How Do Shifts Affect League-Wide Batting Averages?

I was recently asked the following question by Rob Neyer: If infield shifts work so well, why aren't league-wide BABIPs (Batting Average on Balls in Play) dropping? It's a great question. Shifts are designed to to take hits away from certain pull-heavy hitters, and with the huge increase that we have seen in the number of shifts used across baseball over the last few years, intuitively we would think that this would affect the league's batting average. And it does! However, the effect is almost imperceptible because the number of batted balls against a shift is still a small percentage of all batted balls put in play.

First, for reference let's look at what the league-wide BABIP has been over the last 10 years, as well as the shifts data that we have been collecting at Baseball Info Solutions since 2010:

*Projected by year end

Based on research that we have done at BIS, we know that the shift lowers the batting average on grounders and short liners (the ball in play types most affected by the shift) by about 30 points. So far this season, the batting average on grounders and short liners on shifted plays has been .230, and on non-shifted plays it has been .265. That's a significant difference. However, despite the shift being employed far more often this season than any previous season, it has still only been used about 10% of the time. Therefore, the overall batting average on all grounders and short liners in baseball has been .262, only a 3 point difference from the .265 average on non-shifted plays.

And that's just grounders and short liners. If you factor in ALL balls in play, that 3 points gets diluted even further, because the infield shift has no effect on balls hit to the outfield. The league-wide BABIP this season is .299, but it would be .300 without the shifting. So, in general the shift is only going to lower the overall BABIP by about 1 or 2 points, and that gets lost in the noise when looking at year-to-year BABIPs.

However, just because it might be difficult to see the impact that shifting has had when looking at year-to-year numbers doesn't mean that shifting hasn't had a meaningful effect. So far this season teams have saved 127 runs throughout baseball by shifting. If we assume all those runs would have been earned, that means the league's overall ERA of 3.80 would actually be 3.85 if teams weren't shifting. So, the shift does make a difference.

On Tuesday, Tom Verducci published an article for Sports Illustrated supporting the idea that MLB should at least consider making the defensive shift illegal. The thought is that scoring in baseball has declined too much in recent years, so let's regulate the options available to the defense to keep things more exciting for fans. However, as the data above shows, the shift is just a small part of run prevention. A difference of 1 or 2 points in league-wide batting average is nothing compared to, for example, when the pitcher's mound was lowered after the 1968 season. While shifting definitely makes a difference, regulating it isn't going to reverse recent run-scoring trends. In fact, by taking away the shift and limiting the strategies that teams can use to gain an edge, MLB would actually be making the game less exciting.