in basketball don’t (usually) start by throwing some crazy no-look pass through four defenders. They start with, say, a jab step that gets a defender to move 4 inches backwards.

When the defense is set, that’s when it’s hardest to play offense. So really what we should be doing is gaining a tiny advantage against a well-prepared defense, then a slightly bigger advantage against a slightly out of position defense, and so on until we reach the “full-sized” dominoes at the end—an easy goal.

(I do think dominoes are still a useful metaphor even if all the dominoes are the same size—you still want to play quickly to keep advantages rolling instead of letting your momentum stop—but the real power of the domino chain comes in starting with a tiny push that eventually ends with toppling a huge obstacle.)

Some examples

Let’s look at a few examples of first throw turnovers and/or first throw stagnation, plus a couple times where teams did things I like more.

Here’s a sequence I noted from a recent 2026 College Nationals game. The UNC handler picks up the disc and stares right into a poach defender for a solid 4 seconds before turning and shoveling a pass to the other handler. Even though the poaching defender was probably ~15 yards away from the reset handler, UNC doesn’t really manage to generate any advantage out of it (they don’t gain yards and they don’t really gain free access to the break side).

Later in the game, the UNC center handler throws a first throw turn (for not the first time, eg see here or here) and one of the announcers comments on it: “It’s just another first throw turn...I think they’re asking her to do a lot, and maybe too much, because it’s just giving the disc back.”

The 2025 South Central Regional Final between Johnny Bravo and Doublewide that I was watching recently likewise had two first throw turns—one from each team—both on throws trying to reach the breakside.

The 2025 Toronto Rush started out a game against the Boston Glory with a first throw turnover. It doesn’t seem like the highest degree-of-difficulty throw, and you can see in the background trees that there’s some swirling wind, but they did look away from an open pass back to the handler on the open side in order to throw a pass to the break side.

And here’s an example from the UFA that I used in a previous article. Much like the first clip above, the handler looks downfield a little too long, and doesn’t really generate any advantage out of some pretty extreme handler poaching:

This clip from the 2025 U-24 world championship match has the honor of involving 2 first throw turnovers (followed by a close call, that’s only saved from being a third consecutive first-throw turnover because it’s caught by an unintended receiver). The first throw, by Chloe Hakimi, isn’t a break throw like some of the examples we saw before, but it’s a bit aggressive to throw a first pass into basically a double team on a player who’s got barely a half-step of separation. (The announcer says: “quickly picked up...and thrown away. Somewhat surprising...” But I’d have to disagree...I think it’s actually less surprising than it should be.)

After the US turn, Canada hucks it deep on the first throw. Not the craziest decision as the cutter had some separation, but not a great throw. Then the US nearly turns the disc over on an overthrown 25-yard under caught by an unintended receiver downfield. But there was really no need to default to the big 25-yard throw when there was so much open space in front of the thrower.

Let’s take this clip from 2022 Medellin Revolution as an example of what I’d like to see more of. The disc is “centered” after a pull, but, instead of the handler turning around and looking downfield for the big gain, there’s two immediate swing passes and then another swing back to the middle before the disc makes its way to Levke Walczak cutting from downfield. The extra bit of side-to-side passing leaves Levke’s defender just a bit more unsure where the threatening spaces are, making Levke’s juke just a bit more effective:

The New York Empire have also been doing a good job of starting small, based on a quick skim through one of their recently streamed games, see here or here or here for example.

Here’s one final clip, this time from universe point at 2025 club nationals. I don’t think this is the most egregious example, it’s just a good defensive play as much as anything. But I don’t love running an offense where no player even gets within 10 yards of the thrower. With the disc stuck on the sideline no less, a known danger zone for offenses everywhere.

Why not just isolate Ella Juengst 5 yards in front of the disc, then focus on the bigger chunks once you get the disc off the sideline? Compare that to Revolver, who, when they get stuck on the sideline after a pull in their 2025 nationals run, immediately throw a short backwards pass on stall zero:

Hero-ism

Let’s change tack a bit and talk about the psychology of these kinds of decisions.

In her book Right Kind of Wrong (subtitle: The Science of Failing Well), author Amy Edmondson writes about the “hero feeling” that comes from individuals finding workarounds to the problems they face, and how those feelings of heroism cause people to be less interested in fixing the system that caused heroics to be required in the first place. It’s a longer quote but I think it really applies to situations we see on the frisbee field too:

Boston University professor Anita Tucker...observed that nurses confronted “process failures” surprisingly often—almost one an hour. A process failure was anything that disrupted a nurse’s ability to complete a task, such as an unexpected supply shortage in bed linens or medications. The nurses were acutely aware of these frustrating daily hurdles. Their jobs were hard enough!

...What we called “first-order problem-solving” was a work-around to complete the task without addressing causes of the problem. For example, a nurse working the night shift who ran out of clean linens to change her patients’ beds simply walked to another unit that had linens and took from their supply. Problem solved. The work-around required minimal time and effort. She’d taken the initiative and been resourceful in caring for her patients.

Never mind that the other unit was now in short supply...Like all quick fixes, the nurses’ work-arounds created an illusion of effectiveness. Confront a problem, implement a work-around, get on with your day. End of story. Except that it’s not. When we analyzed the hospital nursing care as a system, we realized that work-arounds, despite being effective in the short term, actually made the system worse over time...

...many of the nurses we spoke to described a “hero feeling” from using work-arounds that ensured that patients got the care they deserved. Whether walking down the hall to locate the extra linens or going to the pharmacy to obtain a missing drug, the nurses derived gratification from overcoming the many little hurdles their jobs threw in their path. But this hero feeling lessened their motivation to engage in second-order problem-solving [that is, improving the larger system so the workarounds aren’t necessary in the first place]...

Much like getting linens from another department is *a* solution for nurses whose own supply closet is empty, attempting a hard pass to the breakside is certainly *a* way to manufacture an advantage for your offense. But I question whether it’s the most effective way. Could a systemic change more consistently manufacture those easy advantages?

I expect that “hero feeling” exists on the frisbee field, too. Players who take pride in being able to break the mark (a first-order solution to the problem of beating the defense) are less interested in switching to an offensive system that doesn’t require them to complete such high degree-of-difficulty throws (which would be a second-order solution to the problem of good defense).

But, as other sports have shown us, offenses are more robust and more effective when they’re built on strong systemic principles instead of on individual brilliance. (Which is not to say individual brilliance is bad! Just that it shouldn’t be your main offensive system). These supremely talented players would look even better in dominoes-based, start-easy systems that manufacture easy opportunities for them instead of big-chunk-only systems that result in occasional early turnovers. (See: Mac Hecht being ready and willing to throw a backwards pass to get off the sideline on stall zero.)

Abusing poaches is your offense

To me, one of the lessons here is that offenses need to learn to be happy to make the defense pay, immediately, for poaching. Especially the kinds of obvious handler poaching we saw in a couple of these clips.

Too many handlers still seem to think their job is to throw the disc downfield, and they try a little too hard not to change that job description, even when the defense puts an extra player in the way.

But really the handler’s job is to make the defense pay, in whatever way the defense lets them. It’s only by showing the defense that you know how to punish their poaching that you’ll earn the right to throw the disc downfield. If the defense shuts down plan A (as any smart defense would do), you’ll maximize your chances to use plan A by immediately showing the defense you’re perfectly happy to go with plan B, too.

Caveats

The plural of anecdote is not data, as they say, so the fact I can find ~6 examples of first throw turnovers admittedly isn’t proof of anything. I’m sure there are tons more examples of turnovers that happened trying to start too big, but also there are many examples of turnovers that happened when doing “easy” stuff, too.

So I’m not offended if a few examples aren’t enough to fully convince you. But I do think there’s a real tendency here. Whether it’s a tendency to always look downfield first, or a tendency to personally enjoy successfully doing hard things like breaking the mark, or just the systemic style of playing “big chunk” offense looking for 25-yard gains and breakmark throws that solve problems with one pass instead of a system that builds up advantages to make gaining yards easy.

And I’m not saying never, ever make a hard throw. I just suspect the hard/easy ratio is off. The threat of a big throw opens up the small stuff, and the threat of small stuff opens up the big stuff, too. I’m just saying it seems like not enough teams do a good job threatening (and carrying out the threat of) the small stuff.

Final thoughts

I’d like to see the frisbee community establish more cultural memes around the idea of starting with something easy instead of something hard.

• Partly that means more willingness to abuse poaches immediately instead of looking downfield past them.

• Partly that could mean embracing a more “dominator” style at the start of points, to get the defense reacting to a few easy short passes before attacking downfield.

• Or it could mean adapting Brown’s 2024 style with short, “front of stack”-esque isolation play, even if only when initiating the offense.

• And partly it means embracing the “dominoes” concept as a mental model of building up your offensive attack. (dominator for dominoes?)

What do y’all think, do offenses actually turn it over on the first pass more than they should, or does it happen at the appropriate rate and I just cherry-picked the few examples I found over the course of a couple years of frisbee watching? (I don’t feel like I had to work that hard to find these examples...)

Update (2026-06-16):

Last year my predictions were barely better than chance, and lost (as they had previously) to Oliver’s data-driven method. I said last year (2025) that I was less confident in my predictions than in 2024. This year my meta-prediction is that I feel more confident than I did in 2025.

A couple new things to note this year:

• I recruited Sammy and Sam to also provide predictions. (If you want me to post your predictions next year too, get in touch)

• Oliver had two AIs (Claude and Gemini) also generate predictions. They both roughly do what he was already doing with his own spreadsheet-based methods, so I don’t think it’s surprising that those three methods came up with generally similar ideas.

Here are the predictions:

Here’s the average of all six rankings:

Lime and especially Pink win the award for being ranked all over the place. Navy has the steadiest ranking, with none of the six predictions putting them below 4th.

The other obvious point of comparison is to look at the average for the three humans compared to the average for the three numbers-only methods (Counting Oliver as “the numbers” since his ranking is purely numbers driven). Those rankings are below:

The humans like Gold and Pink more than the machines do. The machines rank Atomic and Purple higher than the humans do.

For the most part, the machines have very similar predictions to each other—very small spreads. Eight teams have a difference of 0,1, or 2 between their best and worst machine rating. Forest wins the “The Machines Can’t Decide What To Think of You” award, with a 9-spot spread between their worst and best ratings, while no other team has a machine spread of more than 5.

There’s a good chance the full average outperforms any of us. The humans and machines each have legitimate, independent insights1 and averaging them together probably helps combine that knowledge while canceling out each party’s biases.

Most importantly, all teams are predicted to have lots of fun!

So I threw together a few lines of code to add up turnovers per year across each division.

Long story short: I didn’t end up thinking there was anything too interesting to be discovered here. There’s no evidence of any big shifts in the balance of power between offense and defense between 2018 (when Nationals moved to the San Diego area) and now.

But I’ll post it anyway (in the spirit of the Journal of Negative Results) because I’m not aware of anywhere else this data is already readily available.

So here’s the data for...

...the Men’s division:

...the Women’s division:

...and the Mixed division:

Quick Discussion

The Men’s and Women’s divisions both saw their lowest turnover rates in 2018 and 2025. The Mixed division also saw low turnovers in 2025, but 2018 wasn’t an especially clean year for them.

As discussed previously the wind conditions in 2016 and 2017 (in Rockford, IL and Sarasota, FL) had a huge impact on the number of turnovers.

There doesn’t seem to be anything interesting about the San Diego years—no clear overall trend, so it’s probably just driven by random variations in how windy the tournament is each year. I would’ve made a graph but it seems pretty obvious the graph wouldn’t make it any more interesting than a table does.

2021 seems to be the worst or near-worst year for all the divisions...maybe that was a windy year or maybe there’s also a “teams were rusty coming back from COVID” factor?

For more data on conversion rates, the UFA team stats page is surprisingly interesting. A few of the best teams actually had better conversion rates on the D-line than on the O-line in 2025:

There are probably a variety of reasons some great teams have such high D-line conversion rates. One of those reasons is surely the “get a lot of break scores beating up on weak opponents” effect. But while this might explain the Union (padding their stats against the Mechanix?), it doesn’t really explain the Glory who played against more consistently strong opposition.

The Spiders were the one huge outlier from a clear “the winningest teams have a high D-line conversion %” trend, instead getting by on a highly efficient offense (basically tied for best hold% in the league).

