The city of brotherly love has enjoyed the services of many great pitchers in recent years, but none have clutched the hardware like Halladay did. The other three of the “Four Aces” received votes around that time: Roy Oswalt finished 6th the year Halladay won, and Cliff Lee (3rd) and Cole Hamels (5th) placed behind Halladay’s runner-up robbery in 2011. Inching closer to the present: Aaron Nola finished 3rd in his infamous “WARs at war” 2018 campaign; Zack Wheeler finished 2nd in both 2021 (a year in which it could’ve gone any of four ways) and 2024; and Cristopher Sánchez—the hero of this week’s story—came in a unanimous second place to Paul Skenes last year. This year, Sánchez may very well continue the trend of Phillies aces coming up just short yet again, this time at the hands of the singular Jacob Misiorowski.

My purpose here is not to argue that Sánchez should win over Misiorowski. But I will make the case that Sánchez has been rather unlucky to not break this unfortunate Phillies streak this year. While part of his bad break stems from the Miz doing things on the mound we’ve never seen, part is also found in the numbers if we dig deeper.

Beyond the historically great velocity, the heart of Misiorowski’s case lies in his sub-2.00 ERA (1.97). Even though voters value fielding independent metrics more than ever, the “what actually happened” part is still strong (and when we have multiple near-equally elite pitchers, it probably should be!). ERA is likely why Chris Sale (2.06) currently has better odds to win the award than Sánchez (2.52), even though the latter’s superior workload puts him ahead of the ‘24 CY winner in most other facets.

Given what we know about the instability of ERA, an argument could be made that Sánchez is unlucky to have an ERA that’s “only” in the mid-2s while his competitors sit around 2 flat. It’s not as if Sánchez’s “under the hood” numbers are profoundly better than the others’, though—Miz’s peripheral metrics indicate his ERA is no major fluke. So is it really just ERA? To voters: basically. However, the context surrounding the difference is illuminating.

WAR has a lot to say about the NL Cy Young race, of course. Sánchez will seemingly have an advantage here given his higher innings count (175.1) compared to Miz (155.0) and Sale (144.0), but the difference materializes much differently in bWAR than in fWAR. They’re all pretty much within a stone’s throw by fWAR:

1. Sánchez - 5.7 fWAR | .29 fWAR/9

2. Misiorowski - 5.6 fWAR | .33 fWAR/9

3. Sale - 5.3 fWAR | .33 fWAR/9

But bWAR tells a much different story:

1. Sánchez - 7.7 bWAR | .40 bWAR/9

2. Misiorowski - 5.3 bWAR | .31 bWAR/9

3. Sale - 5.1 bWAR | .32 bWAR/9

Woah! Isn’t bWAR the version most folks agree should be used for awards…?

Things have changed, it seems. The aforementioned 2018 Cy Young race with Nola was arguably the start of the shift away from bWAR, as most believed his case to be weaker than deGrom’s (and Scherzer’s) despite boasting the higher bWAR. Voters understood the dominance of deGrom back then even with his 10-9 record, so 2018 can also be considered pivotal in terms of the gradual devaluation of pitcher wins and losses (looking at you, Rick Porcello).1

We are now at a point where a CY-caliber pitcher who exceeds his competitors by roughly two and a half bWAR is more or less no longer in the running for the award due to other factors. I’d say that’s a fairly substantial shift. Maybe voters of yesteryear would have appreciated Cris’s case more, but the Phillie picked a bad time to be on the wrong side of the trend when a guy like Misiorowski is doing what he does. I mean, have you seen him pitch?

I feel like even if their ERAs were closer that voters would still be biased in favor of the Miz due to his unprecedented combination of stuff and command. Simply put, he has been more exciting to watch. Who doesn’t like seeing 103-104 mph regularly from a guy who can consistently zone it? Shock and awe abound every fifth Brewers game. The fact that it has all come together for Jacob this season is incredible and he absolutely deserves the award if he finishes well enough, in my opinion.

…BUT……

How do we reconcile the fact that Sánchez’s BABIP against is over 100 points higher than Misiorowski’s?

Share

Yes, that is a real stat. Among qualified pitchers this season, no one has a worse BABIP against than Cris (.335). Only two (Yoshinobu Yamamoto and Keider Montero) sport a better BABIP against than Jacob (.231). I don’t know exactly how many standard deviations of a difference that is, but I’ll wager it’s a handful.

While not a pure pitch-to-contact archetype (he’s no Sandy), Sánchez definitely has more of that flair than Miz, who has none of that. That means he’ll pretty much always allow more balls in play than Miz will, and that means Sánchez’s .335 BABIP against hurts all the more. So why is the discrepancy that large?

For starters (pun intended), Miz seems pretty much guaranteed to have a low BABIP ceiling thanks to his insane stuff, so that could definitely be part of it. And on Cris’s side, Philadelphia’s mediocre defense may be making it harder to convert ground balls into outs. Overall, though, there’s simply a lot of variance in a pitcher’s BABIP against. And by “a lot,” I mean “I dare you to find a pitching stat that’s less stable” a lot. A trustworthy estimate puts the stabilization point (i.e., the point at which most of the variation in a stat can no longer be attributed to chance) of pitcher BABIP at roughly 2000 balls in play. For reference, league leaders in balls in play nowadays tend to allow around 600 in a season.

Now, if it takes at minimum three—but realistically four or more—seasons for us to be able to trust pitcher BABIP, how much stake can we really put into the 100+ point gap? If we take the 2000 BIP stabilization point as gospel, then we can actually regress Sánchez’s .335 down to .296, and Miz’s from .231 up to .279.2 All of a sudden the gap between them appears much more modest when we account for the instability of the metric. So how much better is Misiorowski than Sánchez at preventing hits on balls in play, actually? The answer, I would claim, is a good bit less than what their results have indicated.

“Awards should be based on results, though!” I hear you. And I agree with you! When you have two stud pitchers vying for the trophy, why not give the edge to the guy whose results were better? It makes sense. My point is that we happen to live in the reality where Sánchez’s BABIP is .335 and Misiorowski’s is .231, and we could have very easily not gotten that reality. And, if their BABIPs against were more like their regressed estimates, the Cy Young race would probably look a whole lot different.

It all circles back to ERA, a stat that depends on BABIP a great deal. Forgive the following math as it’s very rough, but the point remains. If Jacob’s BABIP were in fact .279 rather than .231, that would translate to ~14-15 additional hits conceded, which can be conservatively estimated to be worth ~7-8 additional earned runs allowed. This would bring him up to 41 ER from 34, resulting in a regressed ERA of ~2.40 (this is keeping K% and BB% constant!). If Cris’s BABIP were in fact .296 rather than .335, that would translate to ~18 fewer hits allowed, or a ~9 ER swing in his favor (49 → 40). This would yield a regressed ERA of ~2.05.

If their ERAs looked like that instead, who would you back? Their fWARs would be the same. Their odds at the Cy Young would not be.

Even though Sánchez has been getting hit harder than Misiorowski this season, Sánchez has also made up for it by stepping up in high-leverage moments—a.k.a. “being clutch”:

• Sánchez: 4.15 WPA | 2.65 WPA/LI | 1.73 Clutch

• Misiorowski: 3.41 WPA | 4.45 WPA/LI | -0.86 Clutch

In isolation, Miz has produced outcomes that better contribute to winning. Sánchez has not done as well on that front, but his positive impacts have been more tangibly felt by his teams in his games thanks to better sequencing. Much like with BABIP, I won’t pretend clutch/sequencing is repeatable either, but it does help explain how Sánchez and Misiorowski have manifested their value differently.

The best way to tie all these points together is to view the different components of their RA9-WAR—a WAR variant from Fangraphs that’s more in the spirit of bWAR than regular fWAR is. RA9-WAR consists of fWAR added alongside BIP-Wins and LOB-Wins. BIP-Wins estimate how much value a pitcher has added from his BABIP, whereas LOB-Wins estimate how well a pitcher has done to control baserunners. You see where this is going.

• Sánchez: 5.7 fWAR | -1.9 BIP-Wins | 2.3 LOB-Wins | 6.1 RA9-WAR

• Misiorowski: 5.6 fWAR | 1.6 BIP-Wins | -1.2 LOB-Wins | 6.0 RA9-WAR

Why oh why did Sánchez’s misfortune have to fall into the bucket voters scrutinize (BIP-Wins) and not into the one they don’t (LOB-Wins)? Oh well; that’s baseball. The modern Phillie ace curse persists.

Subscribe now

Apologies to Chris Sale for not giving him enough credit in this article, but he already got his CY over a Phillie. “Only enough room for one C(h)ris here” was my thought process while writing this, and is also about what I imagine they’ll be thinking to themselves as they face off in Philly tonight.

Herein lies the problem that’s going to rear its ugly head yet again come October. No, I’m not referring to the postseason having 12 teams, though there is certainly a lot to wrestle with there. Instead, I am referring to a much more concrete issue: seeding.

Six teams on the AL side and six on the NL battle it out amongst themselves to determine who plays in the World Series. A perfectly sound approach, but its current implementation is lackluster in that the way teams are seeded creates consistent disadvantages that are not commensurate with what teams deserve. These disadvantages surface because the best Wild Card teams are more often than not better than the worst division winners.

Think about it: To be a Wild Card team, you have to be among the best of twelve non-division winners, but to be a winner of your league’s worst division, you only have to be the best among a weak crop of five. The pool Wild Card teams come from is consistently richer! And this is not just a theory. Since the postseason expanded to six-per-league (2022-present), 4/5 seeds (the two best Wild Cards) have averaged about two and a half more regular season wins than 3/6 seeds (worst division winner + worst Wild Card). Whoever comes out of the 4 vs 5 matchup has been more likely to be a better team than the 3 vs 6 victor. So then, why does the 1-seed have to face the winner of 4 vs 5, while the 2-seed gets the comparatively weaker 3 vs 6?

The expanded format exacerbates this issue, but it’s not the main culprit. It has persisted ever since the Wild Card was instituted over three decades ago. From the mid-90s up until 2011 there were four-per-league, which resulted in the 1 facing the 4 and the 2 facing the 3. The vast majority of the time, the 4 was better than the 3, again because the best Wild Card has much greater odds of being better than the worst division winner. Expanding to five-per-league alleviated this somewhat, but still resulted in 2s enjoying easier matchups on average than 1s. Six-per-league has made it even worse since the 2 now gets decent odds at facing the 6 instead of the 3.

Is rewarding division winners so important that 2-seeds should consistently enjoy the easiest path? What incentive do 1-seed hopefuls have to finish as the 1? Is guaranteed home-field advantage in the Championship Series worth having to face harder competition in the more random best-of-5 Division Series?

Share

We are on a collision course this year to have one of the more lopsided brackets in modern postseason history. If the season ended today, the 79-win Rays would face either the 75-win Yankees or the 73-win Red Sox, while the 70-win White Sox would await either the 67-win Astros or the 68-win Guardians (or whichever other middling team slides into the 6-seed). As it stands right now, every team on the side with the 1 would be better than every team on the side with the 2.

There are plenty of examples I could highlight (literally the majority of brackets), but the worst offense in recent years was the 2021 NL. The 107-win Giants had to face (and lose in heartbreaking fashion to) the 106-win Dodgers in the DS while the 95-win Brewers faced the 88-win Braves (who eventually won the World Series, likely coaxing fans into forgetting about the bracket injustice). This was a classic case of two historic ball clubs having to face each other early due to the crime of sharing a division; this season, the problem is surfacing on the other end, with the worse side being profoundly mediocre.

So, we’ve established that it’s better to be the 2-seed than the 1-seed most of the time. But what about 3-6? Turns out those incentives are all out of wack, too. It’s best to be the 3 (because you get the weakest Wild Card and weakest bye team), but after that, it’s probably best to be the 6! Followed by 4, then 5. So our progression from easiest average bracket to hardest goes 2 → 1 → 3 → 6 → 4 → 5. You may notice those numbers are… out of order. Achieving a more harmonious 1 → 2 → 3 → 4 → 5 → 6 in terms of bracket advantage seems more reasonable. Of course each progressive seed should have incrementally harder times in the bracket. Otherwise, why even seed?

Under the current league alignment, I would propose seeding all non-bye teams by regular season record. Simple! And kind of obvious. The NBA has been functioning under a similar format for a long while now, and—play-in gripes aside—no one is really looking to change anything about that (In fact, when they tried keeping division winners as highest seeds after expanding to three divisions per conference, it was so unpopular that they scrapped it after just two years.). The NFL offers division winners seed priority, but assuages the issue somewhat by reseeding after the Wild Card round. MLB does nothing to mitigate the competitive imbalance divisional seeding creates, and has consistently fallen victim to it as a result.

Compared to the current format, this change would seem rather punitive for the worst division winner. After all, winning a division is still perceived to be a big deal by many, even if said division is weak. My counter to that would be to ensure the worst division winner still retains home-field advantage in the Wild Card round. That way, they still reap the benefits of being a division winner while not souring the balance of the bracket.

You’ll notice that I wouldn’t demote the second-best division winner if their record weren’t second-best, since I do feel that distinction still deserves a bye (à la the NBA playoffs in their two-division era). Personally, both bye teams being from the same division would rub me the wrong way in a league like MLB, where divisions shouldn’t be that meaningless. But there’s an argument to go all the way seeding all six by record!

Here’s how this proposal would look for the AL this season:

• 1-seed Rays (79-54)

  • host winner of 5-seed Guardians (68-66) @ 4-seed Red Sox (73-60)

• 2-seed White Sox (70-63)

  • host winner of 3-seed Yankees (75-58) @ 6-seed Astros (67-67)

We can still debate the number of teams that deserve to make the postseason; the AL this year is particularly unflattering in that department. But, this slight tweak to seeding would do a lot to ensure a more reasonable reward structure for each seed. As the worst division winner, the Astros would be properly seeded as the 6 instead of the 3, but would still retain home-field advantage against the better-seeded Yankees. Meanwhile, the White Sox still receive a division winner advantage via the bye, but are threatened with a harder opponent (Yankees) than they’d get in the current format (Guardians). And as the top seed, the Rays aren’t counterintuitively thrust into both of the harshest options. Everyone gets a fairer slate!

Subscribe now

Although I believe this would be a positive change for the MLB postseason, my frustrations extend beyond seeding issues and into the structure of the league itself:

The elephant in the room is that having three divisions per league is just such a headache. Brackets are so much better when things come in twos! If there were only two divisions… you know, I’ll leave my commentary on that to the article above.

In conclusion, MLB seems fine with the postseason as is because it creates drama down the stretch and gives worse teams a fair shot at the crown. However, I would argue that this structure merely serves to fabricate parity rather than foster it. Structural obstacles in the name of balance are anything but balanced, and there are better ways to streamline the structure so that the process serves the outcome rather than dictating it.

But MLB prefers bandages to surgeries.

Yordan Alvarez is not his league’s leader in either WAR, yet is the heavy favorite for AL MVP. This is because he is by far his league’s best hitter, and hitting value is more tactile than other forms of value. Consensus seems to imply that his standing as a DH/LF is reflected too punitively than it should, and/or that his batting value isn’t rewarded enough. His candidacy is receiving much more favor than Bobby Witt Jr.’s, whose premium defense has propelled him to slightly more WAR than Yordan despite playing in 16 less games.

It’s valid to feel like Yordan deserves the MVP more. There is reason to believe current positional adjustments rest on too wide of a range, which is a major point in Yordan’s favor. But are voters going to choose him because of methodological disagreements, or because of aesthetic preferences?

The unfortunate reality for Bobby is that his way of accruing value isn’t as sexy as Yordan’s. Which reminds me: it’s time for an ad break…

“It really is the shoes.” Ha!

This old habit of homer worship will die hard, if it dies at all. Last year’s NL MVP finalists were Shohei (duh), Juan Soto (9th in NL fWAR), and Kyle Schwarber (14th in NL fWAR). It doesn’t really matter since Shohei was so obviously the winner, but isn’t it telling that Soto and Schwarber were the finalists and not Geraldo Perdomo, Trea Turner, Corbin Carroll, Francisco Lindor, or Fernando Tatis Jr., all of whom bested them by fWAR? Isn’t it telling that this season’s AL MVP runner-up according to odds-makers is not WAR leader Witt, but hard-swinging Junior Caminero (9th)? Witt wasn’t even granted honorary finalist status last season despite being clearly superior to J-Ram—a decision I suspect was more political than anything.

