A Roster Platoon Case Study

There are many strategies to employ during a fantasy baseball season and this year I’ve been trying a few new ones. The one strategy that has worked out tremendously well for me so far has been based on Jeff Zimmerman’s early-season analysis of xwOBA’s in-season predictive power. I detailed a practical way to apply his findings to waiver claims in shallow leagues. I’ve also decided, though the decision was sort of decided for me, to punt one category. Some of my own early-season research showed that punting one category can be fine as long as you have a few maxed-out category scores. However, one strategy that I’ve never employed is the usage of platoon splits. Two players, acting like one, and being placed in your lineup as if you were a real manager sounds pretty cool. But, I play in a shallow, 10-team, ESPN roto-league and I wanted to see if this strategy can work in such conditions. Here’s a little case study to see how this strategy would have played out for me in 2021 if I had started at the very beginning of the year and stuck with it until now. 

The first step was to find two players; one who hits best against lefties and one who hits best against righties. Instead of using batting average, I used wOBA (current league average is .314) to get a clearer picture of performance. I took all hitters in 2021 with at least 40 plate appearances in order to include prospects and rookies in the analysis and calculated their difference in wOBA against left-handed and right-handed pitchers. Here are those two lists:

From there I needed to find two players who are not highly rostered (in shallow leagues, that is) and who have comparable numbers. That lead me to Akil Baddoo and Garret Hampson. That was, admittedly, a cherry-pick. But, you can certainly use the tables above to find two players available in your league if you’re interested in doing a similar analysis. Once I had these two players, I used FanGraphs split tools and game logs for performance statistics.

Something to keep in mind; if you were to look up Hampson vs. L on our splits leader boards, you would notice the numbers are different from the numbers I’m presenting here. I’m not isolating Hampson to performance vs. L in this analysis. Instead, I’m looking at how Hampson performed when he was started against a left-handed starter. He may have faced a righty out of the pen later in the game and his performance against that righty is reflected here. You wouldn’t have been able to sit Hampson, once you started him when a righty out of the pen comes in. So I needed to incorporate that into the analysis. Finally, I needed to be sure not to double count. Since Hampson performed a little better overall, I defaulted to Hampson when both players had favorable matchups. To make matters just a little more confusing, I defaulted back to Baddoo whenever there was no favorable matchup. Here are the numbers for comparison.

This means that if you rostered these two players and you were diligent enough to switch them out in the right platoon situation, you would have the stats represented in the ‘Totals’ row. By these numbers, rostering Robbie Grossman and keeping him in the lineup against RