Do Defensive Shifts Work? If So, Are Bunts the Silver Bullet?

On opening night, Carlos Martinez had his way with the Cubs to the tune of 10 strikeouts over 7 1/3 innings. In the eighth inning, however, the Cubs stirred to life. They loaded the bases, knocked out Martinez, and had Bryzzo coming up with less than two outs. Both failed to deliver, and that was sad for Cubs fans, but something jumped out during Rizzo’s plate appearance. With the bases loaded and two outs, the Cardinals presented Rizzo with a defensive shift: a lone defender to the left of second base in the lumpy form of third baseman Jhonny Peralta, standing halfway between second and third base.

Given that alignment, Rizzo could bunt . . . nay . . . drive a Mack truck through the left side of the infield. Even a mediocre bunt would likely result in a single, allow Willson Contreras to score from third, and tie the game. The announcers questioned why Rizzo did not attempt to bunt. THE ANNOUNCERS. Obviously, Rizzo and Cubs fans alike would prefer a grand slam or a double in that situation, but home runs and doubles are harder to execute than bunts, right? Or, perhaps, the shift does not have much effect on Rizzo, so he should not deviate from his ordinary approach.

Placing three or more infielders to one side of second base is not new to baseball. We know anecdotally that teams shifted against as early as the 1920s, and in 1946 Ted Williams faced his first defensive shift when Lou Boudreau ordered six fielders to the right-field side of second base. After a 50-year hibernation, teams are much, much more willing to flood one side of the infield these days. Joe Maddon’s Tampa Bay Rays woke the shift from its slumber, deploying a then-record 517 shifts in 2012. And in 2016, batters contended with shifts on 28% of the balls in play, with the Houston Astros leading all teams with over 2000 shifts.

The rationale undergirding the shift is well-known: with more fielders in the area where batters hit ground balls, the defense will turn more of those grounders into outs. How have offenses performed against the tidal wave of shifts? Take a look at these numbers are on balls in play (i.e. batted balls excluding home runs) over the past five years comparing performance against shift versus no shift:

On an aggregate basis, the shift does not have a consistent downward effect on BABIP. The shift “won” less than half the time. Interestingly enough, the Cubs deployed the fewest shifts in MLB with “just” 603 in 2016 – and still recorded the best defensive BABIP in the history of the universe at .255, proving that having all-world defenders at half your positions is a better defensive strategy if you can pull it off. Six years into the shift revolution, then, the jury remains out on how effective it is.

In general, teams deploy the shift most frequently against sluggers, especially against left-handed sluggers. Via Fangraphs, take a look at the 25 of most-shifted against batters from 2015 and 2016:

As shifts become the norm rather than an anomaly, and given the context of Rizzo’s plate appearance against the Cardinals on opening night, I wondered how the Cubs’ sluggers – Rizzo, Bryant, and Kyle Schwarber – fare individually against the shift. To analyze their performance, I used data from 2015 and 2016 for Rizzo and Bryant, and data from 2015 and opening day 2017 for Kyle Schwarber.

Near the top of the list above, Rizzo faced the second most shifts last season in all of MLB, on over 50% of his plate appearances. Teams shifted against Kris Bryant 302 times in 2016, or 22% of his plate appearances, placing him fourth on the list of most-shifted against right-handed hitters. And in 2015, teams shifted against Schwarber on 88 occasions, or 32% of his short-season plate appearances, which is a relatively small sample size and warps his numbers to a fair degree as discussed below.

How do these Cubs sluggers fare against the shift from a BABIP perspective?

Interestingly enough, Rizzo has only a slightly worse BABIP against the shift, Bryant’s BABIP is considerably worse against the shift, and Schwarber absolutely destroys the shift to an interstellar degree. This made me think that BABIP may not demonstrate the true value of the shift. Many batters have noted that the shift has a mental effect, causing them to change their approach even subconsciously, leading to more favorable results for the defense. Looking at batted ball data, it appears that the shift may have at least a minor effect on Rizzo, Bryant, and Schwarber.

First of all, each of these sluggers hit at least 50 percent more in-field pop-ups against the shift:

Second, each of Bryant and Rizzo have lower wOBAs against the shift, while Schwarber continues his interstellar success against the shift while looking weak against normal alignments. Note again that home runs are removed from shift stats, so these wOBAs look particularly low:

It is interesting how Bryant’s wOBA against the shift drops from an All-Star level to relatively poor, and yet he sees far fewer shifts that Rizzo. My guess is this apparent disparity is related to the effect that shifting defenders to the left side of second base with runners on base has on double play execution and other defensive coverages, but that’s pure conjecture at this point.

And Schwarber, how does he do it? Looking a layer deeper, it turns out that his BABIP on ground balls against the shift is .366. League average BABIP on grounders against the shift was .234. Perhaps Schwarber is really adept at going the other way and simply hits these grounders the opposite way? Nope: He pulled over 61% of these grounders into the teeth of the shift, and hit 30% of them up the middle. Nor did Schwarber hit these balls abnormally hard. The secret to his success is plain old blind luck over a small sample size; don’t expect that kind of success to continue.

Given this discussion, it seems like these Cubs should stick to their approach, especially because bunting against the shift does not have a drastic effect on defensive positioning, and hulking sluggers tend not to be particularly adept at bunting. Indeed, the MLB success rate on bunt attempts over the past two seasons is just 18.5%.

Over that same period, Bryant never attempted a bunt against the shift or any other defense. Rizzo, on the other hand, attempted to bunt against the shift multiple times, reaching base with a hit 66% of the time. Moreover, Schwarber led off a Spring Training game with a perfect bunt single against the shift. So the latter two sluggers may be pretty adept bunters.

Coming back to our initial question, what should Rizzo have done with the bases juiced and two outs? Under the influence of outcome bias, I initially thought Rizzo foolish for refusing to even threaten the bunt. After working through this analysis, however, I think he played the percentages just fine by swinging away, even if he would have looked like a g*ddamn genius if a bunt had worked out.

Lead photo courtesy Richard Mackson—USA Today Sports