Jeremy Peña Is Starting Out Strong but Coming up Short

Jeremy Peña is off to an excellent start. He’s also been one of the least productive hitters in baseball. How’s that for a lede?
If I told you that without any additional context, perhaps you’d think Peña was struggling at the plate but making up for it in the field. Yet, the former Gold Glove winner currently has -2 OAA and -5 DRS on the season. So much for that theory. Much to the contrary, Peña is on fire at the plate. Over the first six weeks of the 2024 campaign, he is batting .313 with a 129 wRC+. And while his .351 BABIP is likely unsustainable, his .327 xBA ranks second among qualified American League batters. His .363 xwOBA ranks in the 80th percentile, a big step up from his .305 xwOBA (22nd percentile) in 2023. Most impressive, he has cut his strikeout rate down to just 14.0%, ninth lowest in the AL. His strikeout rate has improved from the 30th percentile in his 2022 rookie campaign to the 61st percentile last season, and now it sits in the 92nd percentile in year three.
However, if you glance up from those percentiles on Peña’s Baseball Savant page, you might be surprised by the most important number of them all: His batting run value is zero. The line on the value spectrum is the faintest shade of blue, sitting about a quarter of an inch closer to “poor” than “great.” That doesn’t seem right. Indeed, out of 485 batters to see a pitch this year, Peña is the only one with a wOBA and xwOBA above .350 and a negative batting run value, according to Savant. It’s not hard to understand why he’s an outlier. Typically, when a player is hitting anywhere close to as well as Peña, he provides at least some positive value to his club.
Metrics like wOBA and xwOBA are context neutral, while Baseball Savant calculates run value by considering the runners on base, the number of outs, and the ball and strike count for each discrete event. If you take that general methodology a step further and also consider the inning and the score, you get a statistic like Win Probability Added (WPA) – although Peña might ask that we please, please stop taking the methodology a step further. According to WPA, Peña has cost the Astros far more than he has given back in 2024. Houston ranks second to last in the AL with -3.82 offensive WPA this season. Peña (-1.03) is responsible for more than a quarter of that negative WPA. Only two players have contributed to the team’s misfortunes more than Peña: the now-optioned José Abreu and a deeply slumping Alex Bregman.
Don’t let the fact that Peña has only the third-worst WPA on his own team overshadow just how much negative value he has contributed this year. His -1.03 WPA is the 11th lowest in the majors and sixth lowest in the AL. In other words, his actions at the plate have decreased his team’s chances of winning by more than all but 10 other players in the game. Yet, if you take a closer look at the bottom of the WPA leaderboard, you’ll quickly see that one thing is nothing like the others:
| Player | wRC+ | WPA |
|---|---|---|
| Andrew Vaughn | 37 | -1.58 |
| Jeff McNeil | 94 | -1.50 |
| Alex Bregman | 60 | -1.48 |
| Elehuris Montero | 37 | -1.34 |
| José Abreu | -20 | -1.33 |
| Tim Anderson | 36 | -1.29 |
| Ceddanne Rafaela | 63 | -1.19 |
| Andrew McCutchen | 82 | -1.06 |
| Seth Brown | 63 | -1.06 |
| Christian Bethancourt | -34 | -1.03 |
| Jeremy Peña | 129 | -1.03 |
| Kris Bryant | 46 | -1.02 |
| Bo Bichette | 50 | -1.00 |
| Paul Goldschmidt | 65 | -0.95 |
| Hunter Renfroe | 39 | -0.93 |
| Nolan Gorman | 65 | -0.93 |
| Carlos Santana | 69 | -0.92 |
| Eloy Jiménez | 78 | -0.91 |
| Randy Arozarena | 70 | -0.90 |
| Gleyber Torres | 67 | -0.89 |
| Bobby Dalbec | 7 | -0.89 |
| Mickey Moniak | 56 | -0.87 |
| Oliver Dunn | 70 | -0.87 |
| J.D. Davis | 78 | -0.86 |
| Leody Taveras | 97 | -0.86 |
Yes, Peña is the only player on that list with a wRC+ that’s better than league average. You’d have to expand the table to 53 players before anyone else would appear with a wRC+ above 100. (For what it’s worth, that player is Colton Cowser.) It’s not that good hitters can’t have a low WPA – right now, 30 qualified batters with a wRC+ over 100 have a negative WPA – but it’s unusual to see a player hitting this well with a WPA this poor. When a player is getting as many hits as Peña, you’d think some of them would have to come in higher-leverage spots. Similarly, Peña is making outs at a low rate this year (his OBP is 15% better than league average). In theory, that should limit the number of times he can come up short with the game on the line.