This season, the NL MVP finalists will consist of Shohei again, Pete Crow-Armstrong, and a mystery player who will probably not play a premium position. I will temper any scorn I may have for recent voting trends when it comes to Ohtani as his greatness doesn’t need empiricism to be justified, but the fact remains that PCA bests him by more than a full WAR right now. Ohtani will likely not close that gap by season’s end, yet he is still neck-and-neck with the Cubs center fielder according to the odds. The question between them is basically whether 85 innings of sub-2 ERA pitching is worth more than generational outfield defense. WAR seems to think the defense is better, by a margin that is not insignificant. Many feel the opposite, and it’s fine to do so, but is the evidence for feeling that way cohesive enough to contend with the WAR frameworks that have been sculpted to be as accurate as possible for decades? Or are the frameworks behind the times?

Share

If you’re not averse to diminishing fielding contributions, then surely you also thought Aaron Judge deserved to win over Cal Raleigh last year. Right? Or is it okay to enhance defense/position in the equation when it’s a catcher, but not when it’s an outfield player?

Judge had Raleigh beat in fWAR last season by one win and in bWAR by about 2.5. Yet, Cal was only two voters switching to him away from winning. Of course, him reaching 60 HR while Judge settled for a mere 53 was a big selling point for the Mariner, but another main one was him being a catcher. The argument was “WAR is not accounting enough for what a catcher does. He manages the game, calls pitches, and takes a physical toll no other position has to. Cal deserves it because WAR is clearly missing those things, and he was more valuable than Judge as a result.”

…Does WAR really not account for those things? Catchers actually receive the kindest positional adjustment already—better than the next-hardest positions by a half a win and better than corner outfielder by a whole two wins. The gap between Judge and Raleigh already swung in Raleigh’s favor by two wins! (It was even more than that since Judge DH’d in 56 games). Regardless of whether they were MVP candidates or replacement-level hangers-on, Raleigh was always going to have that advantage over Judge. So then the question becomes: Should the net positional difference between catcher and corner outfielder be more than two wins? Maybe! But if it is, what is the evidence for that? And can that argument then also apply to Dillon Dingler’s MVP case this season? Ah, I forgot—Dingler doesn’t have the requisite dingers.

I imagine a narrowing of positional adjustments in WAR would better reflect how we tend to perceive outfield players. However, I think such a change would also create some serious friction with how we perceive catchers.

Maybe there is evidence that Raleigh was in fact more valuable because his positional adjustment wasn’t kind enough. And maybe there is evidence that Yordan is in fact more valuable this season because Witt’s positional adjustment is too kind. My main point here is that these two beliefs do not coincide.

If we assume that catchers deserve more love from WAR, and also that corner outfielders and DHs deserve more love from WAR, then where does that leave us? If the low end and the high end get more, we’re basically giving everyone more WAR at that point. That’s fine I guess, but “when everyone’s super…”

Subscribe now

This rant was inspired by Lance Brozdowski’s recent video on WAR as it applies to this season’s MVP conversations. I agree with some of his points and disagree with some, but overall I’d be hard-pressed to not recommend his content! Check it out below:

Although some batted ball luck is probably repeatable, today we are going to dip our toes into a world where none is. If every team enjoyed equal spoils on balls in play, how might they be doing in the standings right now?

A couple weeks ago we explored FIwOBA (and GBkwERA) on the team level. Now, we will create versions of Wins Above Replacement based on those stats, then derive expected win-loss records from those (in a different manner from xW-L for pitchers).

First: position players (FIWAR). Here we take each team’s (park-adjusted) FIwOBA and calculate Batting Runs from it like so:

FI-Batting Runs = (FIwOBA - lgwOBA)/wOBAscale*(PA-IBB)

where lgwOBA (league-average wOBA) = .316, and wOBAscale (to adjust the numbers to a wins-based scale) = 1.237. These numbers come from Fangraphs’s Guts! page and are subject to slight changes daily.

We then replace standard Batting Runs with FI-Batting Runs in the WAR formula:

FIWAR = (FI-Batting Runs + Base Running Runs + Fielding Runs + Positional Adjustment + League Adjustment + Replacement Level) / (Runs per Win)

where Runs per Win = 9.825 (again from Guts!) and the rest of the components come straight from Fangraphs.

And voila. Team FIWAR:

The Cubs’ dominance in this department is subdued by FIWAR, but not enough to relinquish the top spot. Meanwhile, clubs like the Cardinals, Phillies, and Mets appear much better if we ignore their BIP outcomes. On the reverse end of the luck spectrum lay none other than the Dodgers.

Now: pitchers (gWAR). Here our approach is different from position players as we estimate pitcher WAR directly from GBkwERA:

gWAR = IP/9 * (repGBkwERA - GBkwERA)/(Runs per Win)

where repGBkwERA is based on league average ERA, K-BB%, and GB% to yield a replacement-level baseline, and Runs per Win is again 9.825.

That’s pretty much it! All that’s left is adjusting the scale of gWAR to match that of fWAR—otherwise pitching would be inaccurately represented in total WAR.

This time, the Cubs experience a much less favorable reality on the pitching side, as have the Nationals, Athletics, and Royals. Interestingly, all of the top half pitching squads by gWAR have even better fWARs, a circumstance which may have arisen from scaling gWAR down. Overall, variance in gWAR (as in FIWAR) is much smaller than in fWAR, indicating that isolated talent trends towards stability.

Share

The final step to calculating a makeshift low-variance record estimate is to combine FIWAR and gWAR like so:

FIWin% = (.294*G + FIWAR + gWAR) / G

That .294 represents Fangraphs’s replacement-level win percentage, which comes out to ~48 wins over a 162-game season. With being about three quarters of the way through the season, that replacement-level number is currently ~35-36.

And now we have FIWin%!

The best ball club by win percentage as of August 12th is the Rays, but they are merely slightly above average without batted ball variance. Instead, the Cubs reign supreme by FIWin%! They are trailed by a clustered quintet of NL teams with varying levels of batted ball misfortune. The Brewers and Dodgers have outperformed their FI-expectations, whereas the Pirates and Nationals have underperformed.

Mets fans look away (again), for your team has underperformed their fielding independent metrics the most. At least the Giants and Athletics would still be below average in an equal-luck world—the Mets would be above average! Their fierce rivals in Atlanta have experienced significantly greater fortune despite being merely a win or two better by FIWin%. The Diamondbacks are in a similar boat, which in their case has propelled them from mediocrity to a potential Wild Card finish.

Subscribe now

Think of FIWin% as “what could have happened with equal luck,” as opposed to “what should have happened.” That question is better addressed by Pythagorean and BaseRuns W-L, which do not strip batted ball variance away from the equation. Since FIWin% does, it makes sense that the difference between actual Win% and FIWin% would correlate with weighted BABIP as it does (r = .36).

Although FIWin% can’t change how the playoff race looks right now, it does potentially offer better insight than current record for how the race pans out for the rest of the season. We shall see!

Nick Martinez and Eduardo Rodríguez are each of their leagues’ poster children1 for the “they can’t keep getting away with this!” pitcher. Why do I say this? Well, we are about 70% complete with the 2026 season, and they both still sport ERAs under 3 as starters with 120+ innings. Can’t ask for much better if you’re a 1998 expansion franchise right now! But… their “under-the-hood” numbers muddy the waters:

• Martinez: 2.65 ERA | 4.01 FIP | 4.61 xFIP | 4.79 SIERA

• Rodríguez: 2.71 ERA | 3.97 FIP | 4.55 xFIP | 4.83 SIERA

Those ERAs are among the ten best in the majors among qualified pitchers. Those SIERAs are two of the three worst. The more we parse out noise (i.e., fielders behind them, home run luck) from their profiles, the worse they look. My preferred pitching metric based on K-BB% and GB% paints a similar picture.

And yet, their ERAs are better than half of that top 10. They’ve both displayed this trend all season, in fact, without really letting up.

Shrugging off strikeouts, Rodríguez and (especially) Martinez do something far braver: pitch to contact. Egads! I thought throwers pitchers weren’t allowed to do that anymore… how could this be? Despite what traditionalists might tell you, MLB hasn’t outlawed the practice. It can still be done, and is being championed this season by the old school workhouse extraordinaire himself:

Nobody is Sandy nowadays, but Nick and Eduardo get close while succeeding in preventing runs to an even greater degree. However, given what we know about their xFIP, SIERA, GBkwERA, xqzplrsmvERA2, etc… it’s only a matter of time before they regress, right? Everything points to that being the case… any day now…

Pessimists may point to their hearty LOB%s (82.2% for Eduardo, 86.3% for Nick; league average is 72.3%) and say “well, a pitcher’s ability to strand baserunners is super unreliable. Of course they’ll regress,” before taking another hit from their vape. And truthfully, I doubt I would bet on them having lower ERAs than they do now at season’s end. However, there is an element to their profiles that deserves mention, because it reveals part of why they may be more valuable than their peripherals show.

Share

Enter Grid WAR.

Another WAR, yay! We don’t have enough WARs, and we’ll never have enough.3 Grid WAR—invented by Professor Adi Wyner and Ryan Brill at UPenn—attempts to better distill the extent to which a pitcher helps his team win games. The idea behind it is that aggregating outings in one average like WAR traditionally does is flawed because it penalizes terrible outings too much, and that using context-neutral win probability added above replacement is superior.

Their paper dives deep into the math, but for an example: Suppose Pitchers A and B have five outings where A gives up 0, 0, 0, 0, & 10 runs (7 IP for the 0s, 2 IP for the 10), and B gives up 2, 2, 2, 2, & 2 runs (6 IP each). A and B are equal in that they have a 3.00 ERA over 30 IP, and will be granted 0.6-0.7 WAR across those 5 games. However, their effects on their team’s context-neutral win expectancy are decidedly unequal: By using Fangraphs’ WPA Inquirer, we can see that A will grant his team an average of 3.8 wins over his four scoreless outings (his team on average would carry a 3.5-run lead into the 8th). Meanwhile, B will grant his team an average of just 3.5 wins over his five outings (his team on average would carry a 1-run lead into the 7th). A’s 10-run disaster does not make up this difference, as a game can only be lost once. If these two pitchers repeat this pattern over a full season (33 starts), A will afford his team 2 more wins than B will his, even though by all aggregate stats they will appear identical.

Thus GWAR’s4 contention is that traditional WAR underrates streaky pitchers, and that this variance is partially a trait. GWAR has a year-over-year stickiness of r=.26 (about the same as RA9-WAR). Although fWAR has better reliability (r=.41), it doesn’t predict GWAR as well as GWAR itself does, indicating there is some value in run distribution that GWAR is reliably conveying. Specifically, the paper found that pitchers who exhibit especially high streakiness are most underrated by traditional value metrics, whereas those with especially low streakiness are most overrated.

Here are the top five pitchers by GWAR this season:

1. Jacob Misiorowski - 4.6

2. Cristopher Sánchez - 4.3

3. Eduardo Rodríguez - 4.0

4. Nick Martinez - 3.8

5. Cam Schlittler - 3.7

Unlike Mis, Cris, and Cam—who ooze dominance under the hood, over the hood, and everywhere around the hood—Nick & Ed have relied on their good starts being great while exhibiting favorable selectivity with their bad. They may not be shiny Lamborghini’s (their fastballs sit around 92-93 mph, which is a few ticks below average), but they don’t constantly sputter, either.

Let’s compare our duo of honor to newly-minted Padre Robbie Ray, who has received similar levels of fortune to his run prevention (3.08 ERA | 4.36 FIP | 4.74 xFIP | 4.76 SIERA). Ray has 2.3 GWAR, placing him just outside the top 40—a far cry from top 4. Now, I didn’t create GWAR so I can’t say for sure why Ray’s is that much lower. But I do know that Ray is less streaky:

Ray’s curve being less favorable at first glance make sense since his ERA is 11-12% worse than the other two, but that doesn’t entirely explain why his GWAR is merely 60% of Nick’s/Ed’s. Part of it is that Ray tends to have more “meh” outings. Meanwhile, Rodríguez—the GWAR leader of the bunch—has had the fewest such outings. His “boom-or-bust” rate (percentage of starts that are not “meh”) is 83%, Martinez’s is 77%, and Ray’s is 68%. In other words, Ray has been nearly twice as likely to leave his team in limbo upon exiting the game. He’d be better off concentrating some of those runs into already-bad outings so that he could allow his team more room to win the others.

The linear ERA estimators don’t see why Ray hasn’t positively impacted his team’s winning chances as much as the other two have. Grid WAR does. Nick and Eduardo have leveraged their quality distributions to their advantage this season, which has been key to ensuring their impact is fully appreciated by their teams.

Have they just been lucky? That’s certainly part of it. I would hesitate to argue that either is better than Paul Skenes just because their GWARs are double his (1.9). They have been more valuable to winning, though, and they may have more control over that than we initially thought.

The two metrics I’m interested in today are FIwOBA and wBABIP. FIwOBA is a custom formula that essentially measures the “three true outcomes” (home runs, walks, & strikeouts), with some modifiers (HBP, FB%, IFFB%, Pull%). On the other hand, wBABIP looks at the other forms of hitting productivity: singles, doubles, and triples. The difference between it and regular BABIP is that it weights each event according to its wOBA run value (1Bs are worth .889, 2Bs 1.259, & 3Bs 1.592) instead of counting them equally. Park and league adjustments have also been applied so that the stats can be seamlessly compared with not only themselves, but also the “all-in-one” wRC+. All averages are 100 as a result.

Subscribe now

By wRC+, the best lineups in baseball have belonged to the Dodgers (113), Nationals (112), and Cubs (111), with the Pirates (107) occupying fourth place. In last are the Reds (89), followed by the Rockies (90), Phillies (91), and Blue Jays (92). Eliminating high-variance batted ball events via FIwRC+ paints different pictures, however. The Pirates (112) actually lead the majors in that, with the Nationals (111) not far behind. Third place is perhaps surprisingly occupied by the Cardinals at 107 (94 wRC+), indicating a substantial gap between their fielding-independent performance and general results. Indeed, their wBABIP+ ranks dead last at 92. The Brewers (110) and Dodgers (109) have dominated in the BIP dept., followed by the Pirates (106).

The Cardinals are the most extreme example of a “good under the hood” lineup with poor batted ball fortune, but the trend is not random. In fact, the difference between wRC+ and FIwRC+ (w-FIw) correlates with wBABIP at r = .47, indicating over 20% of what FIwRC+ misses in overall production can be explained by base hit production.

With the trade deadline looming Monday, I wonder if contenders will swing for the fences on some bats. The most prestigious one to grace the rumor mill—Yordan Alvarez—would ignite any offense no matter the metric. However, with the Astros sitting at .500 in one of the weakest divisions in modern history, I anticipate him staying put.

Will the Dbacks, Guardians, or Red Sox (struggling by FIwRC+) trade for a power bat or two? Will the Cardinals or Phillies (struggling by wBABIP+) trade for a contact hitter? Should be an exciting weekend!

Minus the senseless killing and trauma and money drain and propaganda and xenophobia bad stuff, baseball operates in the same way. Now that ties are impossible, there is always a winner, and a loser. Score 15 runs and only give up 2: 1 win. Score 1 and give up 0: 1 win. The standings aren’t dictated by run differential, or wOBA, or average bat speed, or average cost of a beer at the stadium. It’s just wins and losses.

Early baseball statistics understanding the importance of this binary brought us to tracking a pitcher’s win-loss record. As the pitcher is responsible for conceding runs, the onus of the game’s result rests on his shoulders. At least, that’s the logic.

Nowadays we understand that logic to be flawed, since awarding any pitcher full credit for a result is not commensurate with his actual contributions much of the time. Sometimes pitchers are phenomenal, but receive paltry run support that results in a “no decision.” Sometimes pitchers earn decisions because they got (un)lucky from the timing of their run support. And sometimes batters are the ones who win the game for their team by producing runs when it counts.

