The Shift isn’t dead, but it’s limited, and that’s good news for the San Francisco Giants and their quest to build a cost effective lineup around Aaron Judge or some other big fish.
Joc Pederson had a tremendous year, but it’s reasonable to expect a negative regression to some sort of mean. Boy, a limit on defensive shifting might just cushion that particular fall. And what about Mike Yastrzemski, a hitter the Giants desperately need to produce for them? Well, it looks like he’d be helped out a decent amount, too.
Over on his blog, Patrick Brennan, a former analyst for the Phillies and Reds and writer at Beyond the Box Score (RIP), “exported every batted ball with a shift and without a shift in the Statcast-era, built separate xwOBAcon models that considered the EV, LA, and spray angle of each batted ball, and then compared” to generate a list of hitters most impacted by defensive shifting and argue who might most benefit from limits on that shifting.
He wrote all this up in response to the Pirates signing Carlos Santana to a free-agent deal, because as a switch-hitter, Santana saw the most impact to his offensive line batting left-handed. The results involve the expected weighted on base average based on contact (xwOBACON) with the shift — as Statcast covers shifted versus non-shifted batted ball events (BBE) — and expected weighted on base average based on contact (xwOBACON) with a standard defensive alignment.
If you read through the post, he walks through the process and how he wound up settling on a model that would present as a “standard” defensive alignment and in a later tweet clarifies that the xwOBACON doesn’t include “softly-hit balls or pop ups.” He also makes it clear that not all shifts are created equal, but also that left-handed batters are disproportionately impacted by them. His analysis also doesn’t factor in things like sprint speed and non-obvious shifts.
But it’s a real effort to determine just how much the shift impacts outcomes and the result of his analysis is a top 20 chart of players most impacted by the shift in 2022 (minimum 200 BBE). You can read the full list on the blog, but I plucked out some key people here:
Removing some of the softer contact looks to have added about 100 points of xwOBA, so you can’t just go onto Baseball Savant and see what’s going on, but on the season, with all contact included, Joc Pederson’s xwOBACON was .443 on a wOBA of .373 and xWOBA of .367. By all measures, his quality of contact was better than the results. Brennan’s work speaks to how much defensive shifting played a part in it.
With Joc, I’m trying to think about how big a step back he can take next year before his value to the Giants is irrelevant. As a DH and with some limits on the shift, he might not fall back to the 100 OPS+-type player he was in 2020 and 2021. Doesn’t need to be a 144 guy like he was in 2022, but somewhere around 120 seems well within reach.
In Yaz’s case, it’s good news. If you suspected he was a little unlucky, then here’s data to back it up. He hit the ball harder in 2022 than 2021 (42.6% hard hit rate vs. 39.5%) but his pull rate also went up to nearly 40%. Even an xwOBACON up around .380 (it was .369 this past season) would help, because the Giants can’t just rely on him as a glove and occasional pop kind of guy. He might be the key to the lineup, Judge or no Judge.
And obviously, I included Cody Bellinger because he’s a free agent and a former Dodger during Farhan Zaidi’s time with that team and for a player whose sole value right now is defense, seeing any data that would suggest he could actually improve some on the hitting side of the value ledger would seem to boost his stock as a free agent worth pursuing.
Not mentioned on the top 20 for 2022 is Joey Gallo, who did, objectively, have a dreadful season (79 OPS+), but whose hard hit rate, barrel rate, and walk rate were all in the top 6%. His average exit velocity was 63rd percentile as well. On defense, he also rated 86th percentile in outfielder jump and 90th percentile in arm strength. In terms of the wOBA stuff: .286 wOBA, .295 xwOBA, .427 xwOBACON. And in Brennan’s blog post, Gallo pops up a few times on a list of 20 Largest Decreases in xwOBACON with the shift since 2015. In 2018, the difference was 0.083, 0.080 in 2017, and 0.051 in 2021.
The teams have all this data — and probably even better data not available to the public — so it’s reasonable to believe that the markets for Bellinger and Gallo will be healthy and maybe even surprising. Maybe even interesting options for teams that can’t sign Judge or Carlos Correa. Right now though, for the Giants, simple rule changes have improved the value of players already on hand.