It’s hard to compare UFA and Men’s club directly since I’m not sure if the UFA data excludes buzzer beater attempts. Though I was intrigued by the most recent post from Katfish Corner where he shares that the Colorado Summit got a higher percentage of their scores from breaks than Johnny Bravo did (with basically the same players on each roster):

Unfortunately there’s no way to separately look at club O-line vs D-line conversion rates with the data we have available. There’s one club data point I do have — Joe Marmerstein told me via email a couple years ago that 2024 national champion Rhino Slam had 58 scores on 89 O-line possessions, a 65.2% conversion rate.

2024 Rhino is actually another team that played statistically cleaner D-offense than O-offense: accounting for the team’s remaining goals and turnovers via the team’s stat page tells us the D-line had 30 scores on 42 possessions — a 71% conversion rate.

The differences between gender divisions matches up with previous data like this Reddit post, this Ultiworld post, the Reddit post linked below, etc.

Notes:

• This 2019 Reddit post gives me confidence I did my math right in adding up the goals/turnovers. We get identical numbers for Men’s/Women’s. Their number is slightly lower for Mixed for some reason...but I trust my data, I think they may have missed something small.

• As pointed out in the comments of that post, these numbers are obviously not perfectly accurate, statkeeping is hard. But I do expect they’re generally close enough to reality to be worth looking at.

• Conversion rate is number of goals divided by total number of possessions, i.e. the percentage of time that you score when you have the disc. Since every possession either ends in a goal or turnover, the calculation is then Conversion Rate = goals / (goals + turnovers).

• I discovered in the process of putting this post together1 that Turnovers weren’t recorded in the 2014 data. That’s why 2014 isn’t included in the tables above. I’ll add a note to the original Nationals Stats post to point that out.

• The number of games varies between divisions because some games are missing their box score data (or weren’t played at all due to teams sometimes forfeiting their final consolation game). The details of this are included in Update on Career Nationals Stats.

• Turnovers per game is the total for both teams, so “20 turnovers per game” means (on average) each team turns the disc over about 10 times.

—I added a few games (20ish, details below) to the database that were originally missed due to a combination of the limitations of the script I built and some broken links on the USAU website. All the tables in the original post have been update to reflect these small changes. Mainly it added a few stats for Women’s division players who were playing in 2015.

For example:

• Kami Groom’s 2015 is now on the leaderboard of best single year +/-, and is also added to the list of single-year triple doubles

• Robin Fennig is now correctly shown at over 200 career assists.

• Many women’s division players (Groom, Fennig, Sandy Jorgenson, Kaela Helton) have moved up a couple spots on various career leaderboards.

—I updated the tables in the original post to correctly display ties between players, i.e. players tied with 7 single-game goals will all show up as “T-2” instead of “2,3,4,...”

—I updated the script to directly pull the game’s final score (instead of calculating it based on the sum of goals recorded in each team’s box score). I believe this method is much more accurate, so I expanded the sections in the original post discussing wins/point differential/winning percentage. Here’s a couple of the tables I added (a few more are in original post):

— I added a table of “near triple-doubles” to shout out the players who’ve come one stat point from the goals-assists-blocks triple double:

—I reformatted the way player names are displayed in the single-year database, which should make it easier to sort by name and see one player’s career at a glance:

This same change should also resolve the previous issue that multiple players were merged into one if they had the same exact name. (For single-year stats at least, I don’t have a good way to fix that for full career stats).

Some notes on data quality

I added some code for error checking to be able to provide more specifics about where the data is imperfect. Here’s what I found out.

There have been, by my count, 1,779 games “played” at Nationals since 2014 (including 2014). Take the games per year, and multiply by 3 for the three divisions:(55+55+53+55+53+53+56+53+53+53+54)*3 = 1,779.

Of those games, data is not always available, or not always fully available. Here are the details on what’s missing and what may be slightly incorrect:

1. 18 games from the 2015 Women’s division (the entire 15 game championship bracket, plus the “Pro Flight 5th place” consolation bracket”) have ‘broken links’ on USAU. The data is there (if you go to an individual team’s Team Page and from there click the correct link to the game), but the link from the tournament schedule page won’t take you there.

   These games were MISSING from the data when I first published Career Nationals Stats. However, I copied down all the links manually, and these games are now added to the database.

2. 4 games from the 2014 Mixed division (the 11th place game and the 3-game 13th Place bracket) also had the same issue and have also now been added by hand.

3. 10 games from the 2017 Mixed tournament (the 3rd place game, plus the 9-game “Pro Flight Play-in” bracket) also had broken links from the tournament page, but in this case I could NOT find the box score from the Team Page, so they remain unadded.

   For the same reason, there are 8 missing games from the 2016 Women’s consolation bracket (The two “Pro Flight 7th place” brackets) and one missing game from the 2021 Men’s tournament (Johnny Bravo vs. Chain Lightning for 9th place). (19/1779 = about 1% of total games)

   These games are NOT counted at all in the database -- they won’t count as a “game played”, nor will their point differential be included in a team’s point differential calculation (even though there is a score listed on the tournament page)

4. 6 games (6/1779 = 0.3% of total games) are listed on the USAU website as either canceled or forfeited (I’m assuming that’s what the ‘F’ stands for). They are: the 15th place games for Men’s 2015 and 2016; 5th place for: 2016 Men’s, 2017 Women’s, and 2023 Men’s; and 11th place for 2017 Men’s. These games aren’t counted in the database (i.e. not even as a win for the team that didn’t forfeit).

5. There are 32 games (32/1779 = 1.8% of total games) in the database for which the game was played, and the final score is available, but there is simply no box score data recorded—it’s all 0s. Here’s a 2019 Quarterfinals game between Molly Brown and Schwa that is completely missing box score data. 24 of these 32 games happened in 2014 Mixed, so the issue is pretty limited outside of that tournament.

6. Putting those numbers together, about 96.8% of all games played at Nationals since stat keeping began are included in the database with reasonably representative (see the next note) stats.

7. Now that the script directly scrapes the final score, we can compare that number to the sum of goals (or assists) listed in the box score to get a feeling for how many errors there are in the data.
   
   My script finds 306 games that have a minor goal miscount (For example the 2025 Mixed semifinal, Hybrid vs. shame. has shame. scoring 12 goals but only 11 goals are accounted for in the box score. This one’s actually fixable since the game was streamed!). These errors account for 519 missing goals, an average of less than two goals per game (per “game with error”, that is. Less than one goal per game considering every single game). In other words, the vast majority of games that have an issue like this are missing just one goal. (306/1779 = 17.2% of games)

8. We can also do the same check using Assists instead of Goals. Doing it this way finds 400 games with at least one missing assist, totaling 616 missed assists (again, basically 1 per game). These are NOT necessarily the same games where a goal is ‘missing’, for example the 2025 Mixed Quarters game, Slow vs. shame. has 14 assists recorded for shame. but 15 goals recorded.

9. As for Blocks and Turnovers, obviously I have no way to error check the data against another source. If I had to guess, they’re probably slightly less accurate than the goal/assist data.

Overall I feel pretty good about how the video has been received, but I do want to take a little time to address it more in depth. I only replied to a few comments in the Reddit thread—because I’m a little self-conscious about just spraying out comments left and right on Reddit. And I don’t want it to appear that, because I didn’t reply to a particular comment, I have no good rebuttal to it. I feel less self-conscious about putting my detailed thoughts up on my own personal blog, so, here are a few further thoughts:

1. Decision making / decision making under fatigue

A number of comments pointed out that players make mistakes, especially when tired (each line is from a different comment):

“I think a key aspect here is conditioning. A tired offense will make mistakes...”

“need to take a quick breath or else they’ll throw it away immediately due to fatigue...”

“I think [slowing down is] taught at lower levels when players...do silly things before they are locked in on offense.”

“I think at lower levels, “monkey brain” tries to do too much too quickly…”

“I don’t think there is anyone arguing fast breaks are categorically “bad”, they are just bad for some younger, developmental, or lower level teams because they haven’t learned to capitalize on them effectively.”

I touched on this previously in my post Why I like fast breaks:

Teams (and individuals) aren’t good at what they don’t practice

...

The main reason frisbee players aren’t good at (or comfortable with) fast breaking is that they’ve played frisbee for years but never practiced fast breaking. Basketball players are comfortable with fast breaks because they’ve been practicing them since they were 6 years old.

Think of, say, that college senior on your team who you wouldn’t trust to pick the disc up and start a fast break. Then ask yourself — how good could they have been at fast breaks if every turnover of every practice and game since their freshman year was an opportunity to improve their skills at making decisions in fast break situations?

I don’t (necessarily) deny that players are bad at making calm decisions while tired and while going fast. But if they practiced it, all the time, they would be a lot better.

And again, remember that the overall argument I’m making is not just “people can play faster if you teach them”. That’s just one part of the overall argument that the advantages (the defense isn’t ready, the defense is just as tired as you are, the offensive opportunities are better) outweigh the disadvantages (playing offense fast is perhaps slightly harder, especially if you haven’t practiced it). You can’t just say “kids make mistakes”, because kids make mistakes playing frisbee “slowly”, too.

For one final variant on this topic, here’s another comment:

Around 6:15 is, ironically, an example of why some teams and coaches are hesitant. Starting with a fast pace exploits the advantages the video is talking about to gain yards, but then a rushed huck throws it away.1

I think we can all agree I can find ENDLESS examples of incomplete hucks, whether coming from fast break situations or when starting from a set offense. The argument was never “you’ll score every time on fast breaks”, it was “your average number of points per possession will be higher if you’re fast breaking”.

No one fast break turnover can disprove that hypothesis, nor, I admit, can any one easy score on a fast break fully prove that hypothesis (more on this at the end of this post).

I also hope that viewers will respect my intellectual honesty for including a few fast breaks that don’t end up being successful! I’m not trying to pretend fast breaks are more perfect than they really are, and I trust my thesis enough to show you the downsides that do sometimes happen.

There were some fast-break-related concepts I didn’t focus on in the video simply because I couldn’t find a good way to make them feel compelling in a video format. “The perceived downsides happen because you don’t practice fast breaks enough” was one of those concepts. So instead of an extended discussion on this concept I just turned it into one quick comment near the end of the video (at 10:34 or so):

Even the clips highlighted in this video are barely scratching the surface of what’s possible with fast breaks. Basketball players learn to fast break from the moment they start playing basketball. With more practice fast breaking, frisbee players will get better at playing fast break offense.

2. On the “value of possession”

A common topic commenters brought up is how “valuable” possession is in frisbee. One Reddit comment says possessions in ultimate are “more meaningful”. Another YouTube comment says “the value of possession in ultimate is much greater than similar sports”.

This is something that I’d seen discussed and I’d thought about before making the video. However, I didn’t feel like I had a “killer” answer—at least no answer that’s as easy-to-understand and logically-true as I’d prefer to have—so I didn’t try to rebut it in detail in the video.

I did make the general argument against it at 10:06— Fury “knows they come out ahead in the long run by fast breaking” even if they sometimes throw a quick turnover. I also brought it up in a similar way in the aforementioned Why I like fast breaks:

...a critic would say: but frisbee is different—possession is so highly valued, we can’t risk a turnover just for a better scoring chance! But I think that argument cuts both ways. If you highly value possession, then you should value even more the opportunity to play offense against an unprepared defense.

What else would I say on this topic?

I think the first question I ask myself is what does it really mean to say possession is more valuable in frisbee than other sports?

The first thing I’d say about “possession is valuable” is that, as written, it is false.

You could technically win a frisbee game 15-0 without completing a single pass. Just catch 15 Callahans2. You could win a frisbee game while losing the “time of possession game” 90% to 10%. I bet there’s a good chance this one has actually happened.

Possession is not inherently valuable. Scoring is inherently valuable. Possession is only valuable to the extent that it leads to scoring. But then if you agree with that, we’re back to the question of “how do you maximize the chance you score?” And, IMO, you do that by not squandering the great fast break opportunities that I showed teams squandering in the video.

If you don’t agree with all that, I’d be curious what “possession in frisbee is extra valuable” means to you if it’s not “scoring in frisbee is extra valuable”?

The other question here is whether strategies that are good on average are still good in situations with low sample sizes. As one commenter put it:

Comparing to other “open field constant movement” sports (soccer, basketball) there is ONE key difference...You get about 90 possessions per game in both sports...

Ultimate is played to a point target. You might get 25-30 possessions in a game to 13 or 15.