In comes Win Probability Added (WPA) and its offshoots. WPA factors in base state, out state, the inning, and the score of a game to determine the extent to which each plate appearance affects a team’s chances of winning or losing. Since altering your team’s winning chances by 100% in either direction is impossible (there’s always a chance!), WPA can’t award a full win or loss to one player—though some walk-off home runs have gotten close. There is also a version that normalizes leverage (WPA/LI), which aims to isolate contributions regardless of the intensity of the moment.

So, since we can more precisely deduce win contributions through WPA and WPA/LI than we can W-L record, what if we created more accurate records with those stats?

Expected Record & Gifted Wins

A similar approach to this was done by Eric Sammons over at the Community Blog at Fangraphs, which I highly recommend. Sammons calculated WPA-Wins and WPA-Losses on a game-by-game basis, whereas I will just estimate career W-L records. The gist is the same: Award pitchers not by the arbitrary constructs of a “win” or “loss,” but by their actual contributions.

Truth be told, my inspiration came from NL relievers Aaron Ashby (12-2 thru 59.2 IP this season) and Adrian Morejon (20-8 thru 127.2 IP this season and last). I thought, “Wow, those guys sure accrue a lot of wins in the little time they have to pitch!” And they really do. Of the 3,878 pitchers to ever throw at least 250 innings in major league history, Ashby ranks just outside the top 50 in W/9 (wins per 9 innings) with 0.71. And Morejon? Well, he’s at the very top (1.02). Devin Williams is second with 0.92, and third is Monarchs legend José Méndez with 0.84. Again, these are career numbers all-time, from 1871 to now. Morejon is the king of earning wins efficiently.

Being the party pooper I am, I sought to uncover just how fortunate Ashby and Morejon have been to earn such flattering records. They are of course great pitchers, but would they have as many opportunities to earn wins if they were closers, a role in which it’s much harder to earn wins? Never mind if they were starters.

To calculate expected record (xW-L), we take a pitcher’s WPA and divide by games. Adrian may scoff at this approach, but it helps balance workload differences and stays truer to the interpretation of WPA. Morejon has garnered a little over 5 WPA in his career, and while this proves unparalleled on an inning basis, he’s not actually winning games at that rate, because he averages just a little over an inning per appearance. In reality, his presence has contributed an average of +2.3% win probability for his teams per game, meaning the Padres have been 2.3% more likely to win when he pitches than when he doesn’t. Theoretically, then, an average team with average circumstances would grant Morejon a winning percentage of .500 + .023 = .523. If we assume 43 decisions from his actual career record of 30-132, then his expected win total would be .523 × 43 = 22, for an expected record of 22-21. Given the circumstances he’s been dealt (i.e., run support timing, managerial choices, sequencing, etc.), Morejon has been gifted 8 wins to his actual record; his winning percentage is a whopping 17.5% better than expected. Using WPA/LI instead also grants a 22-21 xW-L.

WPA heavily regressing expected record to .500 happens frequently with relievers, and it persists for starters, too. Some, however, experience the opposite phenomenon. Let’s take Paul Skenes as an example. Skenes happens to have both the best xWin% (.631) and xWin%/LI (.629) among all pitchers with at least a hundred innings pitched since 1974 (when Fangraphs first starts tracking win probability). He is expected to be 32-19, but his actual record is just 30-21, so Paul has been deprived of 2 wins from his record already (Thank you, Mr. Nutting!). Just two wins “lost” for Skenes may not seem like much, but it is arguably worse than two wins “won” would be better, since the binary W-L system allows for much more positive variance for great pitchers.3

Share

Now for the reveal. Who has been gifted the most wins to their record over the last half century, and who has been robbed of the most?

Feel free to sort the below table however you like, shuffle page by page, and use the search box to find names that come to mind.

By both standard and leverage-adjusted expected records, one man has benefited more than anyone.

I love when new stats reinforce my priors, and here is no exception. Andy Pettitte has his historic Yankee lineups to thank for much of his record (his Astros weren’t too shabby either), as 35 (33 without leverage) of his 256 wins were gifted by some of the best run support money can buy. Pettitte was still a fantastic pitcher—as evidenced by his persistent presence on recent Hall of Fame ballots—even if the 250-win argument in his favor rings a tad hollow in light of this.

Similar spoils were enjoyed by Clayton Kershaw, Pedro Martínez, and Roger Clemens, whose top 10 xWin%s didn’t need the fluff. David Price also benefited immensely, going from an xW-L of 131-108 to a dominant-looking 157-82. Jon Lester sports the second-largest leverage-adjusted GiftedWs with 31, going from a solid 169-148 xW-L/LI to a much cleaner 200-117. I look forward to his Cooperstown ballot debut later this year. A special mention should be made for Chien-Ming Wang as well: The former Yankee (a familiar pattern at the top of this list) finished his career at 68-34 despite his expected records being just a couple games above even. And if you’re looking for another reliever like Adrian Morejon, look no further than Brendan Donnelly (32-10 W-L, 22-20 xW-L).

José DeLeón and Rick Honeycutt share the worst Gifted Wins total at -19—brutal in light of their expected records being above .500! Around here we also find some relievers who were closers for awhile, like José Mesa, Lee Smith, Fernando Rodney, and Trevor Hoffman. Nobody really ever took W-L records for closers seriously before, and this offers all the more rationale for that. As for starters, notable modern names include Kyle Freeland, Matt Cain, Jeff Samardzija, Chris Archer, and Patrick Corbin. Cain is a little surprising to see here since the career Giant enjoyed multiple World Series rings. He finished at 104-118, but if he’d received run support as good as his was bad, he could’ve been 130-92!

The only winless pitcher among the thousands in that table is Terry Felton (0-16). Poor Terry deserved much better than that (7 or 8 wins). Anthony Young was arguably subjected to an even worse reality (15-48 W-L) compared to expectations (29-34 xW-L, 31-32 xW-L/LI); his career ERA was league average!

The “ClutchW” column is basically the Clutch stat applied to decision opportunities. It is just leverage-adjusted Gifted Wins minus Gifted Wins. “The Human Rain Delay” Steve Trachsel has the most Clutch Wins in this sample with 7. In a vacuum, Trachsel was a 145-157 pitcher (that is his xW-L/LI), but since he performed better in more important moments, he was expected to be a 152-150 pitcher (xW-L). All told though, Trachsel ended up with a 143-159 W-L due to subpar circumstances.

The only other pitchers with at least 5 Clutch Wins are Vida Blue, Aaron Harang, and Jered Weaver. Weaver provides an interesting comparison point since he benefited substantially from circumstance (+14 GiftedW, +19 GiftedW/LI). A 131-117 pitcher in isolation (xW-L/LI), Weaver was both clutch and blessed (136-112 xW-L, 150-98 W-L). Contrast that with Rick Porcello, another pitcher who finished his career with 150 wins (150-125 W-L). Porcello was not clutch (-4 ClutchW), but he was blessed so much with run support that it didn’t matter (+15 GiftedW). Rick Langford has the worst ClutchW among pitchers whose careers are fully accounted for (Steve Rogers and Nolan Ryan had even less but played parts of their careers before ‘74). Langford’s record experiences more misfortune than most, as he was neither clutch (-5 ClutchW) nor blessed (-11 GiftedW). Perhaps in another life he’s a .500 pitcher, but instead he must settle for 73-106.

The vast majority of pitchers don’t have any more than a couple Clutch Wins in either direction, given how poorly repeatable “being clutch” is.

Subscribe now

Conclusion

The moral of this story is that “Did you win or did you lose?” is a fair enough question on the team level, but tends to be either too compensatory or too punitive when applied to pitchers. Expected records contextualize individual contributions better, and aid in understanding how dependent actual records are on circumstances beyond a pitcher’s control. If a pitcher is getting shorted by his record, that usually means he either pitches in a lot of save situations or plays for a mediocre franchise.

Analytics have come a long way from pitcher wins and losses impacting perceptions, reputations, awards, and salaries, like they did for over a century. Tangible effects on winning still matter though—not just for isolating talent, but for telling the story of the game. I hope Expected Records can help do that.

The mound being lowered was a seismic shift to the pitching meta, but it probably wasn’t the most dramatic. For that one, we have to travel several more decades into the past for a much more severe change—before there was even a “mound” at all.

When the rubber met the road

Picture yourself in 1892. Professional baseball has been going on for some years, but it looks quite different. Pitchers started throwing overhand just a handful of years ago. Batters are still adjusting to the fact that they can’t call for how high they want their pitches anymore, and can only take four balls before getting walked now. Foul balls are still a decade away from counting as strikes. And, the pitcher releases the ball 55.5 feet from home plate, level with the batter.

Seeing as many of these rules favored pitchers, National League leadership sought to make throwing to the catcher more strenuous—by five feet, to be exact. This change proved to alter the landscape of the game, arguably more than any since (and as we’ll see soon, much more than the mound lowering after 1968).

Truth be told, as someone who is not the most adept baseball historian, I discovered this change when I was browsing year-over-year league stats and noticed that the highest-ever league OPS belonged to the 1894 season, at a whopping .814. This is the only season to ever sport an OPS over .800, besting by a considerable margin the “Year of the Hitter” of 1930 and the steroid-infused seasons around the turn of the 21st century. As it turns out, the pitcher being shifted back was the culprit. The first season with the change was actually 1893, but the jump in offense continued into the following season, indicating additional growing pains. Offense eventually came back down to Earth, however, once foul balls were counted as strikes starting in the early 1900s, launching the dead-ball era.

I don’t want to imagine how hitters would be faring nowadays if the mound were five feet closer!

Did any pitchers really improve?

A total of zero pitchers with at least 100 innings pitched in both 1892 and 1893 posted a better FIP in the second season. Although it’s true that FIP was not designed to reflect the tendencies of the game before the live-ball era, this fact is still alarming for the hurlers. At least a decent handful after the “Year of the Pitcher” (1968) improved their FIPs in an absolute sense. No one from 1892-93 can say as much, and in fact, all of them worsened by at least three-quarters of a point.

Nevertheless, the BIP-centricity1 of the game during that time requires us to place more relative trust in ERA/RA9. Unfortunately for pitchers, this view is no more flattering given the jump in average ERA from 3.28 to 4.66. Surely everyone struggled to prevent runs compared to the year before then, right?

Right?

Wrong!

Share

The Man of the Hour

Unlike 1968-69, the “Most Improved Pitcher” award for 1893 requires no nomination process. Theodore Breitenstein of the St. Louis Browns easily takes the crown.

Succeeding in relation to his peers from a young age was not a foreign concept to “Breit” (pronounced like “bright”), as he was the first to ever throw a no-hitter in his first major league start2. Impressive, though it could not save the team from coming into the 1892 season rougher than sandpaper. The Browns finished 56-94-5 that year, with Breit going 9-19 alongside a 4.69 ERA. Tough goings for the youngster, but his roughly league-average 3.55 FIP showed promise.

That would end up being his only full season with the 50-ft “box” as it were, but unlike the rest of his peers, Breit seemed to embrace the challenge. Upping his workload from 282.1 to 382.2 innings in his age-24 season was prudent in light of his ERA shrinking to 3.18 en route to an ERA title. As you may tell from the table above, this difference was nothing like what every other pitcher experienced with the 60.5 ft. The only other one to improve their ERA in 1893 was Brickyard Kennedy, who would take home the “Coolest Name” award if not for Ice Box Chamberlain. Kennedy’s FIP progression was middling, however, whereas Breit’s was still among the best.

Was Breitenstein the best that season? Him, Kid Nichols, Cy Young, and Amos Rusie all hovered around 11-12 bWAR, so he may not have been. But his improvement compared to the year before far outshone the other three, who enjoyed win totals in the 33-34 range versus Theo’s 19.

How did he do it? The answer is… unknown. I highly doubt it was pure luck, considering the Browns offered worse run support than any other team would have. At the same time, it’s hard to find much about the process that led to his one-of-a-kind improvement—probably because it was over 130 years ago. We do have this quote from Hall of Fame outfielder Willie Keeler to give us a sense of Breit’s style:

Breitenstein is one of the few left-handers who can locate the plate when he wants to, and in addition to this he has terrific speed, sharp curves, and there is not a pitcher in the league that fields his position better. — Wee Willie Keeler, 1897

Sounds like he had it all! It’s a shame the Browns were so bad throughout the 1890s and that his owner overworked him as a result. Breit would be dealt to Cincinnati following the 1896 season, where he enjoyed significantly better run support and kinder workloads. St. Louis became even worse in his absence.

You may be wondering if Breit kept it up in 1894, when offense reached even greater heights. His 4.79 ERA and 5.15 FIP that year were decidedly a step down, though still better than league average. Him leading the NL in innings pitched (447.1 IP) as a result of his owner relying on him so much also may have fatigued him. His bWAR didn’t decrease that much in 1894 (11.5 → 9.3), nor did his fWAR (4.8 → 4.8). Nolan Ryan’s pre-incarnation Amos Rusie took the spotlight that year though, winning the Triple Crown with an outlandishly low 2.78 ERA3 (53 ERA-), 195 Ks, and 36 wins (a number Theo couldn’t have dreamt of in the Lou).4

Perhaps due to being mismanaged, Breit was an early proponent of pitchers pacing themselves and outwitting hitters. He once stated, “Never pitch for strike-outs. Their day is over. Always remember that you have eight men to help you.” Definitely a man of his time—though Rusie likely disagreed vehemently.

Subscribe now

Theodore pitched for the minor league New Orleans Pelicans through his thirties—despite receiving several major league offers throughout—largely because it paid better.5 He even threw a no-hitter at 40 years old. By the end of his tenure, folks were starting to pay greater attention to a young phenom nicknamed “Shoeless,” but that didn’t stop fans from revering Breit as the Southern Association sage he was:

Breit’s Wikipedia page fails to emphasize how much of an outlier his 1893 season was in relation to his peers. Perhaps it should. So here’s to you, Theo, for doing what no one else could at the time, and hardly getting recognized for it.

Special thanks to Stephen V. Rice’s SABR biography on Breitenstein for much of the information used in this article.

Coupled with a widened strike zone implemented earlier in the decade, the batter-pitcher dynamic had reached an uncomfortable tipping point. In response, MLB instituted a change which remains today: The mound lowered from 15 inches high to 10. The idea was to diminish the pitcher’s height leverage so that hitters could attack pitches from a more level angle. This immediately accomplished its desired effect as runs per game shot back up above 4.00 in 1969. Even though the strike zone change seemed to be the greater culprit for the increased pitcher effectiveness, the mound was put on the chopping block nonetheless. Seeing as how much faster pitching development continues to improve over hitting nowadays (and the ever-flowing hitter-friendly rule changes in response), the mound being lowered was probably a wise decision.

Naturally, displeasure from pitchers at the time surfaced from both their words and their play. The change in league average ERA from 2.98 to 3.61 was stark, so posting worse stats in 1969 was expected. However, some pitchers raged against the dying of the light mound, by pitching even better.

Improving in an absolute sense

The stats we’ll analyze today are (from most to least complex): ERA, FIP, fWAR, and RA9-WAR. RA9-WAR is Fangraphs’ attempt at accounting for balls in play and baserunners in addition to its FIP-based WAR, so in that sense it pairs nicely with ERA. These are much noisier stats, however, so we will focus more on FIP/fWAR.

The table below highlights these four stats for all 85 pitchers with at least 100 innings in both 1968 and 1969, with the WARs adjusted to a rate basis (per 250 IP) to isolate average instead of compiled performance.

An important note about the WARs: Its average change from one year to the next will always be 0, since it is adjusted to replacement level for each year and allocates a set amount to all pitchers (430 for fWAR). The midpoint of this list displaying WAR/250IPs that are mostly unchanged bears this out. The relative nature of WAR is therefore a bit ineffective for our purposes.

Take Mike Kekich for example. The young yo-yo starter experienced an ERA rise of 0.63 and a FIP rise of 0.90 between his time with the Dodgers in ‘68 to the start of his stint with the Yankees in ‘69. These changes were pretty much in line with expectation. However, his WAR/250IP improved by 0.8 in ‘69, and his RA9-WAR/250IP by 0.4. Kekich did not improve in an absolute sense, but he was actually better in relation to his peers (his ERA- and FIP- both decreased, to his benefit).