From 1974 (as far back as our WPA leaderboards go) to 2024, there have been 42,194 individual batter seasons (min. 1 plate appearance). In only 2,667 of those (6.3%) has a batter had a wRC+ above 100 and a WPA below zero. Only 204 times (0.48%) has a batter posted a wRC+ above 100 and a WPA below -1.00. And only once (0.002%) has a batter recorded a wRC+ above 128 and a WPA below -1.00 — Jeremy Peña in 2024.
I don’t expect Peña to maintain this distinction all year. He has almost 80% of the season left to play, and I’d expect both his wRC+ and WPA to regress toward the mean. Still, what he has already accomplished is surprisingly historic. Through the end of April, Peña was sporting a 138 wRC+ and a -1.27 WPA. Dating back to 1974, there have been nearly 300 calendar months of Major League Baseball (i.e. April 1974, May 1974, June 1974… April 2024). In that time, there have been tens (if not hundreds) of thousands of individual player months. Yet, never has a player had a higher wRC+ and a lower WPA in a single calendar month than Peña in April 2024. Few have even come close. It may sound hard to believe, but having clicked through every month of each of the past 50 years on a custom leaderboard, I can assure you it’s true.
The last hitter to even post a wRC+ above 100 and a WPA below -1.00 in a month was… Peña in April 2023. Huh! Meanwhile, the last hitter with a wRC+ above 130 to rank among the bottom 10 in WPA was Daniel Vogelbach, who had a 132 wRC+ and a -0.89 WPA in August 2022. The last player with a wRC+ above 120 and a WPA below -1.00 was Kevin Pillar in April 2017, who had a had a 124 wRC+ and -1.1 WPA over 110 plate appearances. Going further back, other players this century who came close to Peña’s numbers include Dustin Pedroia in August 2009 (136 wRC+, -1.02 WPA) and Javy Lopez in June 2004 (132 wRC+, -1.16 WPA).
So, how did Jeremy Peña accomplish such a singular feat? Although he didn’t ground into a triple play with the bases loaded and the Astros down by one in the bottom of the ninth, he is responsible for the worst offensive game by WPA so far this season. (Note: For the next few paragraphs, I’m going to be using Baseball Reference WPA. It’s not exactly the same as ours, but it’s quite similar and allows me to use the Stathead tool to compare individual player games over several seasons.)
On April 17, Peña went 0-for-5. In four of his five at-bats, he came up with runners on base and made the final out of the inning. To cap it off, he grounded into a double play to end the ballgame in the bottom of the 10th, stranding the tying run on third. That’s one heck of a bad game. According to Baseball Reference, no other batter has had -0.6 WPA in a single game this season, and only two more have even gotten to -0.5 WPA. Here are all 12 players who have had a game with -0.4 WPA or worse in 2024:
| Player | WPA | Date | Opp. |
|---|---|---|---|
| Jeremy Peña | -0.612 | April 17 | Braves |
| Bryan Reynolds | -0.523 | April 26 | Giants |
| Nolan Gorman | -0.504 | April 19 | Brewers |
| Ty France | -0.490 | April 14 | Cubs |
| CJ Abrams | -0.484 | April 12 | Athletics |
| Jeff McNeil | -0.480 | May 1 | Cubs |
| Eloy Jiménez | -0.453 | March 30 | Tigers |
| Masyn Winn | -0.435 | May 4 | White Sox |
| Jo Adell | -0.424 | May 1 | Phillies |
| Jeff McNeil | -0.420 | April 28 | Cardinals |
| Lars Nootbaar | -0.412 | May 4 | White Sox |
| Ezequiel Tovar | -0.404 | May 2 | Marlins |
Teams start every game with a 50% win expectancy. By the end of the day, the winners will have increased their win expectancy from 0.5 to 1.0, while the losers have decreased theirs from 0.5 to 0.0. Therefore, if an individual batter finishes a game with a WPA below -0.5, that means the rest of his teammates in the lineup combined for a positive WPA. That’s more than a little embarrassing; perhaps we need another golden sombrero-type term to describe such a performance.
If -0.5 WPA is bad, -0.6 WPA is far worse. It’s not easy to compile so much negative value in one game. Only five other active players have ever had a WPA worse than -0.6 in a single contest:
| Player | WPA | Date | Opp. |
|---|---|---|---|
| Sean Murphy | -0.665 | July 15, 2023 | White Sox |
| Sean Murphy | -0.654 | July 2, 2021 | Red Sox |
| DJ LeMahieu | -0.646 | September 25, 2020 | Marlins |
| Starling Marte | -0.645 | June 29, 2023 | Brewers |
| Andrew McCutchen | -0.629 | August 19, 2016 | Marlins |
| Will Smith | -0.613 | September 3, 2021 | Giants |
| Jeremy Peña | -0.612 | April 17, 2024 | Braves |
Instead of getting sidetracked by the fact that Sean Murphy owns the top two spots on that list, we need to keep talking about Peña. Because, while his performance on April 17 goes a long way toward explaining his low WPA, it doesn’t tell the full story. If you remove that one game from the data set, his wRC+ would only get higher. Yet, his -0.66 WPA in April still would have ranked below any other AL batter with a wRC+ above 100.
