When asked about Major League Baseball’s interest in restricting defensive shifts, Pirates manager Clint Hurdle recalled that when he was growing up in Merritt Island, Florida, they were often short on players for neighborhood games. Hurdle said they would arbitrarily cut the field in half to solve the problem. “Sometimes we’d shut down the pull field. … We just would do it to change the game and realized we learned how to hit the ball the other way,” said Hurdle. “What the shifts are telling hitters is, ‘Here’s what you do. Where is your counterpunch? Where is your answer?’”
Many coaches, commentators and baseball observers have asked the same question, bemoaning batters’ seeming inability to adjust to opponents’ shifts — the tactic of moving defensive players out of their usual position to overload one side of the infield, a strategy that has proliferated across the sport in the past decade. The shifts have also become one of the most conspicuous on-field byproducts of baseball’s data age as more and more teams decide how to align their defenses using actual batted-ball data. In December, The Athletic’s Jayson Stark reported that MLB commissioner Rob Manfred had “strong” backing from baseball’s competition committee to limit defensive shifts. Texas Rangers slugger Joey Gallo’s Christmas wish was to see shifts banned.
But in all the hand-wringing over the shift, one detail has been overlooked: Batters have adjusted, and they’ve done it without league intervention or legislation. What’s more, there’s reason to believe shifts are actually encouraging increasingly efficient offensive behavior.
Shifts have grown at a staggering rate. In 2011, defenses deployed the shift — counting both the traditional (three infielders to one side of second base) and non-traditional varieties — during 3,065 major league plate appearances that ended with a ball being put into play,1 according to Baseball Info Solutions data housed at FanGraphs. That’s only 2.6 percent of all at-bats where balls were put in play. The number of plate appearance where hitters faced the shift has increased every year since, save for 2017. Last season, batters faced the shift in a record 40,730 total plate appearances ending on balls in play — that’s about 34 percent of such plate appearances.
The era of the shift has coincided with a league-wide decline in batting average, though that is more a product of the record strikeouts rates in recent years as fewer and fewer balls are put in play, as batting averages on balls that are put into play has remained steady despite all the shifting, as you can see on the chart below.
While shift usage has grown dramatically, there’s evidence that batters have adjusted by going over the shift, which reduced the overall effectiveness of the shift across baseball.
In 2011, batters hit ground balls 53.2 percent of the time when they put a ball in play against the shift. Last season that number was 43.9 percent, which is the lowest such rate since at least 2010, the first year for which data is available on FanGraphs. When batters are not facing shifts, ground-ball rates have remained steady. Batters had a 45.9 percent ground-ball rate in 2011 when not facing a shift and a 45.9 percent ground-ball mark last season. You can see the divergence in strategies in the following chart. The drop in ground-ball rates against the shift suggest that more players are trying to bypass the infielders altogether by knocking one over their heads.
Not all hitters try to adapt — Bryce Harper, for example, has a career 1.4 ground-ball-to-fly-ball ratio both when facing a shift and when not facing a shift. But those who do, Joey Votto, tend to go high. Votto’s career ratio when not facing a shift is 1.57 ground balls to fly balls, but that drops to 1.44 against the shift.
“I’ve tried to avoid the right side of the infield shift,” Votto said in 2017. “I’m not excited about hitting balls to that side because I could hammer a one-hopper to the second baseman or shortstop, or whoever they have stationed over there. … Personally, I embrace the fly-ball thing just because of that reason.”
The average launch angle of a batted ball has increased in every year of Statcast era,2 rising gradually from 10.1 degrees in 2015 to 11.7 in 2018. But with the shift on, batters are even more likely to hit the ball in the air. The average launch angle against the shift last season was 14.7 degrees, a notable jump up from 13.1 in 2015.
In addition to MLB-wide trends, I looked at the behavior of the regularly shifted-upon batters in 2018 to see how their approaches changed.3 This group combined for a 42.5 percent ground-ball rate when facing shifts and a 44.1 percent rate when not facing shifts.
“Is [banning the shift] that going to produce more batting average? Maybe,” said Houston Astros manager A.J. Hinch at the winter meetings. “More runs? Debatable. A more energized and entertaining game? I doubt it.”
Left-handed hitters are an interesting study since they now put more balls in play with the shift on (26,076 last season vs. shift) than off (23,214 against no form of shift).
Last season, left-handed batters hit for a higher average (.300), greater slugging percentage (.388) and lower ground-ball rate (44.0 percent) when the shift was on compared to when there was no shift (.295 average, .380 slugging mark, 45.7 groundball rate). And because Baseball Info Solutions can only track shift data when a ball is put into play, those stats do not include home runs, since they are not in play.
In some ways, the shift has backfired. Batters have an incentive to hit more balls in the air, and balls hit in the air are more valuable. When batters faced a shift last season, 5.2 percent of balls they put in play went for a home run. When they didn’t face a shift, 4.1 percent of balls went for home runs, according to Statcast data.
While more batters try to go over shifts, they are not always going to the air in the most optimized manner. Every hitter who has faced a shift has probably been advised to try and go the other way. And as a result, the percentage of batters pulling line drives and fly balls against the shift has fallen off notably since 2010, from a 31.5 percent pull rate in 2010 to 26.2 percent in 2018. But by going the other way, batters might actually be hurting themselves. They are purposely avoiding the most valuable batted ball in baseball: a pulled ball in the air.
Consider that in 2018, 32.7 percent of fly balls to a batter’s pull side went for home runs, compared to 8.1 percent of fly balls to center field and 3.8 percent to the opposite field. Batters across the league had a .429 average and 1.514 slugging percentage on fly balls hit to the pull side and a .135 average and .324 slugging mark on balls hit to the opposite field. That’s not much more valuable than a ground ball. Last season, MLB batters hit .236 and had a .258 slugging percentage on ground balls.
Many have made the case for batters facing the shift to simply bunt more often. After all, batters have hit at least .357 when bunting against a shift every season since 2010. Would bunting be more effective than, say, trying to go over the shift? Not for most batters.
According to weighted runs created plus (wRC+) — a metric that adjusts for ballpark and scoring environments, with 100 representing league average — batters produced a 53 wRC+ mark on bunts against all shift types last season compared to a 127 wRC+ mark when putting the ball in the air against shifts.
Batters seem to unwilling to sacrifice potential power in pursuit of infield bunt singles. The percent of at-bats against the shift where the batter bunted has fallen four straight years, from 2.92 percent (2015) to 2.12 percent (2016), 1.88 percent (2017) and 1.73 percent (2018).
One other issue: Teams are pitching less effectively to the shift.
As more and more batters use an uppercut swing to better combat sinking fastballs, which are designed to produce ground balls, the percentage of sinkers thrown has decreased. Sinkers represented 22.4 percent of all pitches thrown in 2010. Last year? 16.9 percent.
The shift will always be effective against pull-side ground balls and low line drives. Batters who hit those batted ball types often, especially left-handed hitters, can see their batting average drop. But more and more batters might be learning to combat the shift. When factoring in all batted ball types — not just grounders and low liners — the MLB batting average on balls in play has remained stagnant. In 2010 — a relatively shift-free season — league-wide batting average on balls in play for all defensive configurations was .297. Last season? .296. The figure has held relatively steady even while scoring and slugging have increased, despite the growing use of shifts. Maybe shifts aren’t such a problem after all.
“The beauty of the game is all the strategies that we can employ,” Milwaukee Brewers manager Craig Counsell said at the winter meetings. So “attacking strategies to win baseball games, man, I just don’t see that as improving the game.”