I am not interested in pitchers like Kekich. I am interested in pitchers who got so much better after the Year of the Pitcher that they actually gave up less damage despite the more hostile environment.

Let’s work our way down the list.

Share

Larry Dierker

First off, Dierker debuted for the Astros on his 18th birthday (!). The eventual Houston manager of five years was perhaps destined for a solid career after striking out Willie Mays in his first major league inning. Through his age-21 season in 1968, Dierker showed some star potential. His age-22 ‘69 campaign was the best of his career (7.3 WAR over 305.1 IP), and was followed by a few 3-4 WAR seasons until he eventually reached a soft denouement once he turned 30. His improvement may have been the result of natural career progression, but it does make you think about how much better his ‘69 season could have been with five more inches of mound height.

Bob Veale

Here lies our most unfortunate pitcher who improved in an absolute sense, as Veale’s ERA shifted unkindly in ‘69. This might be more expected than usual though, since his ERA in ‘68 was 2.05 (3.23 the following season was still much better than average). At 33 years old, Veale enjoyed his last couple of years as an effective starter in spite of the mound shrinkage, a trend which does not appear to have been the result of simple career progression.

Jim Bunning

The former perfect-game-thrower and Congressman shares a similar story with Veale as an experienced pitcher who caught a second wind. Bunning was just as good relative to his peers in 1970 as well, indicating the initial mound shrinkage season was not a fluke. However, a more macroscopic view of his career reveals that the Year of the Pitcher was actually the Year of Not Jim Bunning in Particular, as the eventual Hall of Famer was posting stellar numbers throughout the decade up until a relatively pedestrian ‘68. Perhaps we should chalk this one up as merely bouncing back from a down year.

Dave Boswell

The young Twin lowered his FIP from 3.66 in the Year of the Pitcher to 3.19 in the next. Consistency was the name of his game throughout the late 60s, but he unfortunately suffered a career-ending shoulder injury in the ‘69 ALCS in an otherwise brilliantly pitched game. We are left to wonder how Boswell would have continued to prove his worth on lowered mounds. Instead, he was out of the league by age 26.

Joe Niekro

The mound transition preceded the younger Niekro’s steady development as an effective knuckleballer by several seasons, but we still saw him noticeably improve in ‘69 thanks to what would end up being the second-lowest BB/9 of his career. Just don’t look at his win-loss record that year!

Dick Bosman

The single biggest beneficiary of the whole list by ERA, Bosman’s improvement wasn’t all luck. Leading MLB in ERA in ‘69 turned out not to be a repeatable feat, but he did maintain great production for a few more seasons. Bosman’s RA9-WAR/250IP boost of +7.7 puts everyone else’s to shame.

Jim Lonborg

The 1967 Cy Young Award winner lost a lot of steam in the Year of the Pitcher, posting only 0.2 WAR in just 113.1 IP. The following season began Lonborg’s decade-long stretch as a solid league-average pitcher, which makes me think the mound lowering wouldn’t have affected his trajectory either way.

Dock Ellis

Dock will be the last pitcher discussed since venturing further down the list leaves us with pitchers within a marginal .1 FIP difference between seasons. Ellis is also the last one remaining with a WAR/250IP difference of over 3. These two seasons represent his rookie and sophomore campaigns, the former of which saw him barely reach the 100-inning threshold. Despite his ERA ballooning over a point in ‘69, Dock was unquestionably better that year, evidenced by his K/9 jumping from 4.50 to a career-best 7.12. His BABIP-against being a measly .236—well below the already historically low .269 average for ‘68—explains why his 2.51 ERA that year was not sustainable. Regarding how the mound change affected Ellis, it feels like we don’t have enough to go off of here. The famous LSD-infused no-hitter came in ‘70.

Subscribe now

And the Most Improved Pitcher Award goes to…

The closest thing MLB has to a Most Improved Player Award is the Comeback Player of the Year Award (created twenty years ago on the back of a Viagra sponsorship). If they had this award at the time for pitchers though, who do you think would win it? Especially since this was the hardest season in which to improve in the live-ball era?

It’d be hard to argue against Dick Bosman, whose ERA improvement trumped everyone else’s at a time when ERA was one of our only options to judge pitchers. Rick Wise’s 1.32-difference would also entice voters back then. By the same token, Bob Veale would be a much harder sell, even though his under-the-hood improvement was arguably more genuine.

If the award were truly for a “Comeback Pitcher,” Bunning’s case is strong in light of his older age and his ‘68 season being such an uncharacteristic blip. If MLB were to follow the NBA’s model for MIPs—i.e., favoring young players finally blossoming—Larry Dierker fits that mold brilliantly. Despite benefiting from more BABIP luck than Dave Boswell, Dierker would probably receive my vote over Bosman due to having pitched many more innings (233.2 vs 139.0 in ‘68 & 305.1 vs 193.0 in ‘69).

Singles are nice, but doubles and triples are nicer because they earn more bases and score more runs on average. Barring speed, those are earned as a result of hitting the ball harder than the average single. Home runs are even better and usually involve the ball being hit even harder than that. Unless the pitcher can’t throw strikes, that pretty much sums up the hitter’s itinerary: Get on base, but especially get on base by hitting the ball hard.

Scraping dozens of statistics from the lovely Fangraphs.com allows us to see just how much each one predicts a hitter’s future success, defined here as xwOBA. The table below highlights to what extent each stat among hitters with at least 250 plate appearances in 2024 correlates with that hitter’s xwOBA in 2025 (left-most number). The next column to the right shows how stable that stat correlates with itself in the following season, and the right-most column simply multiplies the absolute values of those two for the purposes of balancing out the list.

You may notice a pattern:

EV (exit velocity) sits at the summit of the intersection between “predicts future success” and “is reliable.” For a decade now, the flagship Statcast metric has reflected exactly how fast the ball launches off the bat—i.e., how hard the ball is hit. After EV is is HardHit%, which, yeah. Next is EV90, which is just 90th percentile EV (“How hard do they hit the ball when they really hit it?”). Hopping over xwOBA (which is largely calculated by exit velocity and launch angle) takes us to BlastCon%, Barrel%, BlastSw%, maxEV, Hard%, and BatSpd. By now you are getting the point.

If you want to succeed in Major League Baseball, the most effective goal you can have is to make the ball come off your bat as ferociously as possible. Although lifting the ball is important, accomplish alone it cannot. If you’re talented at elevating the ball, you best not do so lazily; if you’re not, you better get it out of the infield fast. Having power is so important that even striking out is slightly positively correlated with higher future xwOBA (r=.049), because doing so confounds so heavily with slugging that its harm is sufficiently accounted for.

Most of this knowledge doesn’t require fancy acronyms to understand. Ted Williams always belabored the point that hitting the ball hard in the air is the most essential goal for a hitter. Statcast or not, we always knew this. And so, it’s interesting that a useful stat can be invented that completely ignores it.

Share

Barry Bonds—you may have heard of him—tended to hit the ball hard. Barry also walked, a lot. He earned more bases on balls in his 22 seasons than anyone in history (2,558) and was intentionally walked 688 times—more than double second place Albert Pujols (316). These facts are all related, as not only was Bonds incredible at hitting the ball hard, but his combination of plate discipline rendered him so feared that teams simply refused to try against him. Six hundred eighty-eight times, they chose to give him first base for free (and many of those 2,558 - 688 = 1,870 uBBs were “unintentional intentional” walks).

If you’re not familiar with Jon Bois, I highly recommend becoming familiar. If you are, you might have seen his video from several years ago titled “What if Barry Bonds had played without a baseball bat?”

It’s a cool exercise that blends humor with reality very well, and it was only possible because of how monstrous those steroid-infused seasons in the early noughties were for the Giant. No one before or since has ever been as intimidating from the batter’s box as practically as Bonds was. Even if teams were probably a bit overzealous in their propensity to intentionally walk him, it stands to reason that both his intentional and unintentional walk rates communicated a great deal about his production by themselves.

Let’s return to the xwOBA correlations now, narrowing in on just box score stats for simplicity’s sake:

If we were inspecting these numbers fifty years ago without the gift of advanced tracking analytics, we would still come to the same conclusions about hitting the ball hard from seeing that home run percentage is most positively correlated with future success as well as being fairly stable year-over-year. At that time we would be much more shocked by the fact that walks of both forms were that indicative of success. They are arguably in the same tier as HR%.

We know this better nowadays thanks to Moneyball, and we’ve always implicitly known hitters who get intentionally walked tend to be great, even if the stat is highly circumstantial. The 2026 IBB leaderboard is currently headed by none other than Yordan Alvarez, who has been unquestionably the most feared hitter in the majors this season by whatever metric you pick. He is trailed by Ohtani and Kurtz (9), Trout and Wood (8), Soto (7), and Caminero (6)—all intimidating presences. For being completely ignored in the wOBA calculation, it sure tends to find the cream.

The circumstantial aspect of IBBs should be noted. For starters, non-DH leagues used to feature 8-hole hitters being intentionally walked for the sake of getting to the pitcher; such shenanigans have since been outlawed. Elite hitters being featured in otherwise meek lineups also tend to experience more IBBs. A similar correlational analysis conducted on just players who switched teams may diminish the perceived impact of intentional passes. Even then, they would still probably predict future success more reliably than singles rate (which actually has a negative r!), doubles rate, and triples rate, which depend on the uncertainty of the dreaded ball in play.

As it turns out, we can get a pretty decent idea of a hitter’s talent just by looking at these walk rates—just by all those times when they stepped up to the plate and didn’t even have to attempt to hit the ball hard. Assumptions about their ability to do that are already baked into their intentional walk rate.

By running linear regression analyses on IBB% and uBB%, we can see how well just those two variables are able to reflect overall production. As it turns out, a combined walk rate can explain 60-65% of the variance in a hitter’s xwOBA if intentional walks are valued at several times that of unintentional. I’m calling this Weighted Walk Rate (wBB%):

wBB% = (uBB + 9×IBB) / PA

Some regressions on future xwOBA actually recommend valuing IBBs as much as 12 times more than uBBs, whereas current xwOBA analyses suggested up to 10 but as few as 7. I feel like 9 is a fair compromise given their rare and circumstantial nature.

What does the wBB% leaderboard look like a little over halfway through the season? Here are all qualified hitters through July 1:

Subscribe now

Not bad for excluding batted balls entirely! Some great hitters occupy the very bottom of this list though (namely Michael Harris II and Otto Lopez), so I guess knowing how a batter is doing when he actually swings is worthwhile after all.

Overall, I think the intentional walk being passed off as something that shouldn’t affect a player’s stats has discouraged digging into the matter further. In actuality, it is among the most predictive and reliable box score numbers despite its rarity. Weighted walk percentage may not be as dependable as wOBA for predicting future production, but it 1) is more stable than wOBA, and 2) doesn’t explicitly involve swinging a baseball bat. Given those facts, I’d say it does a fine job.

And since you’re probably wondering, Barry Bonds’ wBB% in 2004 was ~200%.

I don’t remember what it was called, but I do know there were four teams: red, yellow, green, and blue. And I loved the green team. So much so that when it came time to vote online for which team did the best last episode, I spammed green. Click green team, submit vote, page refresh, repeat. On and on and on. Lo and behold the next week, green team won. “It worked!” I’d determined. So I did the same thing for the week after, but when it came time for the results to be announced, the show heinously, insultingly declared one of the other teams victorious. I was distraught, but I didn’t carry my verde fervor with me too long after that. After all, blue was cool, too.

Thankfully, that setting didn’t require any sort of maturity or discernment in the voting process. Some may feel the same way about MLB All-Star voting, but I can’t bring myself to such whimsy. (Maybe that means I’ve soured too much into adulthood.) I believe in rewarding All-Star appearances meritocratically, despite the inevitability of big-market teams enjoying more representation by virtue of having more fans. Does that make me a curmudgeon? I don’t dispute that fan voting reflects a different kind of truth unique to fandom, but I also feel for those players who perform at an All-Star level but who are not rewarded due to circumstances out of their control.

The first round of voting closed on Thursday of this week and the results have been revealed.

Our AL finalists are:

• C: Shea Langeliers (ATH), Alejandro Kirk (TOR)

• 1B: Vladimir Guerrero Jr. (TOR), Ben Rice (NYY)

• 2B: Ernie Clement (TOR)

• 3B: Kazuma Okamoto (TOR), Junior Caminero (TB)

• SS: Bobby Witt Jr. (KC), Andrés Giménez (TOR)

• OF: Aaron Judge (NYY), Mike Trout (LAA), Byron Buxton (MIN), Cody Bellinger (NYY), Daulton Varsho (TOR), Jesús Sánchez (TOR)

• DH: Yordan Alvarez (HOU), George Springer (TOR)

Like most fans last October, I was rooting for the scrappy Blue Jays to overcome the tyrannical Dodgers in the World Series. Little did I know their phenomenal run would fuel such a Canadian wave! The Torontonian enthusiasm, however, is not commensurate with their achievements. And this is different from nerds losing their noses in spreadsheets to declare that actually This Guy deserves an All-Star over That Guy because This Guy ackshually has 0.2 more qzjWAR than That Guy, acqkshewuhlee. No, this is a matter of Alejandro Kirk playing less than 15 games this season and still finishing in clear second place for catchers. Mysterious.

Blue Jays are ever-present in the All-Star field. The others’ cases are not as egregious as Kirk’s, but rest assured the lineup for the <.500 team is likely not entirely composed of bona fide All-Stars.

Our NL finalists are:

• C: Drake Baldwin (ATL), Will Smith (LAD)

• 1B: Freddie Freeman (LAD), Matt Olson (ATL)

• 2B: Ozzie Albies (ATL), Bryson Stott (PHI)

• 3B: Max Muncy (LAD), Alec Bohm (PHI)

• SS: CJ Abrams (WAS), Mookie Betts (LAD)

• OF: Andy Pages (LAD), Brandon Marsh (PHI), Ronald Acuña Jr. (ATL), Teoscar Hernández (LAD), Juan Soto (NYM), Michael Harris II (ATL)

• DH: Shohei Ohtani (LAD)

The NL is composed of the Dodgers, the Braves, the Phillies, the… Wait, are there other teams in the NL? Many voters seem to be unaware. Those three squads are legit this season, but as we’ll discover soon, the other twelve teams in their league house plenty of deserving talent.

Honestly, teams made up of players who haven’t been voted in might be better than those with players who have. Let’s find out if that’s true.

AL All-Snubs

Catcher: Dillon Dingler (3rd)

Voted in: Shea Langeliers, Alejandro Kirk

We’ve already discussed the Kirk absurdity, but what might explain Langeliers earning triple the votes of Dingler? Is it because “if it Hangeliers, we Bangeliers?” Dillon has a fun-loving baseball name too, you know. Is it the fact that Shea started the season hot, with Dingler suffering from overtaking him too late in the voting process? That’s more likely. Langeliers had him beat on offense through the first quarter of the season, but now Dingler has overtaken him on that front (137 wRC+ vs 128) while maintaining superior Fielding Run Value (11.7 Def vs 0.4). The gap between them has just kept growing (3.7 fWAR vs 2.3).

First Baseman: Nick Kurtz (4th)

Voted in: Vladimir Guerrero Jr., Ben Rice

Vladdy Jr. would’ve been a great prediction given his expected production before the season started, but it turns out that his production was never going to matter. Despite being a league average hitter (103 wRC+) with an uncharacteristically meager 4 home runs to his name and only about 1 WAR, Guerrero Jr. leads the voting. Ben Rice of the Yankees is a much more defensible pick (166 wRC+, 2.7 fWAR), but just as good has been Nick Kurtz (164 wRC+, 3.3 fWAR) of the Athletics. I guess Kurtz should have considered getting drafted by the Yankees. Skill issue. Had Munetaka Murakami not pulled his hamstring last month, the AL first base race could have been even spicier.