Is a strategy with a higher expected value still your best strategy when the sample size is small? I’ve seen this topic come up in other fields as well—even in something like the Pokemon video game, where you might have a choice between using a move that deals more damage to your opponent but misses sometimes (higher expected value—even accounting for the misses—but also more volatility) or a more that deals less damage but won’t miss. On average, you might expect to win more Pokemon battles using the first option, but it would really suck to be knocked out of a big tournament by losing a game one of those random time your attack misses.

I’m not sure if there’s a “theoretically correct” answer here, and I couldn’t figure out what terms to search for to find the relevant science.

It’s also not clear to me whether it’s even true that fast breaking is a “more volatile” strategy than setting up your offense. Is that even a concept with meaning in ultimate? There are no “moves that don’t miss” in ultimate—even setting up your offense is only going to work ~60% of the time. I guess it would mean something like “you’ll score more overall fast breaking but you’re also more likely to have 5 turnovers in a row”?

My intuition is to simply say: maximizing your points per possession is good no matter how many possessions you get in a game. If anyone disagrees with that (and truly thinks that even if fast breaking maximizes your points-per-possession, it’s too dangerous because of its ‘volatility’), I’d love to hear exactly what your argument is.

3. “Fast breaks have been around for a long time”

The most-upvoted comment on Reddit said simply “I thought looking for a fast break after a turn has been a strategy already for a long time.”3. A similar comment came up on YouTube:

Your title is hilarious. It’s the future-past... Back in the day, in the late 80’s - early 90’s, NYNY was famous for jumping on an opponent’s turnover and launching right into their offense.

And there was more of a similar vein in at least one other comment on Reddit:

I graduated college 15 years ago. None of this is new and it’s been standard play at literally every level of ultimate I’m aware of for over a decade now.

To be clear, I don’t think these comments came off as super negative. I don’t mind people leaving them, and I do think I could’ve been more explicit in the video about what exactly the title meant. So let me take a moment to explain what I meant by the title and why I stand by it.

3.1 “Is the future” doesn’t mean “never existed in the past”

By saying “fast breaks are the future” I didn’t intend to make it sound like fast breaks haven’t existed in frisbee previously.

When I describe current frisbee culture in the video, I said that “most” teams slow down to set up their offense (1:46 ish) and that “more and more” teams are embracing the fast break in recent years (1:00 ish). While I don’t have any hard data, I do honestly believe both of these claims are true.

Remember that (per Ultiworld’s Projected Rankings) there were 230+ College Women’s teams, 366+ College Men’s teams, 97+ Club Women’s teams, 254+ Club Mixed teams, and 277+ Club Men’s teams. I obviously haven’t seen more than a tiny fraction of those teams play, but if my experience is in any way typical, I would be surprised if more than 25% of them were any good at fast breaking. In fact, I honestly think I’d be surprised if more than 10% of them were good at fast breaking.

So yes, I do think it’s true that “most” teams set up their offense. If you disagree with this take now that I’ve put specific numbers on it, I’d be happy to hear your opinions! (And if fast breaks are really “standard at every level”, why was it so easy for me to find examples of missed fast break opportunities?)

I also don’t have any hard data on whether it’s true that “more and more” teams are fast breaking, but that’s my personal opinion watching frisbee the past few years.4

There are many trends in society that could be said to be “the future” even after they’d existed for a significant amount of time. I’ll give two examples specifically from other sports.

Playing offense on 4th down has been an option pretty much as long as the NFL existed, but a writer who wrote “Going for it on 4th down is the future of the NFL” still would’ve been right, as late as the mid-2010s:

For the record, this is not a new argument—I brought up the 4th down comparison in a 2023 post on fast breaks.

I also think an NBA writer could’ve written “3-pointers are the future of the NBA” in, say, 2005, and have been right, even though 3-pointers had already been around for a long time:

Even just between 2015 and 2025 the NBA has completely changed the amount of 3-pointers attempted—the team that shot the most threes in 2015 would have shot the next-to-fewest in 2025:

So: yes, I’m aware fast breaks have already been a thing in frisbee. No, I don’t think that makes my title incorrect or even hyperbolic. I really do believe there’s a lot of room for fast breaking to grow, at the elite level and especially at lower levels.

3.2. ...Also, look at all the other comments explaining why fast breaks don’t work

While the top comment on Reddit correctly pointed out that fast breaks have been a thing, I think when you look at all the other comments you can see why I think fast breaking is still underutilized.

The 2nd-most-upvoted comment says “A very real reason you don’t see this enough at regionals level ultimate is that it just requires better conditioning...”.

To me, this is pretty silly. Random dudes I’ve played basketball with at the Y (or rec league soccer with at the park) have good enough conditioning to fast break when the opportunity arises. High school and college basketball teams fast break even though they’re not as fit as the pros in their sport. If fast breaking was more important in frisbee culture, people would find the stamina for it.

(Also...I’m simply not sure I agree with the comment’s claim anyway. Are elite players actually less tired than their non-elite counterparts? I would personally guess that all players more-or-less get equally close to their own personal maximum fatigue levels, and the main difference between Nationals and Regionals players is that the Nationals players are doing everything at higher energy output levels.)

I already mentioned above all the comments explaining why youth or lower-level players might be bad at fast breaking:

“I think at lower levels, “monkey brain” tries to do too much too quickly and probably turns the disc more than it should, so favors settling down a bit”

“...they are just bad for some younger, developmental, or lower level teams because they haven’t learned to capitalize on them effectively.”

Again, this is not an excuse that basketball coaches would use to discourage their team from fast breaking. Fast breaks are good strategy, so coaches encourage them and use mistakes as learning opportunities.

Another type of comment I saw was about the need for team-wide commitment:

“- lack of team coordination. You need multiple players committed to it to make it work.”

But again...this is not a problem in basketball! I could go play with some random dudes at the park and we’d be on the same page about fast breaking. Which tells me fast breaking “is the present” of basketball, but it’s the future of frisbee.

So yeah...I think it’s funny that some commenters are saying ‘fast breaks aren’t the future, they’re standard already’, while at the same exact time multiple other comments are offering various reasons why frisbee players aren’t (or even shouldn’t be) fast breaking.

If I made this video about basketball, EVERY comment would be telling me how fast breaks are already standard and my video is nothing new. The fact that didn’t happen for a video made about frisbee tells me that my title was not wrong.

But of course, I don’t deny that there are countless other smart players out there who know how to take advantage of a fast break! I’m just saying there’s still room for a lot more of them.

3.3 I simply disagree with some versions of this take

I don’t have any disagreement with the comments saying “fast breaks are already a thing”—that’s absolutely true—but I do think the comment that said they’re “standard at every level” is just objectively false. Basically every level I’ve ever played or coached at I’ve had people advocating slowing down, and at some of those levels (semi-competitive summer league with many Sectionals and Regionals level club players; coaching college women’s division) basically everyone played slow.

4. Getting better proof

JoeMama writes (paraphrasing for clarify):

I’d love to see stats on effectiveness. How often does a possession end in a score vs. a turnover, depending on whether you fast break or setup slowly after a turnover? Could separate out O and D lines too.

I’d love to have data like this. At the end of the video I say “The data says fast breaks are the smart way to play in every other sport, and I believe that’s clearly the case in Ultimate, too.” I specifically used that “I believe” because I recognized that I didn’t have actual points-per-possession data for frisbee the way I showed for basketball (at 2:13).

The only problem is it wouldn’t be super easy to generate this data. Especially across an entire division. I could do something like “all streamed Fury games in 2026” in a reasonable amount of time. While that avoids some issues (you’re not comparing good team that fast breaks to less good team that doesn’t fast break), I’m sure other people would have complaints about how the Fury experience shouldn’t translate to other teams that aren’t Fury. So I’ll keep this project in mind but it might be hard to come up with a methodology that satisfies all the doubters.

Thanks again to everyone who watched and/or commented on the video! Overall I’m quite happy with how it turned out and quite happy with how it’s been received. I hope this article has done a good job giving a more nuanced background on how I think about some of the topics in the video. There’s only so much you can fit in a 14-minute video…which is one of the reasons I like writing blog posts in the first place!

If there’s any other feedback on my video that you’d like me to give my thoughts on, please let me know!

A new video I made in collaboration with Hive Ultimate is now out on YouTube.

Click here to watch: Why Fast Breaks are the Future of Ultimate

A big thank you to Felix for his work making this video happen. I know basically nothing about video editing, so from my perspective he did pretty much all of the hard stuff.

As always, feedback is very welcome, whether here or on YouTube. What Some Flow post should become a video next?

Enjoy! More blog posts coming soon.

I’ve downloaded the box scores for every Nationals game since USAU started keeping box scores in 2014—11 years, 3 divisions, 1,779 (ish)1 games. By putting these together in a database, we can explore the greatest games, greatest years, and greatest careers at Nationals2.

I’ve also included the full-tournament stat sheets from the 2013 tournament (available here, per-game box scores are not available as far as I know) along with the 2013 results (taken from here) to add 2013 to the full-year and full-career analyses.

I’m not aware of a project like this that already exists. At least, I’ve never seen it myself, nor have I seen Ultiworld use career stats in their ‘best player of the decade’ articles or the Ultimate Hall of Fame in their messaging.

Click here to access the updated full data via Google Sheets. [Note: Click “Data”→ “+ Create Filter View” and you’ll be able to filter the data the way I’ve done in tables below.]

The data definitely isn’t perfect but in my opinion it’s still worth sharing—some further notes and known errors are at the end of the post. Feel free to get in touch if you find an error in the data or just find something cool you think I should add to the post.

[Update (2026-04-22)]:

A bunch of minor updates have been made to this post. For the details, see Update on Career Nationals Stats.

[Update (2026-04-23)]:

I manually changed two more names in the 2013 stats file to integrate with players’ current USAU names (other changed names listed in the notes):

Nicky Spiva -> Nick Spiva

Georgia Bosscher->Georgia Shriwise

Updated three tables (Career assists, career +/-, list of players to play all 12 Nationals) with the new updates to Nick Spiva’s stats.

[Update (2026-02-23)]: [I’ve removed the details of this update as they’re now fully integrated into the post below, and described in more detail in the update post.]

[Update (2026-05-13)]: I embarrassingly didn’t notice until long after originally releasing this article that Turnovers and Blocks weren’t tracked in 2014. Some of the single-game and single-year +/- totals from 2014 are therefore suspect (turnover aren’t being subtracted). Players that played in 2014 likewise have their career blocks undercounted.

Also: this post was covered on the Ultiworld Deep Look podcast here, and there was some discussion here when I posted it to Reddit.

With that said, let’s go:

Single game standouts

Here are some of the standout single game performances in Nationals history:

• Kelsey Viars of 2014 Tabby Rosa stands alone as the only player to record 8 goals in a single game.

• Seven other players have scored 7 goals in a single game, 5 in the Men’s division and 2 in Women’s (never happened in Mixed—is there a reason for that?).

• A 6-goal game has happened 28 times, twice by Sandy Jorgensen and once by Kaitlyn Weaver (who’s already on the list above as a 7-goal scorer).

• Kurt Gibson’s 7-goal game happened in 2014 semifinals. I believe every other game on the list happened in Prequarters or earlier.

• Jimmy Mickle of 2016 Johnny Bravo stands alone as the only player to record 10 assists in a game.

• Robin Meyers (2017 Wild Card) hit 9 assists in a game. 11 others have recorded 8 assists in a game, including Valeria Cardenas at the most recent Nationals.

• I believe none of these games happened any later than Prequarters.

• I have to express some doubts here…I really want the Florence Dionne 10-block game to be real…But it’s hard to believe it really happened. No one else has ever recorded more than 6 blocks in a game. But she’s got this 10-block game in 2025 Nationals, another 6-block game in the same Nationals, and 0 blocks in any of her other Nationals games that year. (And only 2 blocks in 2024 Nationals).
  My heart wants to believe…but my mind is doubtful. Pretty much every elite team is filming all their games nowadays. Someone must be able to find out whether this statline truly happened or not.

• Outside the top spot, 7 players are tied with 6 blocks in a game. There are 18 instances of 5-block games (Twice by Opi Payne).

• It appears Conor Belfield is the only man-matching player to ever get 6 blocks in a game. Perhaps unsurprisingly, he was the mixed division DPOY that year. (In line with the note I made at the start of this post, the DPOY article doesn’t even mention that he had one of the best single-game performances of all time!)

