Who Gives a Shift?

October 23, 2014

The 2015 Bill James Handbook will be on bookshelves on November 1, but we wanted to give you a bit of a sneak preview. This is an abridged version of the introduction for the new Shifts section.

Let’s address the main question first. Who does give a shift? Answer: the Houston Astros. Prior to this season, the highest single-season total of shifts deployed by a team was 599 by the Baltimore Orioles in 2013. Six teams exceeded that total in 2014, but the Astros crushed it. The lowly Houston Astros, losers of 92 games, more than doubled the Orioles total with 1,341 shifts this past year.

Over the past four seasons the number of shifts in Major League Baseball, as measured by a ball hit into play when a shift is in effect, has nearly doubled every year. Here are the totals by year:

MLB Shifts by Season
Year Number of Shifts
2010 2,464
2011 2,357
2012 4,577
2013 8,180
2014 13,296


It is clear that Major League teams have embraced this new defensive strategy. Some more than others. But is it effective?

Our best measurement of the effectiveness of shifts is Shift Runs Saved. It is an estimate of the number of runs saved by a team when employing The Shift Defense. It is a technique similar to other elements of Defensive Runs Saved in that it utilizes info down to the play-by-play level, including direction and velocity of every batted ball. Let’s repeat that chart above, but add Shift Runs Saved:

MLB Shifts by Season
Year Number of Shifts Shift Runs Saved
2010 2,464 36
2011 2,357 42
2012 4,577 76
2013 8,180 135
2014 13,296 195


As you can see, the more shifts deployed in Major League Baseball, the more runs are saved. And the lowly Astros with their 92 losses? They had the highest total of Shift Runs Saved by a wide margin. They saved 27 runs on their 1,341 shifts. The next best team was the Toronto Blue Jays with 16 runs saved on 686 shifts.

So the answer to all of this is: Shift or get off the pot! Maybe that should have been the tile of this article. If you want to save runs, shift. Simple as that. And the more you shift, the more you save. Despite all the shifting that is already done in baseball, many teams are still leaving runs on the table by not committing even more to this strategy.


The 2015 Bill James Handbook contains the full version of this article as well as more detailed shift numbers, including batting averages vs. shifts and against traditional defenses. Additionally, The Fielding Bible—Volume IV, coming out in the spring, will provide a more thorough analysis of The Shift Defense.


COMMENTS (10 Comments, most recent shown first)

I am interested in this data by batter. In answer to shtar, this is a major change in the strategy of the game that has reduced runs scored by .04/team/game. In answer to flyingfish, what has changed is data. In the past, unless you were Ortiz or Williams, the tendencies were not well enough known that it made sense, plus there was only anecdotal information on whether the strategy worked. Thanks to John and his colleagues that is no longer true. As to whether it makes sense for the batter to take a different strategy at the plate, doncoffin's analysis suggest that the average shift costs the batter a little over .01 run, while the typical plate appearance is worth .1 runs, so the shifts are reducing their effectiveness by 10%. Adjusting your approach in response to the shift may well cost more than that.
8:38 PM Oct 25th
I really can't see wasting so much time on this.

Especially since the team that does it the most, is the worst team in baseball.

Nobody copies failure.
11:15 PM Oct 24th
The economist in me can't help but put this in an economist's terms. I looked at the average runs saved per shift (RS/Shifts) and the *marginal* runs saved per additional shifts (RS2-RS1)/(Shifts2-Shifts1). As an economist, I'd expect the marginal productivity of shifts to decline as the number of shifts increases--and it does.

Average RS per shift:

Marginal runs saved per additional shift (calculated beginning with the change from 2011 to 2012, because the number of shifts declined in 2011 from 2020):
2011 to 2012.....0.0153
2012 to 2013.....0.0164
2013 to 2014.....0.0117

Obviously there are still potential gains--to say nothing of perhaps more effective strategies for using shifts--but this does suggest that there may be some natural limit to the strategy.

2:49 PM Oct 24th
I'm assuming the "normal" non-shifted positioning of fielders evolved over time because it worked better than other positioning that was being used. That's not to say it was the best possible positioning, only that it was the best of any that had been tried. I can think of two possible reasons for shifts' increasing use: 1. The way people hit has changed, and 2. People just never thought of it before 5 or 8 years ago. Both could be true (or false). I do seem to remember that for many years announcers have said "They are playing Hitter X to pull," referring to outfield positioning. So it seems to me that it's the infield positioning that has changed most. An interesting question (to me) is where this goes in the future. If shifts save enough runs, then hitters will learn to hit the other way. But if, although they do save some runs, they don't save enough runs, then pull hitters will continue to pull.
11:49 AM Oct 24th
Nice title! :-)

Reading your article, I was struck by a similar thought as the one that MarisFan61 had...

OK, there is a demonstrated correlation between shifting and SHIFT runs saved. But how many of those runs would have been saved even had the shift not been employed?

In other words, what is the differential between safe batted balls (ultimately leading to runs) on "no shift" and safe batted balls on "yes, shift?"

9:31 AM Oct 24th
Shift happens.
6:27 AM Oct 24th
.....and come to think of it, I wonder if that factoid came from this work of yours right here.....
9:32 PM Oct 23rd
Heard a stat (or alleged stat) on the radio tonight.
Unfortunately I don't remember it exactly, so I can only give it approximately.....please anyone help out, if you know it better.
I'm giving it because it struck me (at first blush) as pretty remarkable.

Something like, the Giants this year were only about 15th in the majors in number of shifts. (Or maybe even lower.)

But they were 2nd in number of runs saved with shifts.

But actually maybe it isn't that extraordinary, because it's not like doing more shifts is necessarily conducive toward saving more runs. The whole point of not doing a shift at a given time is that you think it would cost runs.....

9:30 PM Oct 23rd
I find the title offensive.......offensive because I didn't think of it.
9:06 PM Oct 23rd
A very shifty article.
8:06 PM Oct 23rd
©2017 Be Jolly, Inc. All Rights Reserved.|Web site design and development by|Terms & Conditions|Privacy Policy