Platoon Splits: Starting and Relief Pitchers

Every fantasy owner looks for an edge over the competition. Using a hitter’s platoon splits versus right- and left-handed pitchers is one such method. When looking at these platoon splits, however, a fantasy owner needs to look a little further than just the basic splits to see if the difference is seen is with all pitchers or just relievers.

Historically, left-handed batters hit right-handed pitchers better and right-handed batters hit left-handed pitchers better (called the platoon advantage). With some players, the extent of the difference can be more or less than the league average. For hitters with extreme splits, it may be best to sit them and play someone else if they are up against a certain handed starting pitcher.

The problem with looking at most handedness data is that both starting and relief pitching numbers are lumped together (as it is here are FanGraphs). Relief pitchers historically pitch better than starters because the hitters are less familiar with them and they are able to throw harder in short stints. This is especially the case with left-handed specialists that are brought in to get one out against a left-handed hitter. When an owner is thinking of sitting a hitter because of a large split, they should make sure that the hitter has problems hitting both starters and relievers. If they just have problems with relievers, they are just as likely to face one later in the game no matter from what hand the starting pitcher throws.

For example, here are Stephen Drew‘s triple slash lines against all pitchers:

vs LHP: 0.247/0.304/0.412
vs RHP: 0.278/0.340/0.453

He does not have a huge platoon split, but it is noticeable. Depending on an owner’s bench, it may seem like a chance to play someone else when he is against a lefty. That is not the case. Here are his stats versus starters:

vs LHP: 0.271/0.331/0.443
vs RHP: 0.269/0.330/0.442

He has almost the exact same stats against lefty and righty starters. It is the left-handed relief pitchers that are giving him problems and he will face them no matter the handedness of the starter.

To further the example, I looked at all the data from 2010 (sorry, I have been a little lazy about updating my database with the 2011 retrosheet files) for right- and left-handed hitters against right- and left-handed starters and relievers. Here are the numbers:

LHH vs LHP Starters 0.248 0.314 0.382 0.696
Relievers 0.231 0.310 0.350 0.660
Difference 0.017 0.005 0.032 0.037
LHH vs RHP Starters 0.263 0.331 0.417 0.748
Relievers 0.254 0.333 0.396 0.729
Difference 0.009 -0.002 0.021 0.019
RHP vs RHP Starters 0.258 0.312 0.405 0.717
Relievers 0.245 0.310 0.379 0.689
Difference 0.013 0.003 0.025 0.028
RHP vs LHP Starters 0.261 0.326 0.409 0.736
Relievers 0.261 0.336 0.403 0.738
Difference 0.000 -0.009 0.006 -0.003

As it can be seen, “LHH vs. LHP” has the biggest split between starters and relievers.

While we don’t track starter and reliever numbers here at Fangraphs, our good friends at do have the data. Goto a hitter’s page, Select Career Splits and scroll down to the Platoon Splits section (link to Stephen Drew’s data). While the reliever data is not there, it can be figured out since the handedness data for All and Starting Pitchers is available.

When considering benching a hitter with a large platoon split, make sure the split exists for hitter against both starters and relievers. This is especially the case with left-handed hitters vs. left-handed starters.

Print This Post

Jeff writes for RotoGraphs, The Hardball Times, Rotowire, Baseball America, and BaseballHQ. He has been nominated for two SABR Analytics Research Award for Contemporary Analysis and won it in 2013 in tandem with Bill Petti. He has won three FSWA Awards including on for his MASH series. In his first season in Tout Wars, he won the H2H league. Follow him on Twitter @jeffwzimmerman.

Sort by:   newest | oldest | most voted

tremendously helpful. i spent all last season benching drew against lefties. well, at least those parts when he was still playing, grumble, grumble..