• Not shown in the image: shout out to DiG’s Gus Haflin, the only player in the Men’s division to have a recorded 5-block game (or more). Every other 5- or 6-block game happened in Women’s or Mixed.

• Above are all the instances of a double-digit turnover game. (Updated on 2026-04-20 to add 2015 Jenny Fey vs. Molly Brown, see updates post.)

• The vast majority on this list are from 2016 (Rockford, IL), with a few from 2015 (Frisco, TX) or 2017 (Sarasota, FL). You think it might’ve been windier at those tournaments than it is near San Diego?? There are some 9-turnover games in the post-COVID era, though.

• No shade meant towards any of these players. Sometimes a team’s rational best shot at winning is to have your best thrower take a bunch of risky shots (plus the already mentioned wind factor).

• Mischa Freystaetter stands alone with a +13 game at 2017 Nationals. The image includes all the double-digit plus/minus games.

• Florence Dionne’s game is discussed in Blocks section above.

• Daan De Marree shows up twice on this list, with two +10 games from 2025 Nationals alone! We’ll see more from him later.

• I believe the only game on this list that happened in semis or later was Levke Walczak’s 2022 game against Fury.

• Not shown: 13 players have put up a +9 game, including repeat appearances on the list from Mischa, Daan, Levke, and Sandy Jorgensen.

• Levke’s 2025 Finals performance of 5 goals / 4 assists / 4 blocks / 5 turns leaves her next on the list at +8, the highest ever in a finals game. (If there was a stat just called “plus” that didn’t subtract turnovers, this would be one of the three games ever at +13, along with Mischa’s and Kurt Gibson’s games in the image above.)

  ---

Single year standouts

Highlights of players who had a great year at Nationals:

• Sandy Jorgenson (2016 Scandal) stands a head above the rest with 27 goals. She, Kristen Reed (Love Tractor/Molly Brown) and Sean Ham3 (Mischief/Lawless) are the only players to score more than 20 goals at Nationals multiple times.

• This image is every Nationals performance of more than 20 goals. Not shown: there are 14 more cases of an exactly 20-goal nationals, including twice by Sarah Norton of Traffic.

• (Update on 2026-04-20: Sarah Itoh’s 21-goal 2015 performance with Scandal was added to this list with the new data included, see update post.)

• Ryan Osgar (2019 Sub Zero) is the single-year assist king, with 32 in 2019. Two other players have also crossed the 30 mark. Osgar also did it while tying for the fewest turnovers on this list (6 turns). 2022 was also the year he went 95% on 21 hucks in the UFA, so he was absolutely on one that year, throwing a frisbee in a way perhaps no one else ever has.

• The image shows all instances of more than 25 assists at Nationals. Only two players have done it twice: Robyn Fennig (Heist, NOISE) and Travis Dunn (Lawless). We’ll see more from them below.

• (Updated 2026-04-20: Cassandra Swafford’s 26-assist 2015 performance now included.)

• Ashlee Davison (2016 Traffic) tops the blocks list with 20 blocks in a single Nationals.

• Sandy Jorgensen got 15+ blocks at Nationals twice. I believe the next closest person to that record is Nathan Champoux, who’s done 11+ blocks twice (12 in 2024, see above, and 11 blocks in 2022). Kami Groom has done 10+ blocks twice (with 10 in 2023).

• Honestly this list has a higher % of people I’ve never heard of than any other list we’ve seen so far.

• Jordan Sorensen shows up for the 2nd time (see single game +/- list). She’s really exploded onto the scene in the past couple years.

• (Updated 2026-04-20: Now including Kami Groom’s 15-block 2015 tournament)

Sorting by Blocks also makes it easy to pick out players who’ve put together what you might call a “triple double”—10 or more goals, assists, and blocks at a single Nationals. I believe these are the only three players to ever record a Nationals triple-double:

• It’s somehow fitting that each of these three players had the most of each triple double stat (Goals, Assists, Blocks).

• Kami Groom’s triple double wasn’t included in the original version of this post, but we now have the data—see updates post.

• A few players have come tantalizingly close to the triple double. Here are all the times a player was one stat point away. This table is picked out by hand so it’s possible I missed someone:

  

• Ben Jagt is the only Men’s division player to come within one stat of a triple double.

• As mentioned above, 2015/2016/2017 was a very different environment from the more recent Nationals.

• Yina Cartagena (2023 BENT, 29 turns) is the only other player anywhere near the top of the list from within the past few years (2022 onwards).

• It’s pretty hard to argue against the claim that Daan De Marree’s 2025 tournament was the most complete frisbee we’ve ever seen. A +/- of 48 on zero turnovers—in over 14,000 player-years, we’ve never seen another +/- over 43. No one else with 20 goals has ever put up more than 15 assists in the same tournament—he had 24!

• The list above has all the years of +/- over 30. It’s full of names we’ve already talked about: Daan, Osgar, Sandy, Kurt Gibson, Travis Dunn, Sarah Jacobsohn, Jordan Sorensen, etc.

• Kaela Helton and Sandy Jorgensen are the only players to put up a +/- above 30 multiple times.

• (Update 2026-04-20: Kami Groom’s 2015 now included on this list after being missing from the original post)

I wanted to include dominant single year teams here, ranked by point differential at Nationals, but for reasons discussed in the notes below, I believe my data currently isn’t good enough to actually give an accurate list for that. I’m pretty sure that the most dominant team ever in terms of point differential is 2021 Fury at +53. No team in pool play scored more than 4 goals on them.

The original version of this post didn’t include a section for best team point differential because of the data issues (see update post). I now feel reasonably confident that the data is at least pretty good, so here is the best single-year teams in terms of point differential:

• The most dominant team ever in terms of point differential is 2021 Fury at +53. No team in pool play scored more than 4 goals on them.

• 2017 Revolver is the most dominant men’s team, while 2016 Slow White is the most dominant mixed division team. (I skipped a few other teams to include Slow White on the list, hence why the bottom of the table looks the way it does.)

• 2013 Scandal is the data point I’m least confident about b/c game scores were taken from a separate source, see the explanation at the very top of this post.

• 2025 Fury at +44 has the highest point differential while losing a game—they lost 15-12 to Brute Squad in the final.

Career standouts — totals

Some of the best careers at Nationals:

[The inclusion of more 2015 women’s data means many small changes to these tables compared to when this article was first posted]

• Kami Groom is on top for goals, and it’s not even close.

• Catherine Menzies and Peter Prial are the only other two players with over 100 career goals. (Sandy Jorgensen surely does too if we had been able to include years before 2013.)

• They’re also the only players with a career “double-triple”—triple digits in two categories (goals and assists). Tannor Johnson seems like a lock to reach that milestone as well if he keeps playing. Kami Groom can as well by keeping up her average assist production for another 1-2 years.

• Not shown but worth noting: Ella Juengst is at 15th on the list with 79 goals. It may seem not that impressive when you see she’s played at Nationals 8 times…but that’s only if you don’t know she was still playing for UNC in college in 2023!
  If she stays healthy and stays interested in frisbee, she’s basically a lock to end up at the top of this list (with Tannor Johnson). She had 4 goals combined at her first two Nationals4, but has averaged 12.5 per year since then (and, again, she theoretically hasn’t even reached her prime yet).

• The inclusion of more 2015 women’s data added goals for Groom, Menzies, Jorgensen, and Stern (the latter two also jumped up a few places in the ranking).

• It’s interesting to see how Catherine Menzies has played 10 more Nationals games than Kami Groom over the years (due to Menzies’s teams ending up in consolation brackets while Groom’s teams almost always make deep bracket runs. That’s more than one tournament’s worth of games, it’s like she had an extra year to rack up stats. (In fact, Menzies and her teammate Jessica Rockliff are tied for the most games played in the database.)

• Robyn Fennig tops the list for assists, and again it’s not especially close. With 2015 women’s data fixed, she’s now the only player with over 200 career assists.

• Tannor Johnson is quickly climbing up the career assists leaderboard as well. Ulti-verse has him still on the UMass roster in 2020, so he’s likely the youngest person on this list by a healthy amount. If he stays healthy, he’s almost certainly on track to someday have the most total Nationals stats of any player.

• The career assists leaderboard skews heavily towards men-matching players — only 5 of the top 18 career assist-getters have played in the Women’s division.

• The table includes all 18 players with 100 or more career assists.

• Kami Groom tops the career blocks list as well, again by a somewhat healthy margin. One of the biggest things I’ve learned during this project is just how impressive Kami Groom’s career has been. I knew she was really good, but I didn’t know she stood quite this far above the rest, stats-wise.

• As we saw in the single-game stats, high block totals tends heavily towards woman-matching players—10 of the top 12 in career blocks have played in the Women’s division. Ryan Drost, at 10th, is the highest man-matching player on the list.

• Nathan Champoux and Dena Elimelech stand out for high production in a lower number of total years. Elimelech is less than half a block per game behind Kami’s career rate (6.4 bpg vs 6.1 bpg).

• Robyn Fennig has more turnovers than any other player by far.

• There are (unsurprisingly) a number of very good players on the career turnovers leader list. It’s your classic “you can’t lead the NFL in career interceptions without being a very impressive quarterback” effect. (Well, at least to some extent—here’s the NFL list)

• Every player over 100 career turnovers is included in the current version of this table.

  ---

  

• Unsurprisingly Kami Groom tops the career +/- list by a wide margin.

• Kaela Helton shows up at #2, with a healthy gap separating her from #3. She’s 7th in assists, 5th in blocks and, at 17th, just outside the image I clipped for career goals.

• Shout out to Kurt Gibson, all the way at up 9th on the list in only 7 years played. (Claire Trop & Dena Elimelech—tied at 27th—are the next closest players with 7 or less years played)

• More than any other table in this post, I think the career +/- list is truly a “who’s who” of famous frisbee players. Even the next few players on the list after 15th follow the trend—Khalif El Salaam, Jack Williams, Rowan McDonnell, Matthew Redher, Josh Markette, etc.

Before we look at some more counting stats let’s give a quick shout out to the players who’ve been at every nationals in the database (2013 on):

• 26 players have played at every Nationals from 2013-2025. Shoutout to Jeff Wodatch and his 8 blocks in 12 years, the only instance on this list of less than double-digit totals in any of goals/assists/blocks/turns.

• A previous version of this image was missing Nick Spiva b/c his 2013 data had his name as “Nicky”, he’s now added to the list.

• Another 25 players have played 11 years in the last 12.

Career wins/point diff/etc

With the data quality updates I made in April 2026, I now feel good enough about the win/point differential data to share some tables, although I expect there are still some small inaccuracies.

Unsurprisingly, the top of these lists is populated by players who have spent years on Fury and Brute Squad (though, to be clear, many of them have spent time on other teams as well).

Point differential:

• At the top are the 3 players who’ve been on Fury every year since 2013: Anna Nazarov, Carolyn Finney, and Maggie Ruden. Their career point differential is something like +425 (about +35 per year).

• The highest ranking player I have that never played for Fury or Brute Squad—at 24th on the list—is Simon Higgins of Revolver (and Polar Bears) at +250. A few spots down at 30th is the next one—Kelly Johnson of Seattle Riot/Mixtape5 at +243.

• Thomas Li and Alexa Kirkland (both of BFG & Polar Bears) are the highest ranking players I could find who had ONLY played mixed.

Wins

• For wins, I see basically the same pattern, although with a bit less of a Fury/Brute Squad bias. The same 3 Fury players top the list at 63 wins.

• Simon Higgins is at 8th with 60 wins, and we see a few more non-Fury non-Brute Squadders like Jimmy Mickle with 56 wins (11th) and Jack Hatchett at 53 wins (T-14th).

Win %

I filtered win percentage by players who’ve played a minimum of 3 years. That said, there are 5 players in the data who’ve played exactly 2 years of Nationals without ever losing a game—two from 2024/25 Hybrid (Maketa Mattimore and Calliope Cutchins), two from 2013/14 Scandal (Amy Hudson and Quinne Farenwald), plus Devon Anderson from 2013 & 2015 Revolver.

• The general trends are the same, though without such a strong Fury/Brute Squad bias—Alicia White (Scandal, Molly Brown), Lauren Sadler (Scandal, Riot, Brute) and Opi Payne (Scandal, Fury, Molly Brown) are all in the top 5.

• Evan Boucher (Revolver, Mischief), Aaron Bartlett (Truck Stop, Hybrid), and Ashlin Joye (Revolver) are the highest ranked players to have played in the Men’s division.

