The shift is one of baseball’s many modernizations. It’s collected many extra outs as well as left pitchers feeling disgruntled with their team’s analytics department. Some readers may have experience with the shift, either as an analyst, pitcher or defender. Myself included. But what often gets ignored is some of the complexity, lack of normality and the actual impact that the shift produces.
The purpose of the shift is to eliminate a batter’s pull side or otherwise the zone where the batter hits the ball the most often. In its most simple form, it is playing the percentages. The only problem is that it’s not working. At least anymore.
Hitters are good at what they do. Major League Hitters are the best at what they do. And with flocks coaches and aides to assist in developing them into even better hitters, it’s hard to imagine batters falling too far behind in their constant arms race against pitchers. It looks like that’s the case here too. While the shift has annually burdened batters by around ten batting average points, they’re now at the point where the shift is having virtually no impact. It’s a real trend, highlighted by the consistency of the three metrics. Not shown, wRC+ also follows the tight trend.
While the shift is a contrarian method, peel back another layer and you’ll find the analytic anti-shifters. One of baseball’s most popular examples of statistical influence is caked with thick layers of deception and misinterpretation. The shift relies less on the hitter’s inability to hit the ball the other way and more on their inability to adjust to seeing an open side of the infield. Or to adjust to a pitcher trying to induce contact (in a time of power pitching). On the flip side, these mind games also plague the pitcher. Leaving that amount of grass open can make them insecure. This spikes walk rates.
There’s an obvious contradiction there, inflated walk rates while pitchers pitch to contact. This is one of those incalculable intricacies with the shift. The strategy’s success is so dependent on the given batter, pitcher and defence it cannot be generalized. Analyzing individual cases doesn’t help to draw conclusions about discreet players, pitchers or batters. There are simply too many factors.
That doesn’t make the data any less intriguing. Some quick analysis of the trends (or lack thereof) and individuals cases demonstrates just how inconsistent the shift is. For one, surprisingly, going opposite field doesn’t always allow a player to evade the shift. The same could be said about ground ball and fly ball percentage.
These instances are exclusive to traditional shifts, so there’s no noise from obscure or irregular shifts. There’s a noticeable lack of cases where players increased the number of batted balls to the opposite field and suffered, so the logic likely isn’t wrong, but otherwise, there’s no strong correlation. Grouping results by pitcher produces a very similar graph, reinforcing the reasoning, but indicating pitchers don’t have any more power than hitters in controlling who reaps (or suffers) the rewards of an effective shift.
Freddie Freeman’s opponents shifted him more than any other player in 2019. Last season he faced a shift in 88.9% of his plate appearances that resulted in a batted ball. But he only pulls the ball 40.5% of the time, 76th in the league. That’s not even the highest on his team – Both Josh Donaldson and Ronald Acuna Jr. pull the ball more frequently. His ground ball rate, another number that should figure into the equation is 38.3% (96th in the league). By no means am I suggesting that ground ball and pull rate are the only factors used to decide whether a team shifts a player or not, but they are large parts of that decision. Especially for the player who was most heavily shifted, those numbers should be higher. To assess the effectiveness of Freddie Freeman shifts, we’ll have to look at his career, as the 2019 no-shift sample is too small.
No one is going to deny the existence of an impact here, but keep in mind, this is the most frequently shifted player of 2019. Although his wRC+ slipped 12 runs, you lowered the probability Freeman gets a hit by 1.2% throughout his career. For context, across the above sample (nine seasons), that’s 4.6 hits a season. Again, if your pitcher and defence are comfortable, great. But otherwise, that’s pretty lowly and insignificant for a hot button issue in baseball.
Here’s another case to put it in perspective. Again, working from the set of players who had 100 plate appearances in both shift and no shift situations in 2019. When facing a traditional shift, no player in Major League Baseball increased the percentage of balls they pulled more than Mike Trout. He pulled balls 14% more against the shift than ordinary alignments, that’s 2.3% more than the next player, Scott Kingery. Here’s where it gets wild. When he’s facing the shift, Mike Trout’s batting average and slugging percentage jumped by 0.029 and 0.032 points respectively. That slugging percentage is a smokescreen, driven by singles as his isolated slugging dropped 0.010. However, wRC+ goes up 16.1 runs. There is one obvious explanation: Trout is hitting more balls in the air. Nope. His ground ball rate increased 13.9%, and while he did hit more line drives (10.3%) the combined total subtracts 24.2% from his flyball rate. Trout is a dream case for the defence, hitting the ball to defenders when they move into strategic positioning, yet somehow, he’s still succeeding. That’s not just fishy, it’s also perhaps the most Mike Trout thing ever.
One of the most beloved parts of baseball is its strategy, adored by traditionalists and analysts alike. The shift is one of the most prominent strategies deployed in baseball. It stands out to the fans, fielders, pitchers and batters. It grabs our collective attention like few other implementations. Yet for all the romanticizing we do, it’s influence on the game remains virtually null. We don’t fully understand how and why the shift impacts outcomes, Mike Trout has proven that. Without the fusion of advanced psychoanalysis and sabermetrics, we may never. But until we do, let’s curb our sponsorship (and usage) of the shift.
All data sourced from Fangraphs.com.