The End of the Shift Helped Left-Handed Line Drive Hitters the Most

The 2023 season marked arguably the most-hyped and most impactful single rules change in recent memory in the form of the shift ban. Previously, teams had deployed ever more complex and radical defensive alignments to reduce the hitting success of batters with strong pull tendencies, particularly lefties. Front offices had reverse-engineered Willie Keeler’s famous axiom of “hit ’em where they ain’t” and were, instead, placing their fielders exactly where the hitters put the ball. Batters seemed powerless to knock the ball anywhere else, and so Rob Manfred pushed forward changes to limit where fielders could stand. Those changes didn’t exactly ban the shift, but they heavily curtailed extreme defensive alignments.

The results have been seismic. On a macro level, league BABIP spiked from a woeful .290 in 2022 to a more historically-typical value of .297. But we can do better than look at a single league-wide number and credit the entire change to one new rule. A more granular model of defense shows that the shift ban predominantly aided left-handed hitters, and had a much stronger effect on line drives than any other type of batted ball. This latter point runs contrary to analyses of the shift in the public domain for years, which have focused almost exclusively on ground balls.

To examine the impact of the shift, we need to know where each batted ball was hit, how hard, and at what vertical angle. Fortunately, each of these ingredients is supplied by Hawkeye’s cameras or, in MLB’s brand terminology, Statcast. In previous work on defensive efficiency, I’ve used a simple model with those ingredients to predict the probability that each batted ball will be cleanly fielded or drop for a hit, and I use that same methodology here. The model is trained on the 2021 and 2022 seasons, and then I use that model–reflecting the pre-rules-changes defensive positioning–to predict what would have happened to those batted balls in 2023. The difference between 2021/2022 and 2023 is thus reflecting (primarily) the impact of the shift, since we’ve ruled out other things that can vary (hitters, their exit velocities and launch angles, home parks, etc.).

I split the resulting maps into two: one for lefties, one for righties. The broad story is the same for both.

(Note: the color scale is not the same for the two hitters so you can’t directly compare them.)

The dark purple areas are where hitters are having less success in 2023 than they did before, and they match almost exactly with where fielders are standing almost uniformly among teams in the post-shift era. In the absence of all of the variation in position with shifting, teams have placed their infielders in four solid blobs, and the concentration of bodies in those positions has reduced BABIP right where they are, while increasing it almost everywhere else on the field. (The outfield is a notable exception, for the same reason: outfielders aren’t moving around as much, so where they are standing and immediately around it is more likely to be an out now.) You can see this effect in the following maps of fielder positioning from Baseball Savant: in the shift era, there was tremendous variation in tactics and placement, and now, by comparison, the fielders stay put.

Broadly, however, these maps show the largest increase in hitter success just behind the infielders, in the shallow part of the outfield, which is precisely what we’d expect given that the new rules require the infielders to play with at least one foot on the infield dirt. The second basemen and shortstops who used to pull back for hard-hitting lefties can’t get away with that anymore, and the result is that BABIP is up for lefties in particular.

The total change from the 2021/2022 model compared to the batted balls that dropped in 2023 is +750 extra hits. (I presume this is mostly due to the shift, but we can’t rule out every other factor.) Of those 750 extra hits, fully two-thirds, about 500, accrued to left-handed hitters, and the remaining 250 were with righties at bat. Parsing still more finely, 273 of the 500 extra hits to southpaws were on line drives. If you had asked me beforehand to forecast the type of batted ball most impacted by a shift ban, I certainly would have predicted it to be grounders, but those constitute less than half as much as the liners (~122 hits). (The remainder are fly balls, but most of them are low-angle: fliners, if you will.)

This flat trajectory-heavy mix isn’t just an artifact of unreliable stringer classification. You can clearly see in the following chart that the increase in BABIP covers not only ground balls, but basically all liners as well (this graph is mashing together lefties and righties).

Lots of analysis has focused almost exclusively on ground balls as the target of the shift, and that makes a certain amount of sense: line drives have exceptionally high BABIP regardless, so it would seem that targeting them by moving fielders around would make only a marginal difference. I’ve certainly been guilty of querying ground balls alone in my shift analysis. But, there are two factors that show that liners were actually the more valuable target of the shift.

The first is that high BABIP itself. On average, worm-burners in 2023 had a BABIP of .243, up from .235 the prior year. Grounders are not and never will be a very desirable result for a hitter because of that low probability of success to begin with: You can decrease it a few percentage points, but it only makes an already unattractive outcome for a hitter marginally less attractive. Compare that to liners, which hitters almost always do want to achieve: Their BABIP went from .690 in 2022 up to .702 in 2023, a larger 12-point increase on a type of batted ball that is better to begin with.

The other important factor here is the impact of grounders vs. liners in terms of linear weights or outcomes. A grounder’s best case scenario is often just a single. In contrast, many liners fall for doubles. Expected run value of batted balls tends to peak in the 10-15 degree angle range, the beginning of the definition of a line drive.

So, because liners tend to have higher BABIP and greater run value, there is more leverage for defensive positioning to affect them. Shift repositioning depressed the value of line drives as much or more than it did grounders, and we saw the effects last year in terms of higher BABIP but also substantially (almost 10%) more run-scoring (another factor here was an abundance of home runs, but that’s a separate topic). Thanks to Manfred’s rule changes, we got the closest thing to a controlled experiment that baseball permits, and it showed that the shift killed all batted balls on flat trajectories from around -20-25 degrees, the most valuable and common of which were apparently line drives from lefties.