According to his game logs, Peña has only had a positive WPA in 11 of 35 games this season. What’s more, he hasn’t had many big moments to counteract his worst showings. Although he has been a top-25 hitter in the AL by wRC+, he doesn’t even have one of the top 200 games in the AL by WPA (using Baseball Reference WPA again). His two most productive games currently rank 225th and 233rd, and both of those outings came in May. His most productive outing in April ranked 443rd that month. He may be racking up hits, but his timing couldn’t be worse.
Peña quickly developed a reputation as a clutch performer during his rookie season. He hit .364 with a 198 wRC+ in high-leverage spots, and his 1.20 clutch score ranked ninth among qualified AL hitters. He was Houston’s best hitter in the playoffs, posting a 187 wRC+ en route to earning ALCS and World Series MVP honors. That said, his performance ever since is a good reminder that clutch (or un-clutch) hitting isn’t a sustainable skill. That was bad news for the Astros last year in the ALCS, when their 2022 postseason hero went ice cold, hitting .160/.192/.160 with the lowest WPA in the lineup. However, it might just be good news this time around. The Astros aren’t going to turn things around unless their bats stop coming up short in big moments, and Peña has nowhere to go but up.
Leo has written for FanGraphs, MLB Trade Rumors, Baseball Prospectus, and Lindy’s Sports. You can follow him on Bluesky @leomorgenstern.bsky.social.
Great article. Need more GIFS!!!!
Now, this is very surprising. And I got a little suspicious knowing that this season has started off very pitcher-friendly and WPA calculation can be sensitive to run environment.
Bottom line: WPA for 2024 does not seem to be scaled properly.
Why do I claim this?
From 2005 to 2023, WPA/162G for hitters leaguewide each season ranged from -0.33 to -2.54.
In 2024, WPA/162G stands at -6.03.
This is the case despite the fact that 2024 is the only season where hitters had a positive Clutch stat.
Now, I am not exactly sure where and why it went wrong, but I do believe the “true” WPA for hitters in 2024 is higher than what the numbers say.
WPA is calibrated for the recent past (the last 3 years I think?), so the negative yearly WPA/162 would fit with a declining run environment since 2005.
Offense is usually down more at the beginning of the season as well. I wonder if you only looked through mid May in some of the previous seasons if 2024 would seem like as much of an outlier.
The positive clutch stat is interesting, and my guess is that it just means that relievers have been relatively less effective than normal this year. Leaguewide, relievers have a 3.94 ERA and starters a 4.02 ERA. This gap is usually much larger.
So this is a really interesting article!
It inspired me to put this in perspective a different way. Based solely on a player’s WRC+ and PA, what is their expected WPA? What is the difference between that “xWPA” and their actual WPA?
Doing a basic linear fit of WRC+ to WPA/PA for players with >80 PA this season (and then multiplying by PA to get expected WPA), we unsurprisingly find that the largest underachiever is Jeremy Pena (expected 0.3, actual -1.0) with -1.33. Other notable players (differences greater than 1) are Jeff McNeil (-1.30), Colton Cowser (-1.18), James McCann (1.04), and Brandon Nimmo (1.12).
Looking at last year through May 8 as a rough comparison, notable players were Thairo Estrada (-1.28), Jeremy Pena (-1.17), Bobby Witt Jr (-1.04), Jesse Winker (1.03), Jose Ramirez (1.07), Jake Fraley (1.21), and Yordan Alvarez (1.77).
For decades we’ve had the concept of “empty” .300 hitters, where they’ll get a hit 3 of 10 PA but are otherwise not very useful in a lineup. This is a nice expansion on the idea, especially backed by documentation.
I will say that the “Batting Run Value” methodology Tango uses actually thinks that Ben Revere was a lot more valuable than wRAA does.
This does seem a little different than the empty average guys though. I always think of empty average as someone who doesn’t walk or hit for power much, so their production isn’t very useful no matter what situation they’re in. Like maybe 2008 Howie Kendrick, hitting .306 but having a 98 wRC+ because he had a 3.3% walk rate and a .115 ISO. Or like Cool Lester said, 2013-4 Ben Revere, hitting 305/306 and having wRC+ of 92/93.