Second Baseman: Travis Bazzana (6th)

Voted in: Ernie Clement

Admittedly, second base on the AL side felt particularly weak this year. Even so, Clement earning more votes than everyone else in his league doesn’t exactly jive with his production (107 wRC+, 1.1 fWAR). My gaze initially fell upon Jazz Chisholm Jr. (102 wRC+, 2.0 fWAR), who leads his peers in fWAR despite his lackluster start. Cleveland rookie Travis Bazzana (125 wRC+, 1.4 fWAR) is the more exciting pick and just as defensible despite debuting in late April. Bazzana doesn’t jump off the page in comparison to Jazz context-neutrally, but with context we see that Travis has stepped up at the plate in much more impactful ways than the inconsistent Yankee (10.8 RE24 vs -1.7). Ezequiel Durán offers a similar argument as Bazzana, but… is he really a second baseman?

Third Baseman: Miguel Vargas (3rd)

Voted in: Kazuma Okamoto, Junior Caminero

Okamoto has gotten off to a great rookie campaign so far (117 wRC+, 1.7 fWAR), but he’s just another Blue Jay case. Then Caminero is an exciting Ray with a ferocious bat (136 wRC+, 1.8 fWAR), but Miguel Vargas (129 wRC+, 2.4 fWAR) has been arguably as productive at the plate while not struggling nearly as much defensively (-0.7 Def vs -7.0). The White Sox aren’t even a “smaller market” than Tampa Bay and have been electric in the Central, so them not getting much love from voters was surprising.

Shortstop: Kevin McGonigle (3rd)

Voted in: Bobby Witt Jr., Andrés Giménez

This is our first instance of voters getting it right as Witt has been far and away the best shortstop in the game this season. But if I had to go with someone else, I would take Kevin McGonigle (132 wRC+, 3.3 fWAR). I adore McGonigle’s game and always believed he would shine bright as one of the crown jewels of his prospect class. Deciding against Colson Montgomery (118 wRC+, 2.6 fWAR) was a hard decision though, what with his 20 home runs compared to McGonigle’s 6. I do worry Kevin might regress more than Colson in the second half, but for now, give me the Tiger.

Designated Hitter: Yandy Díaz (3rd)

Voted in: Yordan Alvarez, George Springer

I won’t pretend for a second Yordan didn’t deserve first honors here. Second place, however, was clearly Díaz (157 wRC+, 2.6 fWAR). The dude mashes baseballs for the Rays day in and day out. The Blue Jay brigade reveals perhaps its worst non-Kirk iteration here, as Springer is a below league average hitter—an undesirable trait for a DH. Seattle’s Dominic Canzone (152 wRC+, 1.4 fWAR) deserved more love instead.

Outfield: Jac Caglianone (20th), Ceddanne Rafaela (not top 20), Randy Arozarena (9th)

Voted in: Aaron Judge, Mike Trout, Byron Buxton, Cody Bellinger, Daulton Varsho, Jesús Sánchez

This one involves more grasping at straws as most of the statistically deserving candidates earned honors despite Blue Jay fans’ best efforts. I’m shocked Rafaela (114 wRC+, 2.5 fWAR) wasn’t even top 20, though as we’ll see later, voters care far less about outfield defense than WAR does. The comparatively heftier support for Arozarena (134 wRC+, 2.2 fWAR) makes sense in light of his respectable 11.5 RE24. Lastly, Caglianone (129 wRC+, 1.7 fWAR) has bounced back in a huge way from his bumpy debut season and would be a great story. There are truthfully many potential answers for this All-Snub outfield, unlike for the NL…

Share

NL All-Snubs

Catcher: Gabriel Moreno (8th)

Voted in: Drake Baldwin, Will Smith

For the longest time I thought I would choose the Marlins’ Liam Hicks (128 wRC+, 1.6 fWAR) as the sneaky catcher option in the face of the obvious Baldwin. Moreno (124 wRC+, 1.7 fWAR), meanwhile, was off my radar due to missing most of April—a death sentence for any All-Star hopes. But in just a little over 200 plate appearances, the Diamondback has impressed in all facets, and his superiority on a rate basis puts him over Hicks and William Contreras for me.

First Baseman: Alec Burleson (5th)

Voted in: Freddie Freeman, Matt Olson

The electorate didn’t whiff here, but Burleson (134 wRC+, 2.0 fWAR) would’ve been a respectable All-Star. The Cardinal is second only to Bryce Harper in xwOBA among qualified NL first basemen, indicating he could very well keep crushing alongside all these future Hall of Famers into the second half.

Second Baseman: JJ Wetherholt (8th)

Voted in: Ozzie Albies, Bryson Stott

Now this was a whiff. This one ranks atop Statcast’s new “miss distance” leaderboards. At least four second basemen would have been wiser choices than Albies and (especially) Stott here, but such is big-market life. Brice Turang (131 wRC+, 2.8 fWAR) is the most popular stat-savvy choice and is perfectly defensible, as are Brandon Lowe, Ketel Marte, Luis Arraez, and Xavier Edwards. I am most partial to the rookie Cardinal JJ Wetherholt (121 wRC+, 3.3 fWAR) though, due to his superior fielding and my belief that his offense is less fluky than Turang’s (.282 BABIP vs .342). The only part that surprises me about being so staunchly Team Wetherholt right now is how I used to not be, instead favoring Nolan McLean and Sal Stewart for NL Rookie of the Year. Whoops!

Third Baseman: Curtis Mead (10th)

Voted in: Max Muncy, Alec Bohm

Mead (121 wRC+, 0.7 fWAR) is probably the biggest reach for a “snub” in this whole exercise, since LAD’s Max Muncy (143 wRC+, 2.9 fWAR) is in a tier of his own. I wonder if Max would’ve also blundered his queen against an International Chess Master.

Shortstop: Otto Lopez (5th)

Voted in: CJ Abrams, Mookie Betts

I think Elly De La Cruz (128 wRC+, 2.5 fWAR) would have had the best odds at this nod had he not missed a few weeks with a hamstring injury (please stretch!), which opened the door for Otto Lopez (136 wRC+, 3.4 fWAR) to flaunt his arguably unmatched contact hitting prowess. The .340-hitting phenom is an unfortunate small-market snub who certainly deserved more representation than Betts (95 wRC+, 1.0 fWAR). I love seeing CJ Abrams (148 wRC+, 2.7 fWAR) receive such glowing support, but triple the votes of Lopez seems a bit off kilter. What else is Otto supposed to do to earn our fickle favor?

Designated Hitter: Kyle Schwarber (2nd)

Voted in: Shohei Ohtani

This is almost cheating since Schwarber (159 wRC+, 2.5 fWAR) only qualifies by virtue of Shohei (162 wRC+, 3.0 fWAR) earning the automatic starting spot. I’m sure many fans understood Ohtani’s inevitability in the All-Star festivities and thus opted for the major league home run leader, but the two-way terror was probably the more deserving candidate anyway (yes, as a DH. Oh, he pitches too?).

Outfield: Pete Crow-Armstrong (10th), James Wood (7th), Corbin Carroll (12th)

Voted in: Andy Pages, Brandon Marsh, Ronald Acuña Jr., Teoscar Hernández, Juan Soto, Michael Harris II

Here we have the opposite of the AL outfield situation in that I’m nearly certain that these All-Snubs not only deserved spots in the top six, but likely top three. Pete Crow-Armstrong (146 wRC+, 4.8 fWAR) is the league leader in position player WAR, yet he only finishes 10th? Is it because of that incident he had with the fan across from the outfield wall? Is it because Chicago as a city isn’t prominent enough?? James Wood (148 wRC+, 3.0 fWAR) and Corbin Carroll (145 wRC+, 3.5 fWAR) are such uniquely talented and likable players that I’m flummoxed at how they lost to Teoscar Hernández (118 wRC+, 1.1 fWAR). Brandon Marsh’s .401 BABIP is impressive, but is it that impressive? NL voters were keeping it mostly steady for the infield but fell off the rails once the train stopped in the outfield, losing focus amidst the coastal LEDs.

Subscribe now

Conclusion

Are the All-Snub teams better than their All-Star counterparts? The American League’s voting struggles concentrated in the infield, whereas the National League’s were undoubtedly in the outfield. All in all, the differences seem to even out between both, with any calculable differences being marginal. The issue with that is that theoretically, an All-Star team should house the best stars, so there being possibly better teams created in their stead is worrisome.

The operative word in that last sentence is “should.” Ultimately, the All-Stars that should earn the distinction are those that are chosen by the fans, since that’s the system that’s in place. But a system is not validated merely by its existence. Fans may want their All-Star Game to feature the players they already see all the time, but should that desire supersede the occasional injustice (which can and will affect perceptions of legacies)? Hall of Fame voters value All-Star appearances a non-zero amount, after all.

I’ll be curious to see which reserves are chosen by whichever of the handful of mysterious committees MLB utilizes to fill out the rosters. Until then, let me know which ones you’d most like to see earn the nod, and thank you for reading!

Jacob Misiorowski of the Milwaukee Brewers has been showcasing his dominance for a whole calendar year now (with last night’s start aptly marking his one-year anniversary as a major leaguer). Ever since he made only five starts before being (controversially) named an All-Star last year, the baseball world has been keenly aware of his special talent: throwing the baseball harder than most anyone ever, with accuracy and durability.

“The Miz” as an idea has fully evolved now in 2026. After the month of April, Jacob sported a 3.31 ERA and 3.23 FIP, which coupled with his high propensity for strikeouts showed great promise for the rest of the season. But I’m not sure anyone could have predicted how his next month and a half would play out.

Beginning with his start on May 1st, here’s how his ERA has evolved up until today:

• May 1: 2.84

• May 8: 2.45

• May 13: 2.12

• May 19: 1.89

• May 25: 1.83

• May 31: 1.651

• June 6: 1.50

• June 12 (last night): 1.34

No, this is not a reliever whose run prevention numbers benefit from small samples (though he pitches with as much dynamite as the fiercest ones ever). This also isn’t someone who’s just been lucky. Jacob Misiorowski has diminished his ERA with as much certainty as a starter possibly could, culminating in last night’s masterclass:

9.0 IP, 1 H, 0 ER, 0 BB, 15 K

Without any other context, that is already by far the best start of the season, and perhaps of the entire decade. But the ruthless fashion in which Jacob crafted this outing is what propels it into truly rarefied air, with some dubbing it the most dominant start of all time.

It wasn’t a perfect game; NL home run leader Kyle Schwarber guessing on a 91-mph slider sufficed for a single and the only baserunner of the game. However, Misiorowski did face the minimum number of batters (27), as Schwarber’s contribution was erased by a double play shortly thereafter. In any case, this game didn’t need to be perfect to be breathtaking in its own way.

Share

Greg Maddux is one of the greatest pitchers in the history of the game (if you ask me, he has an argument for the best). Maddux’s dominance was largely driven by his command and efficiency—so much so that the stat for a complete game shutout completed in less than 100 pitches bears his name.

Other than being a dominating righty, Misiorowski shares little in common with Maddux. The Brewer is built much closer to the mold of Greg’s career rival Randy Johnson, what with his tall lanky frame and blazing speed. Usually, pitchers like that aren’t as well suited to throw a “Maddux,” since their focus on velocity and strikeouts begets longer plate appearances on average. To wit, Randy threw 100 complete games, but none were a Maddux. Maddux, on the other hand, threw 13 of his namesakes.

You see where this is going. Despite Misiorowski’s archetype, his CGSO last night was indeed a 95-pitch Maddux. If you were to tell me that Jacob Misiorowski pitched a 15-K CGSO, I would follow up my “Wow” with “Jeez, he must’ve thrown a crazy number of pitches to get there.” He threw 101 pitches in 5.1 innings back in April for crying out loud. But no. He averaged just 3.5 pitches per batter last night. Over half of his batters faced struck out, which requires at least 3 pitches each. That is some diabolical efficiency. Jacob’s 15 punchouts now represent the record for a Maddux.

Of his 95 pitches, 74 were strikes. Mathematicians got on the task and were able to conclude that he threw only 21 balls. He induced 22 swinging strikes. He induced more swinging strikes than balls thrown. He never got himself into a 3-ball count, and only experienced five 2-ball counts. Clinical.

Oh, and did I mention that Jacob also broke the record last night for the fastest pitch ever thrown by a starting pitcher in recorded history (104.5 mph)? Because of course he did. He always does. I’m pretty sure a search of the 100 fastest pitches ever tracked from a starter would yield at least 95 instances of Miz the Wiz. Talk of his health always surfaces whenever he routinely throws 103+, which makes sense as Tommy John surgery almost feels like a formality in the career arc of a flamethrower these days. He’s only been a major leaguer for a year, so anything can happen in the future, but his unique combination of stature and flexibility shows promise health-wise—contrasting with Tim Lincecum’s form which most understood was unsustainable even at the time. Misiorowski’s smooth yet ferocious delivery perhaps best compares to Aroldis Chapman’s, with the added benefit of being able to start and finish a whole nine-inning game in the 103+ range as he did last night.

I don’t think anyone has ever seen consistent velocity dominance like that before. Maybe we saw it from Nolan Ryan, but his velo numbers are unverified, and potentially unverifiable. The crazy part is that Misiorowski’s stretch of games since the start of May don’t even need to have radar gun evidence to rival Ryan’s very best:

Misiorowski’s 0.17 ERA over his last eight is also an MLB record for starters in consecutive games. We were already witnessing history even before Jacob threw a pitch last night. The Maddux just put a stamp on it.

The classic calculation of Game Score offers an output of 100 for Miz’s domination of the Phillies (the newer calculation shows an even more impressive 107). This represents just the 17th triple-digit GmSc in a nine-inning game. It is the only one of those to be accomplished with a double-digit pitch count.

This Stathead search was made before the Misiorowski Game could populate the leaderboards, but shows all the other starts whose Game Score met or exceeded their pitch count:

Game Score is agnostic to pitch count and velocity. I never thought I would want a version to account for those things until today, because the merciless efficiency from the young Cy Young candidate2 simply isn’t adequately addressed through the box score.

Subscribe now

Other outings, at face value, may be better. Let’s not forget young rookie Kerry Wood’s 20-K masterpiece in 1998, or Max Scherzer’s 17-K no-hitter in the penultimate game of the 2015 season. Integrating context also complicates matters as we then must consider Don Larsen’s perfect game in the World Series, Jack Morris’s 10-inning Game 7 shutout, any of Sandy Koufax’s postseason masterpieces, and more. However, the cocktail of superlatives for Misiorowski’s game—“most strikeouts in a Maddux,” “fastest pitch by a starter,” etc.—had never been brewed before last night. The unicorn status of this start is undeniable.

The best part is that—much like the Ohtani Game—I have a feeling this won’t be the last “Misiorowski Game.”

Basically, many strategies are at our disposal depending on what we want.

This strategic openness reveals itself on the defensive side of position player WAR the most: Baseball Reference uses Defensive Runs Saved (DRS), whereas Fangraphs uses Fielding Run Value (FRV). Contrary to popular opinion, neither of these approaches are incomplete or misleading, but they chiefly differ in their approach to context. Where DRS is context-dependent in its measurement of run value on each play, FRV ignores base-out state and simply judges the player on how well he fielded the ball. So, a player with +5 DRS but -5 FRV may not be a talented fielder, but he probably made plays when it mattered most and saved his gaffes for less punishing situations.

Because of FRV’s desire to describe the player more than the play on defense, and because of the direction we’re about to go with the hitting portion, I’ve opted to stick with Fangraphs’ framework. As a result I am inheriting their positional adjustments, which I don’t know enough about to change.

Coming off my recent research on Fielding Independent Hitting, I grew curious about tinkering with the batting component of WAR. Both BRef and Fangraphs utilize wOBA as the foundation for their batter WARs, because it assigns run contributions to outcomes in a context-neutral manner. This is a highly defensible stance. It should also be noted that these WARs are not trying to predict the future, but to describe the past a certain way.

The more the past is described, the less is said about what may happen in the future. wOBA describes the past well, but not as well as RE24 (which accounts for base-out state) or WPA (which additionally accounts for inning and score). If you care solely about what happened, those latter two are great, but if you want to gain more insight into the actual capabilities of the player, wOBA is the better bet.

