It’s becoming a bit of an annual routine that I write this article. Last year it was about Caleb Thielbar in early June. This year it’s the seemingly snake-bitten Griffin Jax.
Through 20 appearances this season, Jax is carrying a disappointing 4.82 ERA. He’s been scored on in 7 of his appearances, including three in which he allowed two runs. Those seven “meltdowns” – defined as a relief appearance that reduces a team’s win probability by 6% or more – are tied for the MLB “lead” with two other relievers.
Jax currently leads the Twins’ bullpen average leverage of appearance (1.73), indicating that Rocco Baldelli and Pete Maki have consistently entrusted the homegrown and developed reliever with the most tenuous situations the club has faced.
But his rocky results have led many to speculate that he should be moved down the bullpen hierarchy for a little reset. While some lower stress opportunities might be useful from a mental standpoint – beat writer Do-Hyoung Park had an article earlier this week about how tough Jax can be on himself when he struggles – there is nothing in Jax’s data to suggest he does not belong exactly where the Twins have been putting him.
Not sure I've ever seen a reliever be on the mound for more weird innings in a six-week span than Griffin Jax.— Aaron Gleeman (@AaronGleeman) May 10, 2023
In fact, thus far, he’s shown the best stuff on the team (actually, some of the best in all of baseball. Yes... even a little better than Jhoan Durán.) and demonstrated some meaningful improvements over last season. He’s just run into a stretch of terrible luck that makes those things a little more challenging to see.
Stats like ERA or saves are lousy ways to evaluate how a relief pitcher has performed because their smaller samples can more easily be skewed by a bad inning or two. Those bad innings, which invariably happen to all pitchers at some point, count just the same, but they can also make it so a reliever’s top-line numbers look bad for a long time.
Instead, ERA estimator stats like fielding independent pitching (FIP), which focuses on strikeouts, walks, and home runs allowed, or Statcast’s xERA, which is based on the quality of batted ball contact allowed, are better future predictors. While Jax’s ERA is 4.82 and he’s got three blown saves and five losses, his FIP is 2.55 and his xERA is 2.82.
That FIP mark is tops among the Twins’ regular relievers and his xERA trails only Jhoan Duran’s 2.10 on the team. Across the league, his FIP and xERA are both within the top 15% of the 240 relievers to have thrown at least 10 innings through games played yesterday. What’s more, the 2.27 differential between Jax’s ERA and FIP is the 22nd-largest gap among those 240 relievers.
Contact Quality and Batted Ball (Mis)Fortune
These kinds of metrics look more favorably on Jax’s work than his run-prevention results in large part because of the type of contact he’s giving up. He has not allowed a home run yet this season (he gave up 7 in 65 appearances last year) and has been running the highest ground ball rate of his young career (58.9%, up from 47.3% last season). What that means, then, is that he’s given up 20 hits and batters have a higher-than-you-want .260 batting average against him, 16 of those hits have been singles.
This has been another weird inning in a season seemingly full of them for Griffin Jax.— Do-Hyoung Park (@dohyoungpark) May 10, 2023
Ball hasn't left the infield, but the Padres squeeze in a run.
This is a point where Jax, as frustrating as it has been, is different than a reliever like, say, Emilío Pagán. Jax has been getting nickeled and dimed by seeing-eye singles and bloopers, whereas Pagán is more prone to giving up that back-breaking homer.
Strikeouts are king for relievers, but weak ground ball and pop-up contact that is unlikely to go for extra bases can work, too. Jax has been above average in suppressing hard contact as measured by Statcast, and the ground balls have made him especially adept at suppressing slugging and barrels (the most productive batted ball type). He ranks in the 88th and 89th percentiles, respectively, in expected slugging and barrel rate based on contact exit velocity and launch angle.
Unfortunately for Jax, though, more of that contact has gone for hits instead of outs than would be expected this season, as evidenced by his .345 batting average allowed on balls in play (BABIP). That’s about 75 points higher than last season’s .269 BABIP and well above the Twins’ .279 team mark. More to that point, Jax’s expected batting average based on the contact he’s allowed is .224, right in line with last season’s .220. Four of his 20 hits allowed this season have come with expected batting averages below .130, and that’s not including the ground ball double in Los Angeles that was incorrectly called fair.
As tends to be the case, that inflated BABIP has also contributed to Jax seeing more of his baserunners come around to score than you would expect. His 57.1% left-on-base percentage is about 15 points worse than last season and the team average. Both of those are stats that you’d expect to see improve over time, if for no other reason than regression to the mean.
