clock menu more-arrow no yes mobile

Filed under:

There’s something fishy about AJ Pollock’s splits

There’s something to it, but not as much as you might think

If you buy something from an SB Nation link, Vox Media may earn a commission. See our ethics statement.

AJ Pollock bats against the A’s in September 2022 Photo by Michael Zagaris/Oakland Athletics/Getty Images

On Saturday night, Jeff Passan broke the news that the Seattle Mariners are signing outfielder AJ Pollock to a one-year deal for $7 million, though incentives could push that to $10 million. The right-handed hitting 35-year-old is widely expected to be the weak-side of a platoon with whoever wins the Jarred Kelenic/Taylor Trammell/Cade Marlowe battle in Spring Training.

But before we settle into a vision of Pollock as just a lefty masher, we should look deeper because Pollock’s splits aren’t as extreme as they first appear. The freshest data, from 2022, shows a monster gap: a .394 wOBA (161 wRC+) against lefties compared to just .265 (69 wRC+) against righties, largely driven by a .266 gap in his ISO.

But how reliable is that? Tom Tango et al.’s The Book tells us that right-handed hitters, such as Pollock, need about 2,000 PAs against lefties before their splits are reliable. That is to say, under 2,000, you’re better off regressing the data you have against league-average splits. For his entire career, Pollock’s only about halfway to that threshold, with 1,204 times facing left-handers.

We can do that by weighting Pollock’s career splits (a wOBA .032 points better against lefties) as 60.2% and league-average splits (over the past 20 years, righties have fared .018 points better against lefty pitchers) for the remaining 39.8%. That drops Pollock’s splits to just .026 points of difference.

Chicago White Sox v Cleveland Guardians Photo by Ron Schwane/Getty Images

Of course, using Pollock’s career numbers only gets us so far. Do we really think he’s the same hitter in his mid-30s as he was in his mid-20s? I’ll tell you from personal experience, that’s an impactful decade on the body, to say nothing of the fact that Pollock is a noted swing-changer. Much of the excitement I’ve seen on Twitter has to do with Pollock’s 2022 stats. But that’s even less reliable. Obviously, there’s an issue with using a sample that small, but check this out too:

AJ Pollock’s 2022

Split wOBA xwOBA Gap
Split wOBA xwOBA Gap
Split v. R wOBA .265 xwOBA .306 Gap -.041
Split v. L wOBA .394 xWOBA .350 Gap +.044
Forgive the double-naming on desktop—it’s the only way to make this work for mobile

Last year, Pollock was Juan Soto against lefties but David After Dentist against righties. Yet much of that was driven by massively overperforming his expected stats against lefties while massively underperforming his expected stats against righties, exacerbating his natural splits. The gap between his actual splits and his expected splits ranks third out of 227 players with at least 100 PAs against both lefties and righties.

As you might have guessed, that’s unlikely to last. Here’s a chart showing every player with at least 100 PAs against both lefties and righties in both 2021 and 2022. On the y-axis, I have every player’s 2021 left/right gap in his xwOBA over- or underperformance. On the x-axis, I have the same for 2022. You can see there’s hardly any carryover from year to year. For the nerds, this is an R-squared of 0.002.

A scatter plot showing a lack of correlation between wOBA minus xwOBA splits from 2021 to 2022
With a pattern this random, this is more like a post on Jackson Pollock, am I right?
Chart: Zach Mason for Lookout Landing * Source: Baseball Savant * Created with Google Sheets

The extreme outlier that I put in red in the upper left is Jesse Winker. I don’t have anything to say about this, but thought you’d find it interesting.

The black dot in the middle on the left is AJ Pollock, whose 2021 looks a lot more normal. His play against lefties was still pretty good, with a .366 wOBA. While that’s hardly the Juan Soto-esque production he put up last year, it was closer to his true talent, as measured by a .363 xwOBA. And the good news is he wasn’t nearly so bad against righties in 2021, with a .379 wOBA, which actually overperformed his .357 xwOBA.

So, look, that was more numbers than I usually like to put in a post. I’m mostly here to riff on movies from 1972. But the bottom line is that we should pump the brakes on thinking of Pollock as a platoon player. Here’s something he did against a righty just four months ago:

To be clear, I’m not saying he’s not better against lefties. He is. He’s just not as good against lefties as you might think. But conversely, he’s not nearly as bad against righties as he was last year. My expectations are for a good, but not great hitter no matter who he’s facing. So don’t freak out at Scott Servais if you see Pollock facing a right-hander. Between injuries and the team’s current DH situation, he’s likely to do that a fair bit. It’ll probably be fine.