The low-WPA high-WRC+ guy is someone whose production would be valuable if it was distributed differently, it just comes at inopportune times. Much more something that looks like bad luck than an approach that is expected to run up one stat without helping the team.
And RE24 has 2008 Howie Kendrick as a 8 runs more valuable with the stick than wRAA does – using Statcast’s framework would give him 3 more WAR on his career!
The classic example is always Ryan Howard, but I wish I still had SPSS access so I could run some tests on “RE24-wRAA” against just about every offensive stat in the book, haha.
It is different, yes, but the analogy that I was perhaps clumsily trying to make is that of a player who’s seemingly having a good year but not really directly contributing to team success.
Interesting. I watched a game tonight with Pena against the Yankees. The Astros were behind early, so while Pena scored three hits, including a HR late when his team was behind by six or seven runs, does this mean that his WPA went even lower? If so, what is this really telling us?
He had a .033WPA for that game, so his WPA rose.
I’d imagine it is very very hard to go 3 for 4 with a HR and have a negative WPA. The way to do it is something like: come up to bat, top of the first, 1 out, based loaded and GIDP. The home team then put up big crooked numbers in bottom of the first and second innings. Then a single in the 4th and 7th innings, with no one on, and they get stranded. Finally a solo HR with two outs in the top of the 9th, down 10 runs.
I remember Corey Seager, WS MVP in 2020 with a terrific batting line, had a negative WPA for the series.
You can lower your WPA with a hit – if you or another runner runs into an out. A HR will never lower your WPA – though if it’s late and you’re down loads it might round to 0.000 – the chance of winning down 20 runs is pretty much the same as winning down 19 – both 0 to a lot of decimal places.
Would also add that another way to have a negative WPA with a good batting line is to have a bad AB early, have your own team put up crooked numbers, and then to get hits while your team is up big. If the game’s already decided, what you do isn’t going to change the Win Probability much.
Not quite the same, and an illustration of Phil’s scenario rather than mine, but in the game where Jack Suwinski hit two balls into McCovey Cove, he had a -0.08 WPA going 2 for 4 with two homers; he made two outs with bases empty in the earlier innings, then hit a homer while the Pirates were down 6-1 in the 7th and another off a position player while the Pirates were down 14-3 in the 9th. (Though the position player doesn’t matter for WPA.) The outs put his team in a slightly worse position, the homers barely mattered because the Pirates were still almost certain to lose.
Peña did manage to pull a negative WPA while going 3-for-5 on April 12 this year against the Rangers; he grounded out in a tie game in the second, hit three singles while the Astros were down 8-1, 12-3, and 12-5, and then struck out with a man on when it was 12-8 in the ninth. I guess that suggests another way to get a negative WPA going 3 for 4 with a homer–your team falls behind bigly before you get an at-bat, you get your 3 for 4 with a homer while you’re down a lot of runs, and then the rest of the team stages a comeback and you do something like GIDP down 1 with the bases loaded and one out in the ninth.
It tells us that WPA is about fun facts, not performance.
RE24, the main input for “Batting Run Value” is a different story.
wRAA doesn’t distinguish between a Sac Fly and a GIDP.
“That’s more than a little embarrassing; perhaps we need another golden sombrero-type term to describe such a performance.“
All fedoras are embarrassing
More interesting to me is whether or not WPA has any predictive value or is another way of saying a guy is being more or less “lucky” than most. It’s well known that the ability to hit in the clutch is more about being a good hitter than being able to step up in tough spots, it’d be interesting to see which if any players consistently under or over perform their WPA based on their performance
There’s got to be something more said about the SAME GUY grabbing the two worst WPA performances since this stat was created (last 10 years?)!
It’s not like Murphy is some slouch at the plate, either! Must be an effect of hitting in the same lineup as Matt Olson…
And just to clarify, the list in the article only includes active players. So still pretty weird that the same guy has the two worst performances, but not quite as hard to believe! The oldest article I could find about WPA is from 2004, and dating back to 2004, Kevin Kouzmanoff and Casey Kotchman have both had a worse game than Murphy.
After yesterday’s games he’s now up to a 140 wRC+ and the WPA has at last crested down below -1.00! All the way down to -0.99 😀
This has to be heavily attributable to his position in the batting order? hes been mostly 7-8th until recently hes moved 5-6th hole and just yesterday moved into cleanup… went 3/4 with a homer hitting 4th yday albeit within context of low leverage situation given yankees were already up by 5+ runs.
This is probably one of the most fascinating articles I’ve ever read on FanGraphs. What an odd thing we’re seeing playing out. One of THE weirdest starts to a season literally ever.