Career per-year averages

For all these stats I’ve used a minimum of 3 years played as it seemed like a reasonable cutoff of “actually has a Nationals career”.

• Two names we’ve seen before—Kristen Reed and Sean Ham—are tied atop the goals per year leaderboard, both with exactly 59 goals in 4 years.

• Christian Boxley has played more years than anyone else above him on the list—maintaining elite production for 7 years and counting. The next name we can say that about is Kami Groom (20th on the list). Obviously once we get past Kami there’s no one else we’ll be able to say that about (Although Jade McLaughlin is barely 0.1 goals per year behind Kami’s pace). These players are basically the Pareto optimum for goal-scoring longevity—no one else has been that productive for that long.

• I had no idea how productive Simon Carapella has been the past three years, I feel like his performance has been under the radar.

• Yina Cartagena tops the career assist/year leaderboards. Given she didn’t start playing in the US until later in her career6 she’s unlikely to ever approach the top of the career total leaderboard, but her skills are undeniable.

• Chris Mazur and Travis Dunn are the only other players above 20 assists per year. Travis is the only player with >20 assists/year and >3 years played (though Yina may reach that too). He put up 5 assists with Condors in 2019 but has never been below 22 in any year he attended Nationals since then.

• Daan De Marree will likely join this list once he plays one more Nationals. Tayara Romero, Elliot Hawkins, and Tobias Brooks likewise have a chance to enter the list with more years played.

• Career blocks per year looks pretty similar to the career total blocks leaderboard.

• This chart really highlights Nathan Champoux’s current 5-year run (a full 1.5 blocks/year ahead of the next closest player!) and Kami Groom’s entire career (Sarah Meckstroth is the next closest player with 10+ years played).

• Kami Groom and Sandy Jorgensen both moved up a few spots on this list compared to the original version I posted, now that more 2015 data is included.

• Turnovers per year — don’t really see anything new to point out here, but included it for completeness

• I believe this list shares 4 players (Fennig, Cartagena, Bajwa, Bennett) with the career assists per year leaderboard.

• +/- skews pretty heavily towards male-matching players, with only Kami Groom and Levke Walczak in the top 10 on the woman-matching side.

• Again, I didn’t realize how much Simon Carapella has filled up the stat-sheet in the past 3 years. Peter Boerth, also of DiG, has also been filling up the stat sheet. Those two DiG players have a combined average of only 2 turnovers per year (1.33 for Simon, 0.67 for Peter).

• Other players on this list, like Travis Dunn, Levke, and Tannor Johnson, rack up more positive stats but also many more turnovers.

Career stats — various

Here’s a few more stats I calculated for fun, using goal/assist/turnover ratios to highlight some of the more obvious player “prototypes” (again with 3 year minimum):

• Above is the players with the highest goal-per-assist ratio. You could call them “bucket getters”. There are a few other players who might be on this list but they never threw a single assist (and I can’t exactly calculate a ratio if the denominator is zero). Three players in Nationals history have 20+ goals and no assists.

• Not shown: the first player on this list with >5 assists is Jeff Babbitt (35th place), with 56 goals and 7 assists (8.0 ratio). The next player after that with >5 assists is Ella Juengst (74th place), with 79 goals and 13 assists (6.1 ratio). Let’s say those two plus Kristen Reed are the standard for high output bucket-getting.

• Above are the leaders in “assist per goal” (minimum 1 goal). I’m not sure what to call this group— “pocket passers” maybe? Lia Schwartz is a young star who I wouldn’t be surprised to see fall off this list as she gets older.

• Shoutout to Matt West of 2013-2017 Mischief, who put up 34 assists without ever scoring a goal. I believe no one else with 0 goals in their Nationals career has more than 18 assists.

• Not shown: some top players on this list with a high assist output: Yina Cartagena with 72 assists and 7 goals (10.3 ratio). Further down, Simon Montague ( 133 / 25 = 5.3 ratio) and Nick Spiva ( 136 / 27 = 5.0 ratio) are the first players on the list >100 assists.

  ---

• Shoutout to Jana Krutsinger, the only player with women’s division experience on the list.

• Assists per turnover —it’s clearly very hard to get on this list while also playing more than the minimum 3 years (shoutout to Dillon Whited for being the only one!). Because of that, I went through the list manually and picked out the first few players who also had more than, like, 60 assists. That list is:

Since that list is all man-matching players, I went back and generated a table of the highest A/T ratio among the top 25 or so career assist leaders for woman-matching players:

• Analysis: my estimation of Kaela Helton just went up again.

That’s basically all I’ve got for you…read on if you really want to know all the issues and technical details:

General notes on data and code

This data is absolutely not perfect. But...it’s the best we have, and I think it’s at least good enough to be worth putting together this project.

There are a number of known issues (see below). Newer years have better data than older years. Some but not all of those issues can be fixed by USAU fixing broken links on a few tournament schedule pages.

Basically, you should image every table and every sentence having an asterisk at the end of it—*at least, as far as our imperfect data says. Having mentioned that here, I’ve tried not to talk too much about data quality in the rest of the post. There’s a detailed list of known issues below.

Also: full disclosure, I used AI to write the code that scrapes the data from USAU. In short, my feelings on AI are: I don’t hate AI as much as some people on the internet do. I hate it as much as everyone else does when AI is used as an excuse to not think for yourself. But I think AI’s pretty cool when we use it to do more than we would’ve been able to do otherwise. And while I know some coding, I definitely don’t have the coding skills to have done this on my own.

I’ve taken the time to double-check the data and I pretty strongly believe there is no “hallucinated” data. I double-checked a lot of individual games/years in the process of writing up the text portion of this article and never found a single issue. This isn’t really the type of situation where AI is liable to hallucinate anyway. The issues with the data collection described below are issues of how I defined the program to work, not issues caused by the AI tool I used to generate the code.

I’m happy to share my code7 with anyone who’s interested, but I have to warn you, while I know enough about coding to finish this project, I’m not exactly an expert at either coding or database management and someone who knows what they’re doing would probably find my solutions ‘ugly’.

Known errors / Potential improvements / Other notes

[Update (2026-04-22)]: Some of these notes below are now out of date, see Update on Career Nationals Stats for more info. I haven’t necessarily updated all the relevant bullet points below.

• Partial data from the 2011 tournament is available here. I chose not to include it because:

  • It doesn’t include blocks or turnovers. And,

  • The way the data is presented would have meant I needed to spend a lot of extra time ripping it.

• The USAU archive’s page for 2012 has a note that says: “Individual statistics (not yet available)”. I’ve emailed USAU to try to find out if this data exists on a hard drive somewhere.

• The ulti.info link with 2013 results (here) is clearly not error-free. I noticed at least 1 game that has different scores posted depending on which team’s results are being viewed. So far I haven’t been able to find the firsthand source for these tournament results (the results links here are broken).

• Some games have broken links that mess with my script’s ability to download them. For example, go to the 2014 Mixed tournament here then click on the “Placement Bracket (Elite Flight)” tab and click on the link that loads the game report for the game between Polar Bears and 7 Figures. It’ll take you to this blank game report, so although the box score actually exists if you access it through each team’s results page, it’s not currently included in the data. The 2015 semis game between Brute Squad and Molly Brown is another example. Based on a quick check, these broken links seem to go as late as 2017 but not after that.

• My script calculates the final score (and thus the winner of the game & point differential) by adding up the number of goals in each box score. However, some games as late as 2015 if not later (here’s an example) don’t have the same number of goals in the box score as the final score. This means point differentials and possibly win/loss records aren’t as accurate as they could be.

• There are other games (this one for example) that simply don’t have box score stats recorded. There are examples of this as late as 2019.
  

  Name related issues:

• Name changes: Because all of the data is downloaded at one time, I believe most potential issues for names that changed (due to marriage, nickname use, etc) are avoided. As long as the player is still using the same USAU account, their name is now shown as their current name on the roster, even for old tournaments. I have not tried too hard to see if there are any issues of this sort.

• Name changes, 2013 edition: Because the 2013 data was acquired differently, specifically for the year 2013 there is the potential for names not matching up. I haven’t tried too hard to fix these issues, however I did manually change a few names that came up in the natural course of the project, as below:

  • Robyn Wiseman -> Robyn Fennig

  • Kaela Jorgenson -> Kaela Helton

  • Catherine Hui -> Catherine Menzies

  • Jessi Jones -> Jessica Jones

  • Kami Groom -> Kameryn Groom

  • Sandy Jorgensen -> Sandra Jorgensen

  • Nicky Spiva -> Nick Spiva (added 2026-04-23)

  • Georgia Bosscher->Georgia Shriwise (added 2026-04-23)

• Duplicated names: There is a potential for duplicated names (multiple people with the same name). If people had duplicate names in the same year, this is easy to see in the data, because they’ll have an impossible number of games played. I believe there are only 3 examples of this: 2016 Jordan Taylor (Ironside and DiG), 2017 Jordan Taylor (Ironside and Mischief), and 2019 Rachel Johnson (Pop and Superlame). I haven’t bothered to separate these into different players b/c none of them had enough stats to influence any of the images included in this article.

• Duplicated names, part 2: If two people had the same name but played at Nationals in different years, that’s a bit more challenging to detect, so there may be some instances of that which I haven’t caught. I don’t believe there’s any examples of this in any of the images in this article, however.

• “Unknown”—some stats in the 2013 stat sheet are assigned to “unknown”. I just left them in the data as-is, since it didn’t end up inconveniencing me in any way.
  

• Potential improvement: Ultiworld has detailed stats (yardage, points played, etc) for the championship games that I could theoretically integrate with this data. However since it’s only for the championship game, it’s so limited at this point that it’s not really ever going to tell us anything besides who made finals.

• Potential improvement: I originally wanted to track whether games happened in pool play, championship bracket, or consolation. However, that’s not consistently shown on the results page. For example this quarters game from 2014 has “Quarterfinals” in the match report header, while this quarters game from 2017 does not. I’ve emailed USAU to ask if they’d be willing to go back and add this data in a consistent manner. I could theoretically re-write my scripts to get this information from the tournament schedule page, but that would require a lot of work and I’m not confident I have the coding skills to do it that way.

For example, here’s a video of Hall of Fame coach John Calipari teaching a defensive transition drill. Note how he points out that “you’re guarding whoever you’re guarding” on defense. The drill is designed, in fact, to make it impossible for the defense to know in advance where it’ll have to concentrate its efforts1. Here’s another video with a screenshot that matches how I learned to play transition defense:

Frisbee is full of mini-transitions

But I think there’s an important difference between ultimate and basketball that actually makes this even more important in ultimate. Because a basketball court is so small, the ball will basically never be 10 yards or more downfield (downcourt) of you, except during a fast break.

In frisbee on the other hand, the offense will often complete passes (deep in cuts, short hucks, basically any sequence of 2-3 passes in a row that gain yards) that leave defenders 10 or more yards behind the play.

The typical ultimate player’s response here is to jog downfield at whatever speed the offensive player they’re guarding is jogging downfield. This is poor defense. You should be trying to accelerate away from the player you’re guarding, in order to get downfield and clog up passing lanes while your assignment is still in a position where the thrower wouldn’t want to throw to them.

It’s good offense for the offensive players to jog downfield. The offense is happy for the 2-4 players still involved in the play to have all that space to work with. But on the defensive side, you should be denying them those same advantages.

There’s a more general point here, too. I think frisbee defenses still generally underrate the yardage game.

As I said in my defensive philosophies post:

...even in high-level frisbee, the simplest example of what you should and shouldn’t give up is still one of the best strategies: make it hard for the offense to get close to the end zone. Even if that means you make it easier for the offense to go backwards.

There’s quite literally no point in stopping the offense from going backwards if you can’t stop them from going forwards at a high enough rate. An offense can literally never score on you by throwing backwards. (With the possible exception of, like, this throw.)2

This is one thing that zone defenses in frisbee get right—when you say “we’re playing zone”, suddenly many frisbee players manage to understand that being 10+ yards in the backfield is a bad thing that you should hustle to fix. But as Thinking Basketball and Shane Battier have pointed out, good man defense includes zone defense principles, and vice versa. Getting downfield of the disc is a ‘zone principle’ that most frisbee man defenses still haven’t fully incorporated.

So let’s look at a few examples where players do or don’t work hard to get downfield of the disc.