Measuring the Quality of the Stuff
The baseball analysis community has made significant progress in recent years in sussing out more granularly the things that pitchers have control over. Much of the information I’ve presented above remains rooted in the outcomes of pitches and plate appearances – that is, the things that have happened after the pitcher has released the ball and the batter has done something with it. We’ve long used those kinds of outcome measurements as proxies for the quality of a pitcher’s work.
Thanks to the development of several public-facing pitch quality models – Pitching+, Pitchingbot, and Pitch Level Value (PLV), to name three – that use detailed measurements about pitches before they are dealt with by the batter, we’re now able to assess pitchers earlier in the process of a pitcher-hitter interaction.
These new stats are helping us move beyond using the output of a play to quantify pitch quality and toward using things that pitchers have more influence and direct control over, like pitch type, velocity, velocity differential, spin rate, spin efficiency, spin direction, location, movement, and release points. These measurements break down the process of throwing a pitch and enable us to more objectively measure the pitcher’s skill and the quality of their offerings.
And it’s by these measures that Jax really stands out and has reasons for optimism going forward. Again among the 240 relievers to throw at least 10 innings this season, Jax is tied for the MLB lead with a 119 Pitching+ Score (100 is average). By Pitchingbot’s model, which is normalized on the 20-80 scouting scale, Jax leads all relievers at 69. Similarly, Pitcher List’s PLV metric ranks Jax’s 5.51 score in the 98th percentile among all relievers. All of these are notable increases over last year.
Back to the point about Jax’s unfortunate luck, Pitcher List uses PLV to estimate a measure called “hit luck” that quantifies how many hits a pitcher should have given up based on the quality of their stuff, and Jax is currently sporting a +17 (meaning he’s allowed 17 more hits than his stuff quality suggest he should have) that ranks in the 1st percentile. Whether you fully buy that measure or not, it supports the point that there’s nothing about what Jax is throwing that indicates he’s declined or gotten worse. If anything, he’s gotten incrementally better.
Griffin Jax, Dirty 87mph Slider...and Sword. ⚔️ pic.twitter.com/bko9vkaN8b— Rob Friedman (@PitchingNinja) July 13, 2022
Jax’s amazing slider (sweeper) drives the bus for his arsenal – it’s a top 3 slider by any of the pitch quality measures – and that pitch is moving a little more this season (about 2.5 inches more horizontally and 2 inches more vertically than last season despite staying at the same ~86 mph velocity). Jax has again improved the velocity of his four-seamer and it’s now averaging about 96 mph.
However you want to measure it, Griffin Jax’s “stuff” based on the physical characteristics of his pitches is flat-out nasty and among the absolute best in the game. Moreover, these pitch quality models have shown to be more predictive of future performance (read the primers linked above), especially for relievers within the same season, than anything else we have to use today, including FIP and strike-out to walk rate differential (the previous best predictive measure).
Luck Isn’t Everything
There are all the reasons to buy into Jax seeing better results going forward, but not everything can be chalked up to poor luck. Perhaps because of the bigger breaking sweeper that he throws about two-thirds of the time, Jax’s walk rate is an elevated 9.4%. He’ll need to get that down as the season progresses, but he’s always been a strike thrower and that does not figure to be a serious problem going forward.
Perhaps more concerning, especially in light of the quality of the stuff, is that Jax is missing fewer bats. His strikeout rate has dropped to about 22% after being 27% last season and that largely appears to be due to a swinging strike rate that’s dropped down to 10.6% from 13.1%. Given the stuff quality and velocity, it’s not clear what would be causing that, but my intuition is that it has to do with approach adjustments on both sides.
Hitters are now well aware that Jax is going to throw a metric ton of that sweeping slider and they appear to have adjusted to sitting back and looking for the pitch. They still chase it out of the zone at the same rates as they always have, but they’re making more contact against it, even when chasing.
In addition, Jax’s batted ball contact profile has shifted to where hitters are putting him in play up the middle and to the opposite field more than 75% of the time, which is about nine points more than last season and shows up pretty clearly in his batting ball spray chart:
This is not at all uncommon to see for pitchers who throw a lot of a dominant off-speed pitch – Toronto navigated this last season with Kevin Gausman and his tremendous splitter. It will be on Jax to counter this, probably by throwing a few more fastballs (or perhaps the cutter he’s toyed with) and also on the Twins to adjust their defensive positioning on the infield accordingly, particularly against right-handed batters that have given Jax fits so far this season (.344 wOBA allowed vs RHB, just .210 wOBA vs. LHB).
In the end, though, this Griffin Jax is an even better version than the breakout one from last year and almost everything would suggest his results will follow accordingly with time (and some small adjustments). As Rocco Baldelli put it earlier this week, “Griffin Jax isn’t going anywhere. He’s too good for that.”