An odd Jack Leiter statistical quirk

2021 first rounder has spent the entire 2022 season at AA Frisco. Assigning him to AA for his pro debut was a move that the organization acknowledged was aggressive, and Leiter has had his fair share of struggles in the Texas League, putting up a 5.50 ERA in 73.2 IP over 19 appearances for the Roughriders.

Leiter has been better lately — I heard Kennedi Landry on the Ballpark Dimensions podcast mention that Leiter has been much better since his Futures Game appearance (when he threw a shutout inning), and having gone and checked it out, that has been the case. In his last five outings, covering 23.2 IP, Leiter has struck out 24 of 99 batters faced, walking 13 and allowing a .202/,323/.298 slash line, along with a 3.80 ERA.

That is not, however, the split I found particularly interesting. No, what I found noteworthy was the difference in Leiter’s performance with runners on base, compared to with the bases empty.

Bases empty: .188/.335/.257, .253 BABIP, 43 Ks and 28 walks in 177 plate appearances

Runners on: .292/.369/.475, .378 BABIP, 43 Ks and 15 walks in 158 plate appearances

That’s a pretty dramatic split. As a point of reference, in MLB this year, hitters have slashed .235/.302/.384 with a .298 BABIP with no one on, compared to a .254/.325/.412 and .302 BABIP with runners on base. So while hitters are a little better with ducks on the pond (or at least a duck on the pond), there isn’t the dramatic 250 point OPS swing we are seeing with Leiter.

Interestingly, while walks have been a big Leiter bugaboo this year (and were in college as well), Leiter is actually walking fewer batters with runners on base — his walk rate with the bases empty is almost twice what it is with runners on base, while his K rate is almost the same.

The problem for Leiter with runners on base comes from the damage hitters do against him when they put the ball in play, rather than from walks. Along with the BABIP that is 50% higher with runners on, Leiter’s ISO is .183 in those situations, compared to a nice .069 with the bases empty. Not only is Leiter giving up more hits in those situations, he’s giving up a lot more extra base hits.

Per Fangraphs, Leiter has a 4.86 FIP and a 5.18 xFIP this year, compared to his 5.50 ERA. His overall BABIP of .309 is not out of line with what would be expected, but only 57.7% of runners who reach base on him end up not scoring. In general, pitchers are going to strand 70 to 75% of the runners they allow, but if you’re getting hit harder when you’ve got runners on, your strand rate is going to be worse.

There’s a couple of possible explanations for this, aside from it being just random variation, which, given the sample size and huge difference in outcomes, I’m not inclined to attribute this to (or at least, attribute all or most of this to).

One possibility is that Leiter is either on or he’s not, and when he’s on, he’s locating his pitches and dominating opposing hitters and generally not letting batters on base anyway. When he’s off, on the other hand, he has more runners on base because he’s struggling and thus he’s facing more hitters with runners on base. One could say that this is the case with every pitcher, but given every pitcher doesn’t have this dramatic a difference in outcomes when there are runners on versus bases being empty, it would have to be something more particularized to Leiter.

The second possibility is that he’s approaching batters differently with runners on base. It could be that Leiter is willing to err more on the side of missing with a pitch, and thus possibly issuing a walk, with the bases empty, whereas with runners on base he is seeking to challenge batters more and avoid putting another runner on, and so when he is missing his spot, he’s missing it in the strike zone, rather than out of the strike zone.

The third possibility is that Leiter is struggling to pitch out of the stretch, or at least is having trouble repeating his mechanics out the stretch the way he does from the windup. I would be curious to see his pitch maps and velocity splits in the windup versus out of the stretch, and see whether his velocity is significantly lower out of the stretch, or if the locations of his pitches are markedly different out of the stretch than out of the windup.

I don’t know what the answer is, and so I have no idea what needs to be done, if anything to address the problem, which is an unsatisfying conclusion, I know. But I thought the data was interesting, and so it is now shared with you as well.