Finney

Here’s Carolyn Finney from the 2024 US Women’s US National final:

Finney is marking the thrower, but when the throw goes up she immediately accelerates away from her assignment and rushes to get downfield of the disc:

The turnover that results is seemingly due to an illegal double team by Finney that the next thrower didn’t see before releasing the disc. I support playing by the rules, so I’m not saying you should do everything exactly like Finney did here, but I think the effort to get downfield far before her assignment did is exemplary.

To put it another way, which of these wide open options would Fury, as a team, rather give up? I think the answer should be pretty obvious:

A post-turnover transition

Now here’s an example of what not to do, from Fury’s previous game in the 2024 semifinals.

On the last point of the game, a Molly Brown drop on a swing pass means that Molly is on defense with multiple players far behind the disc. This is a pure transition situation, just like the basketball videos we saw at the beginning of the article.

I’m not very impressed by any of the deepest Molly Brown defenders on this play, who all could’ve hustled harder to get back on defense. But I especially want to draw your attention to one player who watched the turnover happen:

and then, while about 20 yards behind the disc, turned their back to the disc to focus on “their assignment”:

Look, could they have actually made an impact on this play? Maybe not. I might even say probably not. But it’s the last point of your season—it’s the last pass of your season. You need to be selling out to stop the offense from making progress. Whatever last ditch effort you can give to try to make an impact on the play is worth it to make your season last a couple passes longer. Sure, your person didn’t catch the next pass...but now your season is over, is that really what you wanted?

GONA

There’s been some discussion in the English-speaking ultimate world lately about Japanese team GONA, after Hive Ultimate made some content about them. For example, see this recent Reddit thread. They run an “attack from the back” offense where multiple players often start behind the disc:

I like what GONA is doing, but I also think poor defense is making it look better than it should.

The anonymous YouTube commenter gets it right on this one: if they want to be behind the disc...let them! It shouldn’t matter if you’ve never seen this exact offense—”I’m going to sag off them unless they’re actually making themselves a threat to help the offense make progress” should be a core defensive principle you apply no matter the offense’s formation.

A superdump

This play happened in Toronto Rush’s first game of the 2025 season. DC is stuck on the sideline late in the game, and instead of staying tight to the far side handler, Rush’s Luc Comire (#15, near the middle of the screen 6-12 seconds into the video) sags downfield and back towards the middle of the field:

DC completes the “superdump”, losing 20-25 yards while swinging the disc across the field. But I wanted to include this clip because it’s a memorable pass and because this, in my opinion, is a win for the defense. Not every instance of you getting downfield of the disc is going to end with a poach block. When you can force the offense to go backwards where they otherwise would’ve gone forwards, you’ve made defensive success slightly more likely.

Japan 2016

A recent Hive Ultimate video featured another Finney-like example of a player actually hustling to get downfield of the disc:

I love this one for the same reasons I love the Carolyn Finney example above. Check out the visible effort he’s putting into not just staying with his assignment, but accelerating away from him. He makes a quick peek back at the player he’s guarding, but his main focus is rightly on the bigger threats—the yardage gaining threats.

As in the Finney example, we can ask ourselves the question of which of these passes would we rather give up:

The answer to that question is clear to me. Do you play frisbee like it’s clear to you?

A couple ‘mini-transitions’

In my experience, it’s basically trivial to find examples of players not working hard to get downfield of the disc after medium-sized gains. Here’s one I called out in Revolver vs. Machine, although I don’t think it’s the most compelling example:

After the pass is completed to Mac Hecht (i.e. the pass before the score), watch how the Machine defenders who are deepest behind the disc are not really trying to separate from their man, and are instead happy to jog alongside them maintaining a constant buffer3.

Would it have affected the play in this case? Maybe not...again I could even say probably not. Hecht throws the scoring pass pretty quickly. But look at how open the downfield throwing lanes are because there are 3 Machine defenders in the backfield. But I might at least say the defender who gets scored on might not have gotten beaten to the open side so easily if he could trust there would be help sagging into the break side.

Shout out to a commenter, Sam, who pointed out on one of my older posts that Ultiworld has highlighted Revolver having players intentionally lag behind the disc as early as 2013:

Once again, the rest of the cutters stay out of the way and allow that clearing cutter to set himself up. Most of Revolver’s cutters have found themselves behind the disc, and they are in no rush to change that. Revolver currently have a 2v2 matchup downfield, and bringing three more cutters into the mix would only clog up space.

And that as early as 2018, PoNY knew the way to counter it (paywalled):

An under-appreciated key to Revolver’s successful isolation schemes over the years has been their ability to pull defenders out of the active space by having their cleared cutters and handlers lag behind the play. Teams having been playing into their hands for years by continuing to guard these players tightly....

PoNY simply refused to cooperate...every time a Revolver player...tried to lag even with or behind the disc, the New York defenders punished it by recognizing the ploy and then moving downfield to shrink the isolation space.

So, there’s nothing really new in my article today—these tactics are known and used. They just haven’t fully penetrated frisbee culture, even at the highest levels, as we’ve seen in some of these examples. I’m not claiming to be inventing a tactic no one has seen before, I’m simply raising awareness.

Here’s another mini-transition example I spotted skimming through PoNY vs. Machine in the 2025 semifinals:

Again we see what we saw in the previous example: as the disc moves downfield, the defenders behind the play stay with their assignments, instead of attempting to accelerate away from them. We don’t see them show up onscreen again until their assignments are also back in the play.

Where Team Japan or Carolyn Finney were getting in the way of the downfield passing lanes, here the passing lane is wide open for the continuation to Marques Brownlee:

Final thoughts

Care about getting downfield of the disc. It’s kind of that simple.

I’ve mixed together a couple related topics here, mainly:

• The general principle of caring about getting downfield of the disc (whether in a stopped disc like the GONA clip, right after a turnover like the Molly Brown clip, a sideline trap like the Rush clip, or in mini-transitions like the other clips)

• The specific situation of trying to win the “mini-transition” opportunities that come every time the offense gets a 10+ yard gain (like the Finney/Team Japan/Machine examples)

While I think the general principle is important, I actually started writing this article because of the “mini-transitions”. I’d love to see more players and teams consistently hustle downfield when the disc moves downfield. The goal is not to stay even with your assignment, the goal is to beat your assignment downfield so you can clog up the passing lanes, even if just for a second or two. Smart players and teams have been doing this for years, but it’s still nowhere near fully absorbed into our culture. It’s not sexy like a layout block, but it’s a hidden key to great team defense.

I myself have written a post highlighting missed help D opportunities as well, so I certainly don’t disagree with the basic point that “there are lots of missed opportunities to play help D”.

But the article doesn’t really dig down and analyze on a detailed level why these missed opportunities happened. And one of them in particular highlights a concept I’d wanted to write more about.

Here’s the very first example from the article, from Brute Squad v. Scandal at 2025 nationals. I’ve included a few extra seconds at the beginning that help highlight what happened during this play:

Here’s the thing about this play: I predicted it. They say “hindsight is 20/20” to point out how you shouldn’t always trust someone who has opinions on what went wrong after the fact. But I’m not going to give you my analysis of this play after watching it. I’m going to tell you what I wrote about the mistakes made on this play, months before the play even happened.

In a post (from June 2025) on my philosophies on defense, I wrote this:

Footwork is so important to good defense. But honestly I think about footwork when your opponent is inactive perhaps more than I think about the footwork used to play defense against jukes in small spaces. Being able to backpedal/shuffle while your assignment is jogging down the field is a magical key to playing defense with better awareness. When a cutter is jogging down the field and their defender is looking only at them & jogging behind them in their wake, I immediately know I’m not watching an elite defender.

Now let’s look again at what happened when Brute Squad missed this help D opportunity. Watch the pair in the middle of the field: #19 on Scandal with the long red pants, guarded by #16 for Brute Squad. The disc is first swung to the center, and the Scandal cutter makes a non-committal under cut, then immediately “clears out” a few steps away from the open side.

After this short clear out, the Brute defender is slightly on the cutter’s under side. They have their back to the disc, and their eyes only on the cutter they’re guarding. At this point there’s 29:26 left on the game clock:

As the play goes on, the Scandal cutter moves downfield at a slow jog, and the Brute Squad defender simply jogs in their wake, not (as far as I can tell) seeing the disc at all—not really looking at anything besides the cutter they’re guarding.

A full six seconds later, with 29:20 left on the clock, they’re still just jogging after the cutter they’re guarding:

In this time, Kami Groom has run nearly 40 yards on a deep cut, and the potential help defense hasn’t even found time to look at anything other than the player they’re guarding.

This is pretty much a perfect demonstration of what I wrote: Being able to backpedal/shuffle while your assignment is jogging down the field is a magical key to playing defense with better awareness.

Footwork helps your brain work

There’s a few mutually-reinforcing concepts at play here. I’m not saying the answer is as simple as “just backpedal more, bro”. The way you use your feet can’t be fully separated from your mental approach to the game.

First, while I just said it’s not the only answer, physically being able to backpedal/shuffle, and to comfortably do it well and fast, is certainly part of the answer. I grew up playing American football, so backpedaling has always felt natural to me playing frisbee, but I know a lot of other players don’t have the same level of backpedal training. I always include a few reps of “backpedal turn and run” in my warmups:

My friend Zane (his Substack) recently sent me an Instagram post of former Alabama football coach Nick Saban talking about how backpedaling isn’t literally always the answer. Some players may prefer to shuffle/run sideways (”play in a half turn”, Saban calls it). So don’t obsess over backpedaling specifically. Any type of movement that maintains your ability to see both “your man and the ball” is what you’re looking for.

Second, it’s about having a mindset of trying to get ahead of the offense instead of simply reacting to what they’re doing. Here’s another segment from the defensive philosophies post:

Good defense often comes down to a mental challenge—fighting the instinct to be reactive in order to be proactive. If you’re just responding to what your assignment is doing, you’re almost certainly playing suboptimal defense. Great defenders are constantly reading the play and deciding for themselves where they want to be positioned.

In other words, the problem isn’t just “the defender wasn’t backpedaling”. It’s that they were content to just follow the offensive player around instead of using that moment when the cutter was moving slowly to gain an advantage. When offensive players are less active, mediocre defenders relax...but good defenders use that moment to gain an advantage. There’s no evidence here that the Brute defender was making any attempt to gain an advantage against an inactive cutter.

Third, it’s about valuing information as a defender. Both Hive and Travis Norsen have also written about the importance of “facing the space”. Facing the space is just as important on defense as it is on offense—if not more.

At the moment the switch opportunity was missed, about halfway between the two screenshots I included above, the Brute defender arguably couldn’t see a single other player on the field besides the one she was guarding:

Any time you’re in a situation like this, you should be fighting to gain more information. If you don’t have a good view of the field, and you don’t see that as a problem that you need to fight hard to immediately rectify, you’re most likely not playing good defense.

Fourth, I suppose it’s also simply about caring about team defense. If you’re not sold on the idea of helping and being helped on defense, then you won’t truly care whether or not you can see those help opportunities.

But when you realize how important it is to help your teammates, it quickly becomes obvious how important it is to put yourself in positions to notice help opportunities and execute them.

How to do it?

Here’s a pretty good1 positive example that I caught in the men’s final—Revolver vs. Machine. I’m not completely sure I can read the jersey but it looks like #10, Colby Chuck. [edit: I originally said here that it was Adam Rees, who is on the roster as #18, but commenters have informed me it’s definitely not him. The jerseys look a lot alike and I spent significant time trying to figure out which number it was an still got it wrong!]

Watch in this clip how, after his man throws the swing pass, he keeps his chest open to the disc for pretty much the next 8 seconds straight, using shuffling and side running. Yes his man eventually gets the disc back, but it’s on a 2-yard gain that keeps the disc right on the sideline—I’d call that a small win for the defense.

Here’s a couple screenshots to help you see what I’m looking at in the video above:

Final thoughts

[Updated 2026-01-28]: Brummie and I ended up continuing our discussion over email, and I no longer think we have a significant disagreement. Over email he said to me, “If you give me five years to coach a team, I’m going to double down on help defence.” So, we basically agree on the important part (help defense is a stronger strategy when the team knows how to execute it), though we may still disagree slightly on the much less interesting (to me) question of how quickly a team can implement a successful help D strategy.

I’m adding this edit instead of re-writing this entire section to save myself time, but I no longer think the rest of this conclusion section is an accurate summary of my and Brummie’s beliefs. The analysis in the rest of this article is unaffected, as this particular concluding discussion (help D) was just a framing for the actual topic of the article (backpedaling / awareness).

