Incorporating a Shift Ban into Hitter Projections

With the announcement of a likely shift ban in 2023, fantasy baseball analysts ought to prepare.

Many good articles have been written about the shift ban recently, with most echoing the sentiment that the leaguewide impact will be modest. Further, it is conventional wisdom that pull-heavy lefthanders would benefit the most.

However, the size of the impact has yet to be estimated for particular players. This article aims to fill that gap.

Previous Research on the Shift

In this article, the shift refers to a defensive alignment where three or more infielders are on one side of second base (other strategic shifts are ignored). A rigorous approach to measuring the impact of the shift, popularized by MLBAM Senior Data Architect, Tom Tango, is to first separate batters by handedness, then find each batter's wOBA with and without the shift when the bases are empty. Then, find the average effect of the shift on wOBA, weighting each batter's wOBA difference by plate appearances, either using the harmonized mean of plate appearances with and without the shift, or using plate appearances with or without the shift, whichever number is smaller.

I did this for 2015 to 2020 using Baseball Savant's Statcast Search, finding the bases empty shift hurts left-handed bats by 13 wOBA points on average relative to a standard defensive alignment (Tango found a 15 point effect in the same direction in a slightly different sample, capturing 2015 to the first half of 2020). One can take this method a step further and control for opposing pitcher, as Tango does, looking at the wOBA difference with and without the shift for each batter-pitcher matchup.

After controlling for pitcher, he finds a larger negative effect of the shift on left-handed bats of 24 wOBA points over the same sample. Alternatively, the shift does not appear to work against right-handed hitters. In the same sample as above, Tango finds the shift improves right-handed hitter production by 30 to 38 wOBA points. The right-handed hitter shift would likely 'ban' itself as more teams realize this.

Sam Sharpe continues Tango's work on researching the shift, finding a negative effect of 22 wOBA points against left-handed hitters across all base states (not just bases empty).

Incorporating Previous Research into wOBA Projections

To get a decent approximation of how banning the shift might impact each left-handed bat, I projected every left-handed hitter's wOBA with and without a hypothetical shift ban. Following a simple MARCEL approach, the projection captures the last four years of wOBA performance, weighted 5/4/3/2, from most to least recent season. It also adds 1,200 PA of regression to the mean (the mean is set at .317 wOBA). The "current" wOBA projection does not adjust for the shift ban.

To adjust each player's wOBA to account for a shift ban, I produced a second wOBA projection that adds 22 points to their historical wOBA in shifted PA (I took this number from Sharpe's study, which looks at the wOBA effect across all base states). Finally, I looked at the difference between each hitter's projected wOBA with and without the shift-ban adjustment. To get a rough idea of how much the shift ban would impact fantasy value, I used data from the FanGraphs auction calculator (default settings) to find that one additional point of projected wOBA improves a hitter's projected hitter rank by 1.5 spots (wOBA is not itself a category in standard formats, but as an all-in-one measure of offense, more wOBA is tantamount to more offensive production in the categories that do matter).

The table below shows the projections and improvement in hitter rank after incorporating the shift-ban adjustment.

As readers might have guessed, Joey Gallo is the biggest projected winner, though a lot of big names stand to benefit. Further, the benefits are neither massive nor trivial, with Gallo jumping 27 spots in projected hitter rank.

A weakness of the approach in this article is that it assumes each left-handed hitter is hurt by the same 22 wOBA points in their shifted plate appearances. This is an approximation, though probably a pretty decent one, as there is a lot of noise in looking at wOBA with versus without the shift for particular players. For instance, the table below shows a .003 correlation between hitter shift effects in adjacent seasons.

Further, depending on the form of the shift ban, these wOBA improvements may be overstated, as teams shift the third baseman and shortstop as close to second base as the new rules allow them to. These weaknesses leave room for future research.

Download Our Free News & Alerts Mobile App

More Fantasy Baseball Advice