Pushing further in that direction, if the goal is describing what would be expected to happen (or what “should” have happened), then xwOBA is the logical next step. Using xwOBA as the foundation instead of wOBA describes the player better at the expense of describing the play worse. xwOBA doesn’t predict the future, but it does allow a clearer picture of underlying talent because it reflects process more than outcomes.

Similarly, FIwOBA (accounting for just HR, BB, HBP, K, & FB%) was also designed with this player > play philosophy, especially in small samples. Although it doesn’t predict wOBA as well as xwOBA, it’s still much stabler than wOBA and does alright for just five outcome variables.

PLAY — WPA - - RE24 - - - wOBA - - FIwOBA - xwOBA — PLAYER

Share

With all this in mind, I concocted park-adjusted WARs based on xwOBA (xWAR) and FIwOBA (FIWAR):

The best position player in the majors most ways you slice it is Bobby Witt Jr., whose unrivaled five-tool status elevates him above more elite hitters. The funny thing about the top two (Witt Jr. and Alvarez) is that their xWARs are far superior to their standard WARs, indicating they could be having even more explosive seasons (which seems unbelievable in Yordan’s case, but his .491 xwOBA speaks for itself). The two AL MVP frontrunners will likely offer an intriguing debate come awards season.

Yordan’s FIWAR is less convincing, but this is not a knock on him exclusively. Due to including less information in its calculation, FIwOBA generally plays it safer and is less reactive to extreme changes. This is reflected in FIWAR’s smaller standard deviation in this sample (~.78) compared to fWAR and xWAR (both ~.93). Since FIwOBA excludes all balls in play outcomes, its weak correlation with BABIP (r=.10) pales in comparison to wOBA’s (r=.62) and xwOBA’s (r=.35). What FIwOBA loses in descriptiveness, it gains in stability. Though I do wonder if its lower variance means the positional adjustment should also be tweaked down…

Behind Witt Jr. and Alvarez in xWAR and matching the former in FIWAR is Tigers catcher Dillon Dingler, who at the time of writing had a swell 3-for-5 day amidst his team’s 11-0 onslaught on Minnesota. Dingler has started to rebound from his unlucky first month or so and is poised to continue rivaling Shea Langeliers as the best hitting catcher in the AL, let alone best all-around one.

FIWAR’s conservative nature flies in the face of the other two WARs sometimes. For example, Kevin McGonigle, Riley Greene (more Tigers!), Otto Lopez, and Ernie Clement are viewed much less favorably by FIWAR than by fWAR. Ohtani also suffers more by FIWAR (1.1) versus the other two (each 2.4), which mostly has to do with his .364 BABIP being a good bit higher than his usual .328. On the flip side, hitters like Liam Hicks (discussed last week), Brooks Lee, Andrés Giménez, and Manny Machado appear through their FIWARs as if they should be much more productive than they have been. Machado’s FIWAR being higher than Shohei’s is hilarious and illuminates how FIwOBA can miss a lot on the extremes.

I encourage sorting the above table by each variant so that the differences between each WAR can be more clearly noticed.

Subscribe now

Overall, it was fun building these—which should come as no surprise to those of you privy to the state of things here at The Fielding Independent. The goal of xWAR and FIWAR was not only to implement my recent research into a fuller framework, but also to distill talent a little better than standard WAR does. However, the large sways for some players hint at the fact that estimations like these can still be a ways off.

There is truthfully an infinite number of WAR frameworks to choose from. In the future I am likely to try out other variants for position players, particularly those that go full “context” mode just for fun.

The series concludes this week with Part 3 where we will dissect the wOBA formula itself, gnawing at the same principles from a slightly different angle.

Compartmentalizing wOBA

As of June 4, 2026, the current season’s formula for wOBA is as follows:

wOBA = (.702×uBB + .733×HBP + .897×1B + 1.274×2B + 1.613×3B + 2.077×HR) / (AB + BB - IBB + SF + HBP)

The coefficients of each plate appearance event in the numerator change depending on the run environment of the league, but they don’t tend to deviate much. They are derived according to how valuable the event is on average, to produce a stat that communicates overall, context-neutral offensive production.

As we saw in Part 2, wOBA is not the most stable stat, largely because it depends on outcomes which are themselves unstable (namely, hits on balls in play). It is obviously not wrong for including these events, but the goal of FIwOBA and FIbbwOBA was to remove them in order to promote more stability (which they do).

As it turns out, however, we don’t have to create our own formula from scratch—we can also just pick apart wOBA. Enter wFIOBA and wBABIP:

wFIOBA = (.702×uBB + .733×HBP + 2.077×HR) / (K + BB - IBB + HBP + HR)

wBABIP = (.897×1B + 1.274×2B + 1.613×3B) / (AB - K - HR + SF)

wFIOBA includes only fielding-independent events, whereas wBABIP is just standard BABIP except with weights applied to each hit type.

These two stats can be effective if the goal is isolating by defensive dependence, but they (especially wFIOBA) can be misleading. Take Liam Hicks for example, a hitter who has done fantastic in his fielding-independent events. Hicks leads the majors in wFIOBA among qualified hitters with a whopping .733. Yet, just a little over a quarter (60/229) of his plate appearances so far have been fielding-independent. So although he’s converting on his three true outcomes masterfully, his approach restricts him from capitalizing on that gift as much as someone like Ben Rice (.553 wFIOBA), whose outcomes are independent of fielders 43% of the time. Hicks’ wBABIP is nearly bottom 10 among qualified hitters.

That doesn’t mean Hicks is an overrated hitter. Since wBABIP is so unstable, we can expect Hicks to not be that poor on balls in play moving forward. He may not come back to league average by season’s end, but he will likely run into some better luck with enough time. This is why his .392 FIbbwOBA outshines his .361 wOBA. Worth noting though is his more pedestrian .327 xwOBA; perhaps his trajectory over the rest of the season will be an effective test case on FIbbwOBA vs xwOBA.

The issue with wFIOBA is primarily one of scale, as the denominator is too dissimilar from wBABIP’s to just blend them together. Combining wFIOBA and wBABIP in a linear regression was not as descriptive (R-squared = 89%) as if we used FIwOBA, which hints at the efficacy of the negative strikeout variable in FIwOBA. In Part 1 we found that combining FIwOBA with BABIP yielded an R-squared with wOBA of 98%, which is incredibly descriptive. Using wBABIP instead increased the R-squared to 99%, indicating wBABIP is (predictably) more descriptive than BABIP.

Perhaps a more straightforward approach would be keeping the denominators of wFIOBA and wBABIP the same as wOBA, so that the two sum to it. That way, we can clearly see how much of a hitter’s production has come from fielding-independent events and how much has come from balls in play. Ildemaro Vargas is similar to Hicks in that his wFIOBA is sky high (.561), but since FI events have constituted a mere 16% of his PAs, we can see that only .092 of his .331 wOBA (27.8%) is from BB/HBP/HR. Meanwhile, Riley Greene sports a meager .323 wFIOBA, but he’s had such great fortune (league-leading .435 wBABIP) that his .386 wOBA isn’t too burdened by the fact that he’s had over triple the FI events Vargas has had. Even though Greene is far less clinical in converting those opportunities, his sheer volume means he still contributes more fielding-independent value overall (.141, 36.6%).

Ranking qualified hitters by the percentage of their wOBA gained via FI events (FI%) yields the following top 10 (wOBA in paretheses):

1. Munetaka Murakami - 72.5% (.401)

2. Kyle Schwarber - 71.8% (.404)

3. Mike Trout - 64.6% (.391)

4. Aaron Judge - 63.5% (.389)

5. Miguel Vargas - 61.7% (.380)

6. Manny Machado - 60.4% (.270)

7. J.P. Crawford - 60.2% (.342)

8. Max Muncy (LAD) - 59.5% (.374)

9. Colson Montgomery - 59.0% (.342)

10. Yordan Alvarez - 58.6% (.451)

Humorously, Murakami and Schwarber would still be only 15-20% worse than the league average hitter if their BABIPs were .000.

The positive correlation between FI% and wOBA (r=.38) as well as various other hints throughout this series indicate that fielding-independent production is more reliable and more valuable than ball-in-play production, though not overwhelmingly so. This vindicates the approach of isolating the more stable plate appearance outcomes from the unstable ones, though there is still use in mixing them around. For instance, shifting the home run variable from the FI side to the BIP side would essentially leave us with wOBACON and wOBA(BBs&HBPs), the former of which is much more valuable comparatively. Since wFIOBA is generally about twice as valuable as wBABIP to wOBA, it stands to reason that home runs are doing the heavy lifting.

Share

A more dynamic wOBA

If there’s one lesson that I’ve learned from this research, it’s this: Success on balls in play may be less important, but it cannot be wisely ignored. FIbbwOBA accounts for its BIP ignorance by penalizing strikeouts and fly balls, but that can only go so far. Eventually, sample sizes will become reliable enough that BIP success becomes genuine signal.

So, if the goal transforms into wanting a wOBA estimator that grows in its confidence on balls in play over time, we can adjust for this confidence depending on sample size.

Now, beware that what I am about to do is highly exploratory and not the most scientific. Much more exact research should be done to determine stabilization points of wFIOBA and wBABIP, but for now I have settled for my best-guess estimations based on what I know about how stable each event is and how the formulas impact that. Basically, I have estimated that wFIOBA “stabilizes” (becomes more signal than noise) at around 90 non-BIP events, whereas wBABIP does so at 780 BIP events.

What we can then do is create coefficients for both aspects of wOBA that reflect our confidence in them at certain PAs. Our confidence coefficients can be:

FI confidence (FIc) = (nonBIP/(nonBIP+90)) / (nonBIP/(nonBIP+90) + BIP/(BIP+780))

BIP confidence (BIPc) = 1 - FI confidence

As is often the case with math like this, it’s best understood with an example. To reflect the current averages of qualified hitters this season, let’s say our player has ~155 BIP events and ~80 non-BIP events. His FIc would be: (80/(80+90)) / (80/(80+90) + 155/(155+780)) = .47 / (.47 + .17) = .74. We would then apply a .74 multiplier onto his FI components, and 1 - .74 = .26 onto his BIP. In other words, we can be nearly three times as confident in that hitter’s fielding-independent profile than we can in his ball-in-play profile at this point. At the end of a full season, the gap between these coefficients will have shrunk, though likely* still somewhat in FIc’s favor (*Luis Arraez has entered the chat).

Let’s call this Dynamic wOBA (DwOBA). Its formula is:

DwOBA = (FIc × wFI value + BIPc × wBIP value) / (FIc × nonBIP + BIPc × BIP)

where wFI value and wBIP value represent the numerators of wFIOBA and wBABIP, respectively. DwOBA is adjusted to a wOBA scale.

Table time!

If you sort by the difference between wOBA and DwOBA (w-Dw), you’ll notice a familiar positive correlation with wBABIP (R-squared = 85%). FIwOBA differentials tend to be influenced by wBABIP a few percentage points more, whereas the opposite is the case with FIbbwOBA. In any case, seeing Riley Greene atop the w-Dw list comes as no surprise.

DwOBA ultimately ends up painting a similar picture as FIbbwOBA earlier in the season, with the former expected to describe wOBA better as the season progresses (as it includes balls in play more and more). We are probably already past that point as DwOBA correlates more strongly with wOBA (r=.90) than FIbbwOBA does (r=.79). This despite the fact that DwOBA, FIwOBA and FIbbwOBA all correlate with one another around the r=.94-.95 range. As for predictiveness and stability, DwOBA is meant to be more of a snapshot for a certain point in a player’s season, so I’m pessimistic of its ability to be as predictive/stable as FIbbwOBA.

A more sophisticated approach would involve regressing to league average wOBA at first—less and less as the season progresses—but I’ve kept it to just the ratio between FI and BIP events for the time being. Even more sophisticated would be stabilizing each event separately, rather than stabilizing only by whether the event is FI or not (case in point: HBP rate is not nearly as stable as K rate). At that point we start talking about creating an actual projection system, which is above my paygrade for now.

Subscribe now

Conclusion

wOBA, xwOBA, FIwOBA, FIbbwOBA, DwOBA, AbCwOBA, xYzwOBA… Had enough of the silly acronyms yet?

In the future, I will keep a watchful eye on FIbbwOBA in particular as it’s the variant in which I’m most confident. It is the one for those who prefer a BABIP-agnostic evaluation in favor of strictly inspecting outcomes within the hitter’s control à la FIP. I will also be curious to see how DwOBA evolves with growing assurance on balls in play, keeping its more experimental nature in mind. It is the one for the BABIP-sensitive types out there who are mindful of the slow but steady stability of base hits.

I’m content with what I was able to do here by sticking with “classical” sabermetrics as a creative restriction, but there’s no doubt that modern Statcast metrics have been and will continue to be harnessed to even greater heights. We have Tom Tango’s Predictive wOBA, and subsequent riffs by other professionals. Robert Orr’s SEAGER was foundational for advancing our ability to measure swing decisions. I also recommend this paper by Dalton Lowery published here on Substack last summer for a more contact-centric approach. Lastly, Batting+ from Nathanial Thomas breaks down hitting from many angles and shows great predictive promise.

For those who’ve read all three parts, thanks so much for your dedicated interest!

Recap: FIwOBA & FIbbwOBA

FIwOBA is essentially FIP applied to a wOBA scale for hitters. In fact, the regression coefficients for the “true outcomes” (HR, BB, K, HBP) were so similar to those in FIP that I decided to just use those. So, the formula for FIwOBA is:

FIwOBA = (13×HR + 3×(BB+HBP) - 2×K) / (8.6×PA) + C

The current top 10 in FIwOBA among qualified hitters in 2026 is as follows (with their actual wOBA in parentheses):

1. Yordan Alvarez - .425 (.454)

2. Kyle Schwarber - .399 (.411)

3. Munetaka Murakami - .399 (.401)

4. Ben Rice - .395 (.427)

5. Miguel Vargas - .387 (.371)

6. Liam Hicks - .386 (.354)

7. Aaron Judge - .384 (.400)

8. Mike Trout - .384 (.383)

9. Junior Caminero - .379 (.366)

10. Byron Buxton - .377 (.385)

Since the formula is completely agnostic to balls in play, it made perfect sense that the vast majority of the difference between wOBA and FIwOBA was explained by BABIP (batting average on balls in play). However, even though BABIP is very noisy, controlling it is still a skill for hitters. So, in an attempt to draw some valuable info out of BABIP, I found that including fly ball percentage (FB%) and opposite field percentage (Oppo%)—two fairly stable directional hitting metrics—helped explain roughly half of the initial difference.

Incorporating these elements into FIbbwOBA (Fielding Independent Batted Ball wOBA) altered the other coefficients slightly, but not substantially. The current top 10 in FIbbwOBA are:

1. Yordan Alvarez - .440 (.454)

2. Ben Rice - .415 (.427)

3. Munetaka Murakami - .412 (.401)

4. James Wood - .411 (.416)

5. Aaron Judge - .407 (.400)

6. Drake Baldwin - .398 (.406)

7. Kyle Schwarber - .391 (.411)

8. Bryce Harper - .388 (.381)

9. Liam Hicks - .384 (.354)

10. Sal Stewart - .381 (.360)

Differences between wOBA and FIbbwOBA tend to be smaller than between wOBA and FIwOBA, due to the additional information capturing more descriptiveness.

The goal of these metrics was to provide a more stable barometer of hitting talent than wOBA provides, and to potentially challenge xwOBA in that regard even without using Statcast. Do they?

Share

Putting FIwOBA to the test

In order to see how reliable FIwOBA and its batted ball variants actually are (to wOBA and to themselves), I analyzed year-over-year relationships between them. I kept the data modern, after the advent of the new shift rules, so 2023-25. FIbbwOBA was separated into a handful of variants so as to inspect the behavior of each combination of batted ball additions. All players with at least 200 plate appearances in each season were included in the analysis. Separate analyses of just hitters who switched teams between seasons were conducted, though they did not meaningfully impact the results.

