Studies on the Platoon Affect
"It is an old strategy, dating back to at least 1906. It plays a part in every game. It defines careers, it directs strategy. Every manager uses it to a greater or lesser extent. It helps decide who is in the starting lineup, when pitching changes are made, when intentional walks occur. It virtually dictates which pinch hitters can be used and when they can be used. And yet, remarkably, we know almost nothing about it. We don't know who invented it. Ten years ago there were no reference sources for information about platoon differentials, so that we couldn't find out who hit what against left-handed and right handed pitching. How many players actually hit better the way they are supposed to, and how many hit better the other way? How large are the differences? Is there a "standard margin" or is it widely variable? What is the league average for left-handed hitters against left-handed pitchers, for right-handed hitters against right-handed pitchers, etc? What players have the largest platoon differences, and what players hit better the other way?"
This is the opening paragraph of a study called "PLATOONING" by Bill James, that appeared on pages 9 through 15 of The 1988 Bill James Baseball Abstract.
The study looks at the 119 players that had more than 400 plate appearances against both right-handed and left-handed pitchers from 1984 to 1986. The study showed a platoon advantage in terms of OB + SLG of 87 points. As he points out, the 400 plate appearance requirement standard biases the study.
"Why? Because it cuts out most of the players who are platooned. Take Wally Backman, for example. Backman over the three-year period batted 1,122 times against right-handed pitchers and hit .316 against them. Because he hits so well against right-handers, the Mets tried in 1985 to make regular out of him, but he's just hopeless against left-handed pitching. Over the three seasons he batted 221 times against left-handed pitchers and hit a robust .149 against them with a .190 slugging average. Because he can't hit left-handers at all, he doesn't play against them except on rare occasions, such as when every other infielder in the Mets system has to appear in court or something. So he doesn't have 400 plate appearances against lefties, so he doesn't make the study - precisely because he has a large platoon differential."
"So the true platoon differential is probably somewhat larger and even more consistent than it is measured as being in this study."
He points out that the study showed the platoon differential as varying widely from player to player without a pattern as to the type of hitter (i.e. free swinging slugger, high average hitter etc.). He answers the question of weather platooning make sense as a strategy.
"Is platooning justified by the margins that can be gained? Quite certainly, yes. In this study, the players who hit .290 or better had an overall batting average of .304 and an OB + SLG of 823. The players who hit .250 - .269 hit .260 overall and had an OB + SLG of 744. But with the platoon advantage the .260 hitters were better offensive players than the .304 hitter were without it (791 vs. 781). No manager could possibly ignore an advantage like that - and there probably should be more position platooning than there is."
A later Bill James article that discussed the platoon advantage and also talks about the hazards of small sample sizes is at:
http:www.sabr.org/cmsfiles/underestimating.pdf
Another good study on the platoon affect, by David W. Smith is at: