Early Returns on the Shift

It's no secret that the increased use of data by major league teams has led to growing use of the defensive shift.  Rather than calling it a "shift," maybe it would be more accurate to describe the trend as using data and statistics to position fielders for maximum run prevention.  I doubt it comes as a surprise that the Astros frequently changes its fielders' positions for particular hitters.  What are the early results?

Let's start with the caveat that it's a small sample this early in the season, and the results could be different when the season is over.  But we can see what the analyses conclude so far.

DRS (Defensive Runs Saved), the Fielding Bible's advanced defensive metric, calculates a team runs saved value for defensive shifts.  The shift value is shown on the Bill James-Online web site (note that the stats are behind a subscription pay wall.)

So far the 2014 results are very good for the Astros.  DRS calculates +6 net runs saved from the shift. The next closest AL team is the Orioles with +5 runs.   The Brewers lead the NL with +4 runs.

In this New York Post article, Baseball Information System (BIS) vice president Ben Jedlovic describes the net savings calculation procedure:

Ben Jedlovec, BIS’ vice president of product development and sales, explained how the "shift runs saved" are compiled. BIS analyses track each hit and measure how often the "standard" defense would have made a play as opposed to a shift defense.

"Let’s say there’s a ball that’s in the hole between first and second. Without the shift, the defense will get the out 50 percent of the time," Jedlovec said. So if the shifted infield makes the play...then you reward the team with a half-play, which amounts to about one-third of a run.

Now, on the flip side, if a lefty batter strokes a grounder to shortstop, that’s an out 95 percent of the time with a standard defense. So if that hit beats the shift, you penalize the team by 95 percent of a play, which is about -.72 runs...

"Of course," Jedlovec noted in an email, "this does not include the mental impact of a shift, often throwing opposing hitters off their offensive comfort zone."

BIS reports that, in the recent past,  the number of defensive shifts have roughly doubled every year..Teams shifted over 8,000 times last year, and are on pace for over 12,000 shifts this year.  The Astros lead the majors in shifts so far, and are on pace for a ten fold increase in the annual number of shifts used in 2014.  The Yankees are second in number of shifts so far this year.

The TCB Daily Link on Thursday discussed possible causes of the MLB's offensive decline. Jonah Keri writes:

...teams have discovered run prevention gold with shifts. Clubs like the Astros will probably shift even more as long as doing so yields results, and teams that almost never shift, like the Rockies, will likely start once they’re forced to acknowledge how well the tack is working for the competition. There are ways to beat the shift, like shooting for the opposite field or bunting, but we’ve yet to reach the tipping point where a power hitter who sees a lot of shifts, like David Ortiz, decides to sacrifice home runs for singles. We might not even be close.

Is "run prevention gold" hyperbole?  Maybe, the leader in shift runs saved last year (Rays) gained a little less than two wins, according to DRS--which isn't an overwhelming effect.  However, the article makes a good point that the analytic data will only improve in the future, which should increase the effectiveness of shifts.

The Astros' shift tactics also spurred a Fangraphs analysis of pitching against the shift.

Fangraphs used a batter/pitcher match up between Dallas Keuchel and Jose Bautista to demonstrate that sometimes it makes sense to pitch against the shift. In other words, in this case, the Astros both put a shift on, and deliberately pitched the batter outside, or against the shift.  The article shows that this was a rational strategy, underscoring the intricacy of run prevention tactics.  The data shows that Bautista rarely hits groundballs the other way, even against an outside pitch.  Pulling an outside pitch will often cause the batter to roll over the ball--generally a good outcome for the pitcher.  Thus, simueltaneously putting the shift on, and pitching outside, plays the percentages.  This tactic may not work on all batters, but should be based on the specific hitter's tendencies.

The New York Post article, linked above, quotes sluggers who refuse to change their approach (going the other way) against the shift.  And it may well make sense for sluggers to maintain their approach, hitting into the teeth of the shift defense.  In 2013 the slugging percent on pulled balls was about 300 points higher than balls hit to the opposite field..  Hitting the ball in the air is how these hitters do their damage. (Batting average is 70 - 100 points higher for pulled balls too.) It usually doesn't make sense to give up 300 points of slugging in order to improve the odds of getting a groundball through the infield.  Ted Williams, who refused to change his approach against the shift, instinctively knew this.  And the defensive team may feel that it has achieved victory if it can turn a slugging HR hitter into a slap hitter.

The shifts can also affect the pulled line drives which fall for hits.  By putting the fielder where the hitter hits the ball hardest, the number of line outs should increase. Although groundball distribution may be the primary factor in designing a shift, often a pull hitter's line drives are more likely to be pulled in the same direction, where lower trajectory liners can be caught by a well positioned fielder.   Consider the trend in BABIP on line drives in the AL during the period of increasing shifts.

BABIP on Line Drives

2009  .730

2010  .716

2011  .718

2012  .711

2013  .665

Probably a number of factors contribute to the cause of the downward trend in line drives falling in for hits. But it could suggest that teams are defending line drives better.  The improvement could involve both infield shifts and better positioning of outfielders.