In our brief back-and-forth on Reddit, Brummie said that he’s basically unsure whether playing help defense is consistently more effective than “guard-your-own-person D”2, saying3:

I don’t think it’s ever possible to definitively say one approach is better than another scientifically....Some people require shut down defence to give the best outcome, some situations require help.

I’m of the opinion that a help defense mentality—once frisbee players and teams get good at it—is undoubtedly more powerful than pure 1-to-1 defense. The problem is just that we’re not good at it yet:

To me, it’s just so obvious that playing help D will be (at least in the long run) more successful. Yes, there are lots of examples of people sucking at it, as you point out in the article. But for me, that’s just evidence that frisbee players are still bad at playing help defense, not evidence that it’s impossible to be good at playing help D.

The clip we’ve looked at here is a perfect example of why I believe that. We still see players, and not just any players but players on national championship winning teams who lack what I would call some fundamental defensive concepts.

So, to recap:

• Use backpedaling and side running/shuffling to unlock your defensive awareness while you’re guarding a relatively inactive player.

• Look for opportunities to get ahead of the offense instead of always responding to what they do. When the player you’re guarding slows down, that’s not just an opportunity for you to rest (though it’s reasonable to to that too sometimes). Use that time to do more—scanning the field for help opportunities, harassing passing lanes, or getting a step ahead of the next thing your opponent wants to do.

• And learn to value information. When you don’t know what’s going on, you shouldn’t simply accept that state of affairs, you should fight hard to work your way back to a position where you can see what’s going on.

The pros of this type of post: there’s a very high information density here for readers who’re interested in taking the time to analyze all the small moments from this game. The cons of this type of post: it’s less focused on any one particular topic. And I don’t spend as much time explaining/making diagrams to make sure everything I’ve noted is completely clear. Let me know what you think of it!

I tried add names where I could, there were a few numbers I wasn’t confident I could read correctly, so apologies in advance for any mistakes I made there.

Here’s the game:

And without further ado, here are my notes:

Revolver defense:

9:37 Machine scores in essentially 1 throw. The deep guy seemingly kinda gets lost, focusing on the guy right in front of him while the entire side stack goes deep? But also I can understand him wanting to focus on the cutter who’s in an active threatening space, so maybe the weak side help should’ve rotated over the top?

You see some hand signaling from both guys lower in the side stack, weird play:

12:51 Kwon way downfield on the saggy handler defense

12:54 Just look at the positioning generally, I think the thrower misses the guy at the top sideline but there’s a lot of players not super close to who they’re guarding

13:03 Lew passing off Daan De Marree (gonna call him DDM from here on out) as he goes deep, but then he and Vallee both take the same under and the O gains a few yards.

13:06 Lew jogging down the middle of the field while his guy jogs down the far sideline. It’s maybe just enough to discourage the throw to #9 (Bontemps)

also 13:06 Notice how it’s not a pure “keep someone deep” scheme — there is that moment there where 2 white jerseys are deeper than the last two Revolver jerseys. Who has the disc tho? #11 Goff, I think he’s a solid deep thrower

also 13:06 Watch how Seamus Robinson is proactively getting further away from his man

13:10 Vallee shifts on the pump fake to the little upline throw, forcing the swing. intentional?

13:12 as Goff goes backwards, Vallee just stands right in front of the disc

13:19 Chicago sends two deep and two under and there’s enough confusion that the unders get open. You can see the guy on the tip of the raven’s beak signalling to his deeper teammate.

This is where you gotta “talk from the back”... let those further under guys know they have help before they turn around and start running

(It’s interesting how 2 points into the game we’ve already seen like 3 objective mistakes playing frisbee this way. Even the best D in the country is going to make mistakes.)

13:20 I don’t like how Vallee randomly starts playing tight D in the backfield here? After playing more poachy a moment ago.

13:28 what the heck happens here with #50 Lew coming under from the deep D space and by the time you see him, he’s just staring at the ground picking at the turf? Something weird happened.

(Turnover #1), but it’s not especially due to great defense, it feels like they missed the chance for an obvious switch there with Lew staying deep, but the throw wasn’t amazing and the read wasn’t great either (as Keith says from the booth immediately after)

14:29 watch how Lew is guarding the guy running top-to-bottom on the screen but as he sees the swing pass he immediately starts ignoring that to head back towards the top sideline.

14:35 Playing a little too saggy (if anything it looks very zone-y right now), and gives up the pass to that same guy w/o there being any other receivers nearby that he’s really shutting down.

14:38 Watch how #20(?) Liston guards the under and then just stands there as his guy clears out deep

14:43 DDM make a nice hard jab step to deep cut. Liston ended up not having the deep help he thought he had a moment ago, as the field switched and a guy cut under breakside and Robinson was focused on DDM

18:30 Vallee kind of “outruns the (rest of the) pull coverage”-- he gets downfield to set a mark on, but the cutter defenders aren’t in position and Revolver gives up an essentially unguarded first pass. I think you gotta either (a) sag more as the guy running down to the handler. Don’t set a hard mark unless you’ve seen your teammates are there to guard cutters. or (b) get cutter defenders to get down there just a bit quicker

18:35 Clearly someone gets wide open deep -- watch how Lew takes off sprinting deep and one of the Machine cutters is also pointing deep begging the thrower to put it. Unfortunately for them they’re already looking swing pass. Like freeze frame at 18:33 and there’s 7 Revolver defenders on screen but only 5 Machine players.

18:39 Seems like a missed chance for a spontaneous bracket. #35 McGuckin who ends up getting the disc dips his shoulder out of the way of a Revolver defender who just lets him go.

18:40 The open side handler defender looks to do a little “stunt the lane” but is just a moment too soon. But I agree w/ the thought process to be there even at the expense of maybe allowing the zero-yard swing pass.

18:44 Bit of an “overplay” by Lankford running out of bounds going for the D, but McGuckin really has no way to punish it. This is a situation where (with my American football background) I’m trying my hardest to backpedal for as long as possible. Look what happens to Lankford -- he immediately turns his hips to run, then has to turn them again when McGuckin cuts horizontally for the short pass. But McGuckin was just jogging downfield, not running truly fast enough to *force* Lankford to turn and run. Lankford has to turn his hips twice — I think I could’ve made this play with zero hip turns (in a world where I was athletic enough to be playing in this game).

18:50 Ing (like Lew earlier) jogging down the middle of the field while his guy is jogging down the far sideline). Maybe 6-8 yards gap between them? It’s also beautiful to see how **Goff ends up jogging back to him**

18:55 Handler defender once again playing downfield of the disc, allowing the trivial backwards pass.

25:47 At the top of the screen the three cutters kinda end up making a triangle of replacement cuts, and while Robinson(?) stays last back and points for someone to take his man under on the switch, it doesn’t happen smoothly enough to stop the next pass. but theoretically could have...Good news is they keep the disc on the sideline.

25:49 Again see how far Ing is from his man as the disc goes to the top sideline.

26:13 Kwon plays help D to shut down the DDM cut (watch him peeking, watching DDM as his guy walks back into the stack, before committing at the last possible moment when he sees DDM getting open), Robinson switches over to help on the scoring throw, they force the turn (Turnover #2). Probably would’ve been a goal if the throw was a little lower though.

27:42 I’m kinda impressed with how quickly #8 (Whited) starts playing defense after the pass is intercepted. Elite mental transition, he’s immediately looking around figuring out what he can shut down.

27:50 Another nice little switch at the top of the screen to easily shut down a deep cut, but the underneath guy can’t close fast enough to stop the horizontal under cut.

27:53 Again see Robinson putting his hand up calling for & getting the easy deep switch.

29:35 Robinson commits a little too hard when DDM slows down and looks like he might turn back under (give up the small gain instead of the goal!), and there’s no help from deeper in the stack the way there was on the first turnover of this point.

34:36 (Turnover #3) Honestly nothing *too* fancy on the poach here, just your typical “my guy is clearing deep into the stack, I’ll take a peek and see if I can get into that open lane” poach. Tho think about how knowing he’s gonna have deep help leaves him free to take that chance. Also see Clyburn easily passing of DDM deep at 34:32. Also think about how the length of his cut allows more poaching (”saddle point” theory-adjacent thought)

I HATE that he didn’t catch it tho. Could’ve immediately hit Lew on stall zero and be off to the races.

37:40 (Turnover #4) Hard to say anything besides “bad throw” on this one...

38:30 Kyle Lew positioning masterclass again. Even though there’s no mark on the thrower and his guy is right there, he’s still playing like 2-3 steps off, shading open side, and manages to keep a throw from going off. Then lets Bansfield go deep knowing he’ll have a switch

38:39 Robinson from the back just lets his guy have a free under cut to the sideline as he makes sure nobody’s going deep.

38:50 (Turnover #5) Partly just a bad throw. Note how Ing switched at 38:47 then is poaching to help on the throw 6 seconds later.

42:03 Again poachy handler on the first pass.

42:16 #12 (Smith) commits too hard as his guy starts to move under/horizontally and gets beat when it turns into an upline cut.

52:21 (Turnover #6) just poaches everywhere when Machine has the disc deep in their own endzone, daring them to throw something deep. And it works. “Get to the sideline early” on the pull. (Also, didn’t the disc roll out of bounds? By rule I think he was supposed to take it to the front cone)

53:38 easy pass off switch of the cutter running horizontal across the midfield line.

53:54 Ing is pointing out his man at 53:49, then kinda rotates over to the break side (bottom of screen) as if he’s covering that half of a bracket (at least I assume that’s what it is b/c it’s kinda hard to figure out any other way to explain how he positions himself here), but then when a pass makes it thru on the open side he’s out of position (not saying it’s his fault) as “his” guy cuts back towards the open side, and the two deeper defenders aren’t thinking of helping. Maybe a tiny bit of ball watching (+ layout watching) too. But if the other half of his bracket does a better job shutting down the cut to that side, maybe none of this happens.

54:00 (Turnover #7) Seems like just a pure drop / thrower could’ve put more touch on it, nothing really special happening defensively here.

55:58 Again poachy handler setup after the pull:

57:41 (Turnover #8) Technically one of Machine’s turnovers but with the benefit of slow motion the community seems to near-unanimously agree that it was a good (inbounds) catch

1:04:17 Vallee commits a little too far forward even tho there’s already two defenders there to guard two Machine players, and then the next pass gets off too. Hard to say for sure but it seems a little more like they might be actually playing zone on this point, not just poachy.

1:04:28 (Turnover #9) Just a bad throw, really. Offense hadn’t really even been under pressure this point.

1:07:26 (Turnover #10) Kinda straight up drops it.

1:09:45 This time it feels like the handler D was a little *too* poachy off the pull, giving up about 10 yards on a pass to the side handler. I like poaching the lanes but if giving up that much on a pass between handlers is not very sustainable. This was a little too much “stand right in front of the disc” and not enough “stunt and recover”.

1:10:00 Defender chases Smith (#15) a little too hard to the breakside, and then gets beat when he turns back to the open side. Feels like Revolver never really got pressure on this possession overall, I really don’t understand why #8 (I think? doesn’t look like they have a #0) ran towards the disc on the completion at 1:09:49...why not stay downfield with DDM?

1:13:43 Hard to tell exactly what happens here b/c the camera shot doesn’t show the whole field until right before the throw. Revolver poaches off the “side stack” so much that the handler just throws a hammer right to two cutters as soon as the disc is checked in. Maybe they needed to keep one guy a little closer there? Clyburn is helping over the top more in the centerfield, could he have shaded more that way?

Might also count Clyburn’s reaction as “defensive overplays“—he could’ve (a) realized early that he wasn’t gonna get there, and started backing off to guard the next pass, or (b) realized late that he wasn’t gonna get there and stopped early to at least put a smothering mark on, but instead he (c) jumps at a disc he’s multiple feet away from D’ing and isn’t back to good marking position in time to stop the scoring throw from going off.

1:17:31 I like how #8 Whited takes those first few steps deep as another cutter (i.e. not currently his man) starts a deep cut, that’s the kind of “last back mentality” I think good D needs, then he recovers back to a position he likes when he sees that throw is not going up.

1:17:55 (Turnover #11) Just an impressive footblock.