In general, a few relationships stood out in both the 2023-24 pool and 2024-25. One is that, among all of wOBA, xwOBA, FIwOBA, and FIbbwOBA variants, wOBA was consistently the worst at predicting itself the next year. With it being the most purely outcome-oriented of the bunch, it makes sense that all the BIP event noise it bakes in results in worse reliability. Although FIwOBA and the FIbbwOBA variants were more reliable than wOBA year-over-year, they couldn’t really hold a candle to xwOBA, which benefits from precise process data. Even though in the prior analysis FIbbwOBA was shown to describe current-year wOBA as well as xwOBA, none of its iterations could quite match it in stability.

FIwOBA showed better stability year-over-year than any FIbbwOBA. This is logical since FIbbwOBA introduces variables which—though decently reliable themselves—combine to muddy the picture slightly. Interestingly however, FIbbwOBA variants slightly outperformed standard FIwOBA in their ability to predict next-year wOBA. I would hope this to be the case since the goal of FIbbwOBA was to capture the valuable BABIP aspects that FIwOBA omits. It seems this goal was accomplished with better predictiveness at the cost of worse stability.

In particular, Oppo% displayed inferior stability year-over-year to FB%, which also provided much better descriptiveness current-year. Oppo% generally seems like a weaker addition to FIbbwOBA as a result, and I am ultimately skeptical of its necessity alongside FB%. As this Pitcher List article eluded to, it’s possible that including spray angle unconditionally may lose some signal via the fact that handedness plays a substantial role in its effect on outcomes. There is potential for different FIbbwOBA formulas for lefties and righties, with lefties’ Oppo% playing a stronger role than righties’. Is it worth the hassle? Maybe!

As for FB%, the jury is still out on whether the delineation between IFFB% and OFFB% is necessary. The 2024-25 team-switcher analysis indicated that keeping the distinction offered slightly more predictiveness, but the other analyses did not, and it was mostly a wash on the descriptiveness front. IFFB% exhibits worse stability than simply FB%, and behaves a bit differently with BABIP. Because of these facts as well as my inclination towards the simple, I’m left to conclude the distinction may be unnecessary, though I’m not divorced from it completely.

(To be clear about the mechanisms of FIbbwOBA, a high FB% is a detriment, holding the “true outcomes” constant. If a player isn’t hitting a lot of home runs, you probably don’t want them to be sheepishly lifting the ball; at least let them try to keep the ball low so it can squeak by infielders.)

Now that we’ve diagnosed the makeshift wOBAs, it’s time to reveal an unfortunate truth, which is that xwOBA (probably) outperforms them all in the ways that matter. Not only is xwOBA more stable and just as descriptive, but it also seems to predict wOBA the best. Even though the 2024-25 results weren’t as clear about this as 2023-24, its superior stability hints at its efficacy.

Ultimately, the DIPS approach didn’t beat xwOBA. But, it did beat wOBA, which is something! If I had to rank them: xwOBA → FIbbwOBA w/ FB% → FIwOBA → FIbbwOBA w/ Oppo%. FIwOBA still provides some value in my eyes due to its stability, but I fear it misses the mark a bit too much. Infusing FB% allows for better accuracy in relation to wOBA and xwOBA while not sacrificing very much stability.

Subscribe now

Rolling with proper FIbbwOBA

I’ll leave you with a table of qualified hitters so far this season ranked by the version of FIbbwOBA I should probably stick with from now on, with just FB%:

FIbbwOBA = (13×HR + 2.8×BB + 2.9×HBP - 1.4×K - 1.16×FB%×PA) / (6.9×PA) + C

Next time (yes, there will be a Part 3!), we will dive deeper into the wOBA formula itself, re-integrating balls in play in the process.

Sometimes, that’s all you need! When it comes to evaluating pitchers, that is. Since those two variables stabilize so quickly in relation to most other metrics, and since they also resemble generally good pitching, K-BB% (and its twin in ERA form, kwERA) is a trusty stat to turn to 30% of the way through the season.

The beauty of K-BB%/kwERA lies in how simple it is whilst retaining its effectiveness. The modification we are about to make to it is not so simple, so if you prefer elegance above all else, kwERA is definitely enough for you. However, implementing just one other variable to the equation arguably improves its accuracy. That variable is ground ball percentage.

GBkwERA

Just over a decade ago, Jeff Zimmerman published what I perceive as a foundational article of modern pitching sabermetrics. In it, he dives not only into how well kwERA rivals other stats like FIP, xFIP, and SIERA, but also what kwERA misses in relation to ERA:

The above chart was published in that article, so all credit to Zimmerman. What it shows is the relationship between 1) the difference between ERA and kwERA, and 2) GB%. Clearly, there is a significant relationship between the two, which leads us to believe that GB% fills a need in the run prevention space that kwERA misses. Specifically, pitchers with a GB% around 30-40% especially tend to be represented a little more poorly by kwERA.

To rectify this, the formula for GBkwERA was created:

GBkwERA = kwERA * (-3.518*GB%^2 + 2.344*GB% + 0.629)

You’ll have to forgive me as I use this exact version for our purposes in 2026, even though this regression was derived back in 2015. Although the coefficients here are certainly not the same as they are today, they are probably close enough for our purposes.

Five years following Zimmerman’s contribution, Jason Fixelle converted kwERA and GBkwERA into their respective WAR counterparts. This analysis was illuminating as it showed that even though kWAR is a bit more stable year-over-year than gWAR (to be expected with the addition of an extra variable), gWAR predicts future fWAR about just as well, and predicts future RA9 even better.

Knowing this, let’s see how pitchers so far in 2026 stack up, not only by kWAR, but by gWAR as well:

Cristopher Sánchez’s recent CGSO helped propel him to the top of this list, where he’s trailed by sensational youngsters Misiorowski and Schlittler. Also notable is Mason Miller, whose 1.3 kWAR and gWAR see him at #35 on this list despite only pitching in relief.

Sorting by “k-g” (the difference between kWAR and gWAR) is a useful exercise in that it highlights those pitchers whose relationship between their ability to induce grounders and their ability to prevent runs is particularly extreme. The most drastic differential belongs to Tim Hill, whose lefty sidearm style has always benefited his GB% immensely. Jack Kochanowicz and Tyler Rogers follow; Jack’s arm angle isn’t anything crazy, but Tyler’s…

That’s all! I’ll definitely incorporate gWAR/GBkwERA into my pitching analyses moving forward alongside their simpler counterparts.

Some of this may sound familiar if you’re privy to modern pitching analytics. Voros McCracken—whose real name remains a mystery—spearheaded the sabermetric breakthrough known as DIPS theory (Defense Independent Pitching Statistics). This theory posits that pitchers have much less control over balls in play than initially suspected, and that removing such data from the equation can clarify a pitcher’s underlying talent level in smaller samples. FIP (Fielding Independent Pitching) is the most common iteration of DIPS nowadays. Its formula is:

FIP = (13×HR + 3×(BB+HBP) - 2×K)/IP + cFIP

Besides cFIP (which is just a constant to put FIP on an ERA scale), every component of FIP is an outcome that’s pretty much irrelevant to what the defense behind a pitcher does. Fangraphs’ WAR (fWAR) uses FIP as its key component, and as a result is more predictive and reliable than Baseball Reference’s WAR (bWAR), which uses runs allowed. If you wish to have an even more stable version of this idea, you can try xFIP (which normalizes HR to a standard rate), SIERA (which includes general batted ball data), or kwERA (which removes the unstable HR & HBP in favor of just K & BB).

My question then is: Can we apply the same DIPS concepts to hitters? (Should we?) Let’s find out.

High stabilization potential

As alluded to earlier, HRs and HBPs are not all that reliable for pitchers, with HRs taking ~1320 batters faced to stabilize and HBPs ~640. Given Ks (~70) and BBs (~170) stabilize much quicker, I naturally favor kwERA or xFIP early in a season to FIP.

However, for batters, HRs and HBPs are much more reliable, at ~170 PA and ~240 PA, respectively. The former makes intuitive sense considering batters have much more control over the type of batted ball than pitchers do. Whereas pitchers will face a slew of batters with varying tendencies, batters just have their set of tendencies. Add in the fact that home runs are defense-independent and you have the most reliable hit type for hitters. HBPs being more reliable for hitters than pitchers might make less sense on the surface, but comes from the same idea: Pitchers face various batting stances over time, whereas batters just have their one stance that determines how closely they put their bodies in danger to get plunked.

So, if we were to make a “Fielding Independent Hitting” stat, it could pretty much mirror FIP in its components, with the added benefit of all of its parts stabilizing relatively quickly.

The one quirk is that hitting stats don’t function on an ERA scale, with lower numbers being better. Instead, they function on triple slash line scales, with higher numbers being better. Enter wOBA (weighted On-Base Average), the premier hitting rate stat. The formulation for wOBA is powered by linear weights, which is the concept of applying run expectancy averages to every relevant plate appearance outcome, stripping away context in order to more accurately assess talent level.

Let’s make a fielding-independent wOBA, then.

FIwOBA

Since wOBA’s linear weights were derived through linear regression, it’s only right we do the same for FIwOBA. Only this time, we’ll regress our desired variables (HR, BB, HBP, K) to wOBA itself. Basically, given ONLY those four outcomes, how well and to what extent do they explain a hitter’s overall production?

Doing so on all 360 qualified hitters from 2021-2025 yields the following coefficients:

• HR%: 1.51

• BB%: .36

• HBP%: .39

• K%: -.21

Those numbers may not look familiar, but what if we scaled them all up, maybe by some random number like 8.6?

• HR%: 1.51 × 8.6 = 13.0 ≈ 13

• BB%: .36 × 8.6 = 3.1 ≈ 3

• HBP%: .39 × 8.6 = 3.4 ≈ 3

• K%: -.21 × 8.6 = -1.8 ≈ -2

Oh look, they’re FIP coefficients.

Mind blown! As it turns out, the approximate run values for the three (four) true outcomes for pitchers correspond the same way for hitters. Both FIP and FIwOBA ignore anything fielders could interfere with (besides the occasional home run robbery), and in turn draw the same conclusions about the importance of each outcome. This makes total sense of course, but discovering this after running the regression gave me a unique feeling of vindication most stats nerds can probably relate to. What Tango, McCracken, and Dreslough found nearly thirty years ago for pitchers, we’re still finding today for hitters.

How well do these outcomes explain wOBA, though? The R-squared value for this regression was .6798, meaning about 68% of the variance in wOBA can be explained by HR, BB, HBP, and K. That’s… not great, but it’s not bad! We’re completely ignoring base hits, yet we can still explain over two-thirds of a hitter’s overall production anyway. For reference, the R-squared between ERA and FIP for all qualified pitchers during the same time period is about 70%. Pretty close!

Now, would I recommend using FIwOBA as dependably as FIP since they explain their target metrics similarly well? Probably not. The generalized nature of batted balls pitchers face allows for more trust to be placed in FIP, since alterations from it are more likely to be random in small samples. On the other hand, hitters’ profiles are individualized, so the direction in which FIwOBA is wrong will be much more obvious to us. The direction in which FIP is wrong will not be, so it’s safer.

In any case, let’s see how hitters rank in FIwOBA so far this season:

Aaron Judge is indeed the best hitter in the league, and early-season FIwOBA agrees. The stat naturally favors sluggers like Schwarber and Moniak, and doesn’t like Matt Olson as much since it’s oblivious to the fact that he’s been a doubles machine so far. If we sort by w-FIw (the difference between wOBA and FIwOBA), the biggest “ball in play” overachievers so far have been Riley Greene (.094), Troy Johnston (.085), and Otto Lopez (.080), whereas the biggest BIP underachievers have been Cedric Mullins (-.091), Alec Bohm (-.072), and Caleb Durbin (-.068). Differences in the 2021-2025 sample were not nearly that drastic.

FIwOBA is not a better indicator of how productive a hitter has been this season—that’s wOBA’s job. Where FIwOBA is designed to outclass wOBA is in stability. Since FIwOBA is built only with highly reliable PA outcomes, I would not expect to see nearly as much year-over-year or within-year variation in it as opposed to wOBA, which necessarily anchors the muddy base hits along with it. I have not done the math on this, but I am nearly certain it is true.

Sorting that list by w-FIw is particularly illuminating for another reason. Let me just list off the BABIPs (Batting Average on Balls In Play) of the ten highest overachievers in order: .327, .291, .395, .287, .394, .354, .364, .367, .281, .436. League average BABIP is .288.

Now, the ten worst underachievers’ BABIPs: .170, .197, .193, .210, .171, .192, .204, .208, .198, .231.

Stark. The correlation between w-FIw and BABIP sits at an intense r = .7. This indicates that a huge chunk of what FIwOBA is missing lies within BABIP. Regressing wOBA on FIwOBA and BABIP applies the finishing touch:

wOBA ≈ 1.14×FIwOBA + 0.6×BABIP - .221; R-squared = .9805

98% of wOBA can be explained by FIwOBA (primarily) and BABIP (secondarily). That is huge. The math implies the remaining 2% is accounted for in the type of base hit (single/double/triple) and intentional walks.

Share

What to do about BABIP

When FIwOBA is wrong, it’s largely because of BABIP. They negatively correlate, whereas wOBA and BABIP positively correlate. The problem with BABIP is that it is quite unreliable (~820 BIP)—thankfully not as much as for pitchers (~2000 BIP), but still substantially so. Yes, hitters can control it, but not consistently in just 150-200 PA. And yet, what FIwOBA misses is so obviously found within it. What to do?

Now, you may be content with FIwOBA as is, sprinkling in BABIP to taste. The regression formula from a couple paragraphs ago will describe wOBA better than anything else I’ll present here, but will be more unstable and less predictive. If you’re interested in trying to capture some of BABIP while retaining predictiveness, read on.

While researching which metrics best predict BABIP, I stumbled across this helpful article from Pitcher List that digs deep into various hitting stats and how they correlate. Key insights I learned from that piece were that the angle at which the ball is struck is more important to BABIP than how well it’s struck. Line drives are best for boosting BABIP, but are much less reliable (~600 BIP) than ground balls and fly balls (~80 BIP). Darn. For the purposes of keeping our stat predictive, LD% unfortunately does not belong. Whether the ball is pulled or pushed (“spray angle”) also impacts BABIP a lot, and those tendencies thankfully stabilize expeditiously.

That left me with wanting to include one variable for where hitters send balls on the y-axis (up or down) and one for where they send them on the x-axis (pull or push). Through some testing (and with confirmation from that article), FB% appeared more effective as a proxy for the y-axis component than GB%, and including both would be redundant, so I chose FB%. Similarly, I found Oppo% (push frequency) to fit slightly better than Pull% in that it produced a more defined direction in the regression.

The last part of this process involved separating FB% into outfield FB% (OFFB%) and infield FB% (IFFB%)—incorporating some z-axis. I decided to do this because the average BABIP of an IFFB is so much worse than that of an OFFB, and so delineating between them seemed appropriate. This is potentially controversial though as IFFB% is moderately less reliable in small samples than FB%. Theoretically, this separation slightly dampens predictiveness while slightly improving descriptiveness. At least IFFB% stabilizes before season’s end, unlike BABIP.

Including IFFB%, OFFB%, and Oppo% in the regression yielded the following coefficients, which have been multiplied by 5 for ease of viewing:

• HR%: 10.0

• BB%: 1.7

• HBP%: 3.0

• K%: -1.1

• IFFB%: -1.0

• OFFB%: -0.5

• Oppo%: 0.6

Relationships between the original four outcomes remain relatively consistent, though the benefit of walks is dampened by the new inclusions. IFFBs were found to be basically as detrimental as Ks, which tracks with their near-zero BABIP and their inclusion as effectively strikeouts in ifFIP, the version of FIP used in fWAR.

I’m calling this new metric bbFIwOBA, which is potentially the busiest acronym in sabermetrics history (take that, GBkwERA!). Whereas FIwOBA explained 68% of the variance in wOBA, bbFIwOBA explains 80% (generalizing with FB% would’ve had it explain 79%, so very slight). That’s a huge improvement! And the best part is that predictiveness likely doesn’t take much of a hit in the transition to the batted ball version, though if I later find that it does I will pivot back to using just FB%.

Here’s how hitters rank in bbFIwOBA so far this season:

In terms of the difference between wOBA and bbFIwOBA (w-bbFIw), there’s still some correlation with BABIP due to early season instability, but it’s subdued. In the five-year sample, wOBA and BABIP were positively correlated (r = .37), FIwOBA and BABIP were slightly negatively correlated (r = -.21), and bbFIwOBA and BABIP were virtually not correlated (r = .07). Personally, that’s exactly what I want to see. wOBA has the balls in play, retaining noise and all; FIwOBA removes them entirely, missing potentially valuable noise; and bbFIwOBA re-introduces the parts of them that are more reliable, balancing predictiveness with descriptiveness.

Another way of conveying this point is correlating the differences with BABIP. In the five-year sample, BABIP explained 89.9% of the variance in w-FIw, but only 47.2% of the variance in w-bbFIw, meaning a hefty portion of BABIP is reliably expressed through FB% and Oppo%. Furthermore, the standard deviation in w-bbFIw (0.0116) is noticeably smaller than in w-FIw (0.0147), indicating greater stability between wOBA and bbFIwOBA.

Subscribe now

What about xwOBA?

The elephant in the room right now is xwOBA. If we already have that all-in-one, defense-independent Statcast metric that’s meant to be more predictive and stable than wOBA, why bother with FIwOBA and bbFIwOBA?

The primary answer is that I wanted to try. In the age of Statcast where scores of metrics are powered by advanced physics and technology, gleaning valuable information from data that humans can feasibly track means something to me.

Another answer is that xwOBA isn’t crafted with solely prediction in mind. The public formula is unknown, but from what we do know, all batted ball exit velocities and launch angles are included to determine the BIP portion of it. While accurate, this leaves it susceptible to unsustainable batted balls. bbFIwOBA and especially FIwOBA are specifically tailored to ignore such outcomes for the sake of predictiveness.

We see this play out when we run correlations with xwOBA in the five-year sample. Among those 360 qualified hitters, xwOBA explained 65.6% of the variance in wOBA (hang on—didn’t FIwOBA explain 68%?). Also for what it’s worth, xwOBA explained 68.6% of FIwOBA. The entree, though, is its correlation with bbFIwOBA, for which it explained 81.2%.

So… bbFIwOBA is better than xwOBA at explaining wOBA, AND it’s better than wOBA at staying true to xwOBA. It’s basically a less sophisticated SIERA.

Conclusion

I will definitely be returning to this concept in the near future, but I think I should stop for now. There has already been so much to analyze, and amazingly there’s much more still. Some avenues that have crossed my mind include replacing BBs and Ks with even more stable swing decision metrics, actually analyzing year-over-year and within-year predictiveness of FIwOBA and bbFIwOBA (which I simply didn’t have time for), and introducing sprint speed into the equation (though that’s a Statcast exclusive). I will explore at least some of these soon.

Overall though, I did not expect this defense-independent approach to work as well as it did for hitters, so I am pleasantly surprised. I hope amidst the sea of goofy acronyms that this was a fun and worthwhile read!

—Eric Liu

Power carries many definitions. It can mean whether or not the electricity in my home is working, with more power being good in that case. It can mean political, physical, economic, and/or emotional dominance over others, achieved and compounded with questionable levels of legitimacy. That kind of power, I would say, is not as good. It can also mean how reliable a statistical test’s results are (good!), or a superhuman ability often paired with corresponding responsibility (too good to be true!).

The type of power we’ll focus on today is the kind outlined in Merriam Webster’s definition 2, section a, subsection 2: “ability to get extra-base hits.” Why? Because this is primarily a baseball blog and not a sociology journal. But also because power is predictable.

Stabilization

A few weeks ago, I discussed what I believe to be the most important simple stat for pitchers, especially early in a season: K-BB%.

The reason this stat is so effective at communicating the underlying skill level of a pitcher is because it strips away most everything a pitcher can’t control. Give up a high chopper that ends up being an infield single? Ignore. Allow a smoked line drive for an out to the right fielder? Ignore. Surrender a home run that wouldn’t be one in any of the other 29 ballparks? Ignore. Just look at strikeouts and walks, and you remove variance the vast majority of which is noise. At this point in the season, both strikeout and walk rates have already stabilized for both pitchers and batters.

This is great news for those of us who enjoy gleaning information from data, which, deep down, is pretty much all of us. We all want to be able to look at a number and have confidence in it so that we can use it haphazardly in arguments over dinner. It is a national pastime.

So that raises the question: When can we trust balls in play? When can I start hyping up my favorite player’s gaudy metrics with statistical soundness? The answer lies in the type of batted ball. The hyperlink above takes you to a helpful article with all these answers spelled out analytically already, but let’s get there purely with common sense.

There are four types of hits: singles, doubles, triples, and home runs. Let’s put doubles and triples in the same bucket. Those hits are accomplished via a hard-hit ball—usually angled effectively between fielders—the vast majority of the time. The quality of bat-to-ball contact and power inherent in those hits is very often high, with little wiggle room. A hitter is usually not getting a double if he doesn’t hit it hard. Home runs remove fielders from the equation entirely, and even more obviously require great contact and power. Sure, there are some squeakers that get by in right field at Fenway, but overall, you know a home run when you see it.

Contrast extra-base hits with singles. The variability in quality of contact for singles is high. You’ve all seen softly-struck infield hits before, sometimes even balls chopped short of the pitcher. Those are decidedly not high-quality contact, but they are singles nonetheless. Then there are hard-hit singles, some that even bounce off the wall in the outfield. The hitter’s speed and perception of the situation have a lot to do with something like that happening, but the point stands that a ball hit off the wall and a ball grounded ten feet in front of the catcher have and will continue to be rewarded as the same type of hit under the right circumstances.

Since singles are that variable, it stands to reason that those types of hits take a good bit longer to stabilize. That is, we can be more confident at this stage of the season in a hitter’s capacity for extra-base hits compared to their capacity to hit singles, because of the difference in variability between those types of hits.

Isolated Power (ISO)

The formula for Isolated Power (ISO) is (2B + 2×3B + 3×HR) / AB. It is the percentage of a batter’s total bases that come from extra-base hits. ISO is also Slugging Percentage (SLG) minus Batting Average (AVG).

Since ISO considers only extra-base hits, and extra-base hits carry less variability than singles, it stands to reason that ISO probably stabilizes quicker than SLG (which considers all bases). Indeed, ISO stabilizes at roughly 160 AB, a number which qualified hitters in 2026 will be reaching very soon. SLG, on the other hand, stabilizes at double that (320 AB). AVG is the worst of the three in this regard (and others…), taking over 900 AB to stabilize. The record for ABs in a single season was set by Jimmy Rollins in 2007, at 716.

The mathematical elegance of ISO becomes all the more apparent with this information. Have a stat that takes over half a season to be reliable (SLG) and remove the part of it that’s wildly unreliable (AVG) to yield something which is doubly reliable (ISO). Neat.

Share

Power & Discipline in 2026

So, as we approach mid-May, it’s time for ISO to join the stabilization party with K% and BB%. Let’s see how every qualified batter is doing so far with regards to both their discipline (K-BB%) and power (ISO):

What I enjoy most about this graph is the complete lack of correlation between the two variables (r=.05), allowing for various types of hitters along the power/discipline spectrum. Sure, K% and BB% both positively correlate with ISO by themselves, but using K-BB% levels the playing field, in a sense.

The top right of the graph contains disciplined yet powerful monsters at the plate like Yordan, Trout, and Harper. The former is closest to Bonds territory, so here’s to hoping he can bounce back from his slight slump to keep appearing in such rarefied air. It also houses young over-performers in Rice, Vargas, Hicks, and DeLauter, who show great promise for the rest of the season.

Staying high but to the left are the more boom-or-bust “new school” types like Schwarber, Murakami, Walker, and Oneil Cruz. Mickey Moniak’s hot start places him firmly in this quadrant as well, in what might be the most chaotic breakout of the year.

A couple interesting names surface in the bottom left quadrant (aka the quadrant you really don’t want to be in). Rafael Devers’ move to the Bay Area was widely panned on the giving side, and although Boston probably still lost the trade, the Giants are hoping for much better from their star slugger. Fernando Tatis Jr. is also surprising to see down there, as his unique cocktail of struggles has resulted in an odd placement here for his standards.

Finally, the bottom right quadrant reveals plenty of predictable names in Luis Arraez, Steven Kwan, and Chandler Simpson. However, seeing Vladimir Guerrero Jr. down here is odd. His discipline deserves serious praise, but the power hasn’t sparkled yet. Vladdy is hoping to manifest Liu’s concept of infinite power to his advantage.

Subscribe now

Conclusion

If you’re familiar with the wonderful analyses from Down on the Farm, you were likely already privy to the notion that power and discipline are two of the “stickiest” attributes of a hitter. And now that we’re far enough into the season to seek valuable insight from both discipline and power metrics, we can be much more confident in how most players will behave at the plate moving forward. Of course, hitters will make adjustments and move around a bit on the power/discipline axes, but the hope is that you can look at their output in those areas right now and believe in them. And you can, because power is predictable.

Juan Soto has been himself (generationally good hitter), but suffered an injury that set him back a couple weeks. Every hitter with more plate appearances than Soto has performed below league average. Francisco Lindor has struggled. New marquee signings Bo Bichette, Jorge Polanco and Marcus Semien haven’t made good first impressions, and the void left by the absence of Pete Alonso, Brandon Nimmo, Jeff McNeil, and John DeMarsico is fiercely felt.

The part that is so “Mets” about their struggles is that their pitching hasn’t been bad under the hood (fourth-best FIP & xFIP). But even this part of their game would be bottom half if not for the efforts of one wizard on the mound, one whose capacity for dominance used to be more of a question… until now.

Nolan McLean debuted late last season for the Mets as a pitcher—a noteworthy caveat as they drafted him in 2023 as a two-way player. He made eight starts in 2025 and performed quite well in 48.0 IP, though not quite enough to save their postseason dreams. His 2.06 ERA was perhaps a little lucky, but his FIP (2.97) and xFIP (2.78) both broke 3 as well, indicating he might not be a fluke. Still, only eight starts weren’t convincing for some. Could he do that over a full season?

Although that question can’t be answered for another four or five months, so far it’s looking rather promising. His 2.24 FIP and 2.60 xFIP through six starts this season have resulted in the seventh-best fWAR and fourth-best kWAR among all starters. Turns out bettors were right to consider him for NL Rookie of the Year before the season started.1

McLean is able to flaunt such pristine peripherals (45 K, 10 BB) thanks to his amazing combination of velocity and movement on his pitches. His primary sinker sits 95 and moves more than average both horizontally and vertically, meanwhile his sweeper is known to be among the nastiest in the majors. He carries a four-seam, a curveball, and a changeup to great effect as well, though sometimes I feel like he could get away with just the sinker and sweeper. Mike Petriello breaks down this two-pitch combo beautifully here.

I could share more from Nolan’s Baseball Savant page, but you get the idea. Bunch of fiery red bars, not a lot of dull blue bars. “That boy nice” is as deep as the analysis needs to be when he genuinely looks like a Cy Young contender, in spite of his team’s bottom-of-the-barrel record.

McLean’s phenomenal start amidst his team’s woes are reminiscent of a few other young power pitchers in similar situations: Félix Hernández in 2010, Jose Fernandez in 2013, Paul Skenes in 2024. So Nolan’s circumstance isn’t entirely unique, but it is when adding in the fact that his team’s payroll exceeds $350 million. I don’t have to check; Hernández’s, Fernandez’s, and Skenes’s teams did not have payrolls anywhere close to that. A young, ace-level pitcher being the main bright spot on a team with perhaps the furthest placement in the “expensive/bad” quadrant ever2 is unheard of.

The question is, will it last? For McLean being good, I’d say absolutely. For the Mets being bad, I’d say probably not as bad as they have been3, but clawing back from this deficit throughout the rest of the season will be a treacherous journey. To even get back to .500, they’ll have to do a lot of twists and turns and spins in other directions. Much like McLean’s pitches.

So, Mets fans, you may continue to watch as many of their games as you like, though I would advise against it for the sake of your mental health. However, do make sure you catch this guy pitching every week. He is not like the others.

Sandy Alcántara had already shown he was All-Star caliber before the pandemic, so his dominant run in the early 2020s wasn’t entirely surprising. What was refreshing about it was the nature of his dominance as a workhorse. The 2021 season saw him finish 4th among all pitchers in old school pitching score (rewarding outs but penalizing Ks, HRs, and HBPs) as well as in innings pitched with over 205. He wasn’t an All-Star that year, but he did prove that his 97-98 mph four-seam/sinker combo coupled with his 90+ mph slider and changeup made for a blazing yet durable repertoire.

The following year saw Sandy put up a season that honestly feels unheard of just four years later. In 32 starts, Alcántara posted a 2.28 ERA aided by a 2.99 FIP en route to a unanimous NL Cy Young. Those numbers sound like a normal Cy Young season until you discover he did it in 228.2 innings pitched, the largest such total since 2016. This equates to over 7.1 innings per start, the most since David Price and Johnny Cueto in 2014. He of course led the majors in old school pitching score as well, with a whopping 454. All that while maintaining a strikeout rate below league average.

As 2022 was already amidst the era of aggressive bullpen management the league now finds itself in, Sandy’s season is all the more shocking in retrospect. It may not look like an outlier compared to the baseball of yesteryear, but to the modern climate in which it took place, it sticks out like a retro beacon.

Since then, no individual season has quite cracked the old school echelon that seamlessly, though Logan Webb has led the charge in Sandy’s stead. Nevertheless, I am pessimistic about the prospect of a starter ever throwing 225+ innings again, as crazy as that sounds.

Not even Alcántara himself has been able to replicate that level of magic. His 2023 was solid, but subdued in comparison. The FIP jumped over a whole point to 4.03, and what was once a respectable 17.7% K-BB% in his CY campaign regressed to 13.5%. Although his “stuff” was still intact, his peripherals and batted ball profile began to swing in the wrong direction. Tommy John surgery then sidelined him for all of 2024, and unfortunately, the trends continued in 2025: His FIP increased slightly to 4.28, but the ERA ballooned to 5.36. His K-BB% further fell to 11.4%. The quality of his pitches was noticeably worse, and hitters were connecting on him with greater ease than ever.

But there’s a reason we’re talking about him now.

Sandy was understandably emotional after his final start last season, and was put under a lot of scrutiny this offseason as a potential trade target. I have to think his approach early this season has been jolted by these residual feelings. If so, it worked… for a bit.

His franchise-record sixth Opening Day start was a scoreless 7-inning gem against Colorado, and his second start was even better: a complete game shutout against the White Sox (shades of his 6 CGs in 2022). His third trip to the mound was once again profound, as he pitched into the 9th inning and ultimately only gave up two runs. A few weeks into the season and people were starting to believe again.

He unfortunately had a rough go in Detroit the following week, surrendering 7 ER in 6 IP. Slightly better footing was found in his fifth start last week (5.0 IP, 3 ER), but on shady peripherals (1 K, 6 BB). His next start is tonight in San Francisco, and I’m hoping the best for him.

On one hand, Alcántara’s first trio of starts was among the very best in the majors, and felt wholly inspired by his CY campaign four years ago. But these last couple have sowed doubt. If we’re to believe Statcast, the quality of his pitches and his ability to locate them have both regressed from last year—the 5.36-ERA year. He’s added a cutter to his repertoire, which shows some promise as a new compliment to his sinker. However, his slider has lost several ticks of velocity. What’s perhaps most concerning is his 7.9% K-BB%, again worse than last year.

A key difference between last year and this year so far is how much he’s allowed batters to pull the ball. His Pull% last year was high at 45.1% (he’s used to ~40%), but this year it’s only 36.5%—great! In large part because of this, his BABIP thus far is .210, much lower than his usual .270-.280 range and far from sustainable. I want to believe the 3.06 ERA, but unfortunately, my brain knows the 4.42 FIP is probably the better indicator. For these reasons, I do not expect him to rival his electric 2022.

But, so what if I don’t? My friends know better than anyone not to listen to me when it comes to predicting sports. Alcántara has only had five starts—less than a sixth of a full season’s worth—the majority of which were awesome. Who’s to say he can’t reclaim his crown as the ultimate innings eater of our time?

I sure hope that happens. I wouldn’t bet on it, but I will hope on it for as long as I can.