The Beginner’s Guide To Using FanGraphs Shift Data

Earlier this year, FanGraphs began carrying shift data compiled by our friends at Baseball Info Solutions. With shifting on the rise every year, this kind of information has become more and more vital to fans and analysts. This posts walks you through how to access and use the data available on FanGraphs.

First, let’s start by describing what data is available. We have pitcher and hitter shift data going back to 2010 that is viewable on the leaderboards (for players, teams, and league) and player splits pages. The data exists for balls in play only (non-home run batted balls), so if a team shifts mid-plate appearance we only have the alignment for the final pitch. This also means we don’t have data for walks, hit batters, strikeouts, and home runs. If you want to know about how well a team deploys the shift while on defense, you want to look at pitchers. If you want to know about which hitters get shifted against, you want to look at hitters.

The shift data contains four categories as follows:

Shift – All : This breaks out all shifts, traditional or non-traditional.

No Shift : This breaks out all non-shifted plays.

Shift – Traditional : This breaks out all plays where a traditional shift is employed.

Shift – Non Traditional : This breaks out all plays which would not be considered a traditional shift.

Traditional / Non Traditional shifts are classified as follows by Baseball Info Solutions:

Traditional Shifts:

1) If there are 3 infielders playing on one side of the infield, we consider that a Full Ted Williams Shift.

2) If two players are positioned significantly out of their normal position, we consider that a Partial Ted Williams Shift.

3) If one infielder is playing deep into the outfield (Usually the 2B 10+ feet out into right field), we consider that a Partial Ted Williams Shift. If the 2B is only a few steps into the outfield, that does not count.

Non-traditional shifts are situational shifts not covered under the definition of traditional shifts.

On the leaderboards, you can view shift data by using the “Split” dropdown menu in the center of the options section of the page. While you can’t use more than one split at a time, you can view the data by player, team, or league level and can use the other features like AL/NL, year, age, position etc like normal. Here’s our post on how to use the leaderboards.

Screenshot 2016-06-16 at 1.24.42 PM

For example, you wanted to see which hitter had the best wOBA when putting the ball in play against all shifts in 2014, you could find that in the leaderboards (It’s Giancarlo Stanton).

On player pages, if you navigate to the Splits tab, you can select “Shifts” to view the player’s performance by year against each type of shift:

Screenshot 2016-06-16 at 1.30.26 PM

Or you can look at the default yearly splits and scroll down to the shift breakdown for that year.

Screenshot 2016-06-16 at 1.32.10 PM

The data is interesting and useful, but there are a couple important things to keep in mind. First, the data is limited currently by our ability to view the defensive alignment on video. This means that walks, strikeouts, and home runs lack shift data because the defense doesn’t participate in those plays and the video scouts often can’t determine if a shift was on. However, this does not mean that shifts that occur during those plays are unimportant. Shifting impacts the way a hitter approaches a plate appearance and if they hit more or fewer home runs when the shift is on versus off, that is important data regarding the efficacy of shifting. It might not tell you how well the fielders were placed on the field, but the fact that there was a shift at all matters.

It’s also important, when looking at team and league data, to remember that players are shifted for a reason. You can’t simply look at league/team BABIP or wOBA with the shift on or off to determine if shifting works because the population of shifted plays and non-shifted plays do not contain the same hitters. If you want to do a proper analysis, you need to calculate an expected statistic for each group based on the quality of the players in each weighted by PA.

In a perfect world, we would have data regarding defensive position on every individual pitch, but we’re not quite there yet. For now, you have the option to view shift data at FanGraphs on all balls in play and can do so in a variety of ways. Keep in mind that there are some limits to the data and that when looking at team and league level numbers you have to account for the difference in hitter quality.

If you have questions, feel free to comment or contact me on Twitter @NeilWeinberg44.

Print This Post

Neil Weinberg is the Site Educator at FanGraphs and can be found writing enthusiastically about the Detroit Tigers at New English D. Follow and interact with him on Twitter @NeilWeinberg44.

newest oldest most voted

Hi, I know that shifting hasn’t impacted BABIP, which suggests that hitters adapt, but does it impact ISO%? Does adjustment = less power?