1:20:07 Robinson is just a split second too late finding the next cutter he needs to shut down. Or maybe it’s the wing defender at the bottom of the screen (#8 Whited I think?) who messes up by letting the cutter he’s near draw him deep, when he should’ve let him go and taken the under cut. Machine does a solid job of kind of flooding the strong side—maybe the whole D should’ve shifted a couple steps further that way, e.g. Vallee didn’t really get anything done in the spot where he was.

1:20:37 watch Kyle Lew as his man goes upline, and he just stops in the lane as his man keeps going, making it impossible for DDM to get open to that side. In a way they were kind of triple teaming Daan De Marree on this play given what Vallee does a moment later:

1:20:41 Vallee goes for the aggressive poach block, doesn’t get it, DDM smart enough to hit the next man immediately. Revolver has 3 “defensive overplays” in 10 seconds, with the layout at 1:20:32, Vallee’s layout at 1:20:41 and Seamus Robinson coming off his man to try to help-the-helper and guard the man Vallee left, but he’s so late that he would’ve been better off just staying where he was. (Notice how then Ing almost gets the D on the scoring pass, this is a pretty good sequence to exemplify “helping the helper”, just a step too late. Keith calls this out too after the score! Overall I’d have to give pretty high marks to his analysis in this game)

Everything else (Machine defense, Revolver offense)

7:51 good job “beating D to the 2nd option”

13:05 I like how Nate Goff isn’t running, he’s kinda shuffling/sidestepping further into the open space.

13:20 a good resolution to a double cut. also notice how both are calling for the pass.

13:48 really getting a body into the D before making the cut by #79 Ing, honestly I might call a foul if I’m on D here, I love playing that ‘seal the defender’ game but this was maybe a little too much. Tho also the D was trying to block him from getting across the field, so maybe him sliding in front of Ing was a foul too?

13:56 Goff tries to stay in the lane but Lew reads it well enough to get it to the 2nd man.

15:37 The “travel” on the flick huck looks especially bad in slow motion

17:00 Seems like Machine missed a chance for a little switch when Frankenburg slashes across for the score. Also, the defender at 16:58 seems like they made (in retrospect at least) a poor decision -- they come up to help the handler cut but are nowhere near close enough to actually affect the throw. Then they guy they left alone is way open for the next pass

I think this is some good “point five frisbee“ from Revolver. You don’t really think of the forehand inside break being a stall zero throw, but #43 Prior recognizes that’s what he has and just goes for it immediately. (Immediately after the score Keith points out “the defenses are not executing the game plan very well”)

23:24 look where Hecht is standing as the centering pass is thrown and where he finally sets his pivot foot. and as the disc lands at 23:20 he’s still looking downfield so he know how much time and space he has. (see Catch the centering pass facing forwards)

24:15 They chase Hecht too hard to the far sideline then get back back to the middle when he stops hard. Compare to the Lew/Ing positioning previously (screenshots above).

24:18 is a good example of what I’d call bad (not working hard to get downfield of the disc) defense by the guy in the dump space (Smith?) [I haven’t written a full post explaining my thoughts on this topic before but maybe will sometime]

It’s kind of an interesting cut for the score because that guys in a “front of stack” position, and it seems like his defender is thinking “he’s gonna make cuts that are horizontal or towards the disc, he’s not gonna cut deep at an angle towards the endzone, because....who does that in this situation?”

27:05 I kinda hate this pick call, “it’s everything that’s wrong with the sport” (kinda)

27:20 I like this tiny dump-> breakside throw sequence. Feels like a lost opportunity for Machine that they didn’t switch to shut down the throw.

27:43 Kinda an impressively bad throw to be seeing in a national final (as Keith soon comments on)

31:14 Very poachy mark from Bansfield coming out of the pull

31:57 Feels like (in retrospect) Bansfield commits a little too hard to stop an option that he was a little late to anyway, leaving his own guy too open for the help D to help him in time (Tho, if Keith’s analysis a minute later is to be believed, maybe the defender in the middle was supposed to stay in the middle?). Good job “finding the 2nd option” by Hecht (maybe even something of a look off!).

35:10 Again compare what happens on Clyburn’s pass to the help D we saw from Kwon a couple minutes ago. And great “point five frisbee” to be looking for that shovel into the endzone IMMEDIATELY.

35:11 Vallee has the slightest bit of hesitation on this cut. And then, while the disc in still the air his defender puts both hands up as if confused by something. not really clear to me what happened here exactly.

36:41 Kinda embarrassing to not be catching this particular pull (...as Keith mentions lol)

39:26 Maybe another “defensive overplay” though the score a few passes later is more just “crazy good pass” than anything else. But if Lew didn’t go for that pass, they had a big 2-on-1 advantage

44:14 They kinda gave up the pull play BUT I’ll say at least the help defense saw it and was ready to react to it, the disc just got put to a spot where only the cutter could get there first.

50:44 Surprisingly (to me) he may have actually not been in the air yet when he caught it? Didn’t get called tho.

54:01 Double cut from Chicago. [For my notes, might write about double cuts soon]

54:27 Machine defense has seemed really out of gas this whole possession since the turnover (as the announcer immediately says).

1:00:52 Gordon really threads that pass in there. He’s a master of those little backhand passes to space when the defender doesn’t have eyes on the disc.

1:02:18 You don’t have to call a foul. Machine tries to poach it but not quite close enough

1:02:30 Clearly Machine gets kinda confused with the numbers, as Revolver has a 2-on-1 at the disc for a score. You can see in the drone replay (1:03:23) that all the guys at the front of the endzone more or less have their back to the disc. Yet another reason to hate faceguarding -- no one who grew up playing basketball would’ve made this mistake...

1:04:56 DDM poaches way into the lane to discourage the next pass from going downfield. So the offense swings it. And then his guy cuts breakside for the continuation pass. And the defender who’s already there is focused on his own man when he may have been able to switch onto DDM’s man filling that spot and stop the next pass from going off.

1:05:42 DDM lets Robinson go deep, calling for the switch, then he (DDM) floats towards the breakside where the guy who was deep is coming under. Except then Robinson turns and goes back under on the open side. The guy guarding Ing, who’s now going deep, could’ve switched onto this under cut where Robinson gets the disc, but *he* (I believe it’s #17 North) isn’t looking to play switchy in the moment so two people are near Ing deep and Robinson gets the disc. (And then two people go to mark him. Machine D getting a little worn out The 2nd defender had tried to poach off Lankford but was way too late)

Really highlights how against a smart offense poachy/switchy play has to be a whole team philosophy. This switch probably would’ve worked, but it needed to become a triple switch and didn’t. That’s the 2nd moment this possession where DDM plays what I would call smart defense but “his man” gets the pass because his teammates aren’t on the exact same page he is. After the score you can also see the double switch moment in drone view at 1:06:26

1:11:27 Marginally disappointed with how this pull is handled by Revolver. Not the worst mistake in the world but at least slightly suboptimal to let the disc get that far after bouncing.

1:12:00 Defender kinda stops guarding his own man (or was he expecting help) b/c the pass to Gordon is in the air right in front of his face, and as a result Gordon has an immediate 10 yard follow up. And it just so happens at the same moment Adkins’s man had bit too hard when he faked coming back under, and *he* (the defender) slows down a little b/c of the temptation to play D on Hecht with the disc. The two defenders who were furthest downfield both had their backs more or less turned to Adkins and don’t look to help until it’s too late. So kinda just a waterfall of small errors that all happened to go off at the same time.

Also peep the “inside pivot” cut from Adkins at 1:12:02, keeping his body open to the thrower.

1:12:20 “Blink of an eye is kinda what you get with this offense” Keith says (point five frisbee). He follows it up by saying you’re in trouble if the offense is moving the disc at “stall three”, but I think it’s pretty obvious watching that they’re trying to move the disc at stall zero, not stall three.

1:20:20 Notice how Rutledge Smith (#15 Machine) gets open by shuffling slowly into the open side, making it harder and harder for Robinson to keep an eye on both him and the disc (walking before cutting)

1:22:54 Machine is playing their switchy defense, but there’s a bit of a miscommunication as the guy guarding Hecht (cutting top to bottom across the screen as we view it) leaves him to focus back towards the center of the field, but the other defender who was in that area also pinches in, leaving Hecht wide open near the sideline.

Hecht definitely gets a little lucky that his game-winning throw isn’t a turnover.

Also note how, a few seconds earlier (1:22:51), Hecht’s man tries to layout D the reset pass (yet another overplay), leaving him out of position when Hecht immediately runs away. He has to run extra hard to attempt to get back in position and I think him being a half step behind mentally (as he’s over-working physically) leaves him in position to not make the smartest play in communicating all the potential switches.

Honestly I think Machine’s defense was pretty solid too. Revolver was just a tad better at not making defensive mistakes, and a tad better on offense at exploiting defensive imperfections.

Final thoughts on Revolver D

Looking at Machine’s turnovers, it’s not always easy to say that Machine turned the disc over because of Revolver’s schemes.

• #1 at 13:28 — A huck that wasn’t quite good enough. The offense wasn’t under a ton of pressure so far that point.

• #2 at 26:13 — A slightly high throw. Technically an accuracy issue, but Revolver played good help defense in the moments leading up to this pass, perhaps we can say the mental strain that defensive pressure put on the offense helped lead to the throw being slightly more inaccurate.

• #3 at 34:36 — A good (but perhaps not exceptional) poach into the lane—possibly it was actually a switch that just looked like a poach b/c we never got to see the switch completed due to the throw being released. Knowing he had deep help enabled the confidence to poach so I’d definitely give the scheme some credit here.

• #4 at 37:40 — Just a bad huck (at 0 stall count after a stoppage)

• #5 at 38:50 — Throw floats too much. Mostly an accuracy issue but the help D was there to make sure the Machine receiver didn’t have extra time to wait for the disc to come down.

• #6 at 52:21 — Pretty suffocating, poachy defense from Revolver with Machine backed up against their own back line. But the biggest mistake on this point was not catching the pull. Good job by the Revolver D to exploit that mistake to the fullest though.

• #7 at 54:00 — Receiver was open, just dropped it (and/or should’ve put more touch on the throw)

• #8 at 57:41 — Not actually a turnover, slow motion pretty conclusive that it was caught inbounds

• #9 at 1:04:28 — Just a bad throw / bad decision to huck when the offense wasn’t under pressure.

• #10 at 1:07:26 — Stone cold drop

• #11 at 1:17:55 — Nice footblock. Individually great D but not really related to Revolver scheme.

11 turns. 1 was a bad in/out call by observers. 1 great individual mark defense. 2 drops. 3 obviously bad hucks/bad decisions to huck. That leaves 4 that you could give some amount of direct credit to the scheme (being maybe a little stingy here on purpose).

It’s easy enough to point out places where they made small mistakes, but on the other hand, they did enough right to get a big win over Machine. I’ve written before: “Don’t try to figure out how to be one step behind all of the time. Figure out how to be zero steps behind, some of the time. Even if that means you’re 10 steps behind on some other passes” We see some of that here—while there were some plays where Machine got very open due to small Revolver mistakes, there were other plays where no one got open at all.

Overall the philosophies that stand out to me about Revolver’s defense are:

• Poachy handler D, especially right after the pull

• Lots of switching

• Lots of help, particularly:

  • helping over the top

  • helping the helper

• Smart positional play:

  • sagging towards the middle when guarding someone far from the disc

  • Staying in threatening areas when guarding a player who tries to clear out away from them.

• Knowing you’ll have help over the top enables clogging up more space near the disc.

• Focus on scheme over matchups — look at how many different players guarded Daan De Marree over the course of the game.

This is pretty in line with most of my beliefs around good defense. I haven’t written any posts specifically about switching, but that’s just b/c I don’t think I have anything to say about switching that hasn’t been said already. For some of my previous posts that do touch on concepts we’ve seen here, check out:

• Defense is for stopping the other *team*

• Fury’s defense will ignore your weakest link

• There’s still alpha in help defense

• Why you shouldn’t always play tight defense

• My philosophies on: playing defense

• Thinking carefully about defensive overplays

• Stunt and Recover

I think the Revolver defense still has lots of room to grow, too (a scary thought). Three of their biggest D-line contributors—Clyburn, Lew, and Robinson played their final college season in 2025 (Clyburn, Lew) or 2024 (Robinson). (Plus Lankford seems to still be in college) They’re already one of the closest examples I’ve seen to a team playing the way I believe defense should be played...I think if the D line can sharpen up their positioning and instincts just a little more, they can become a truly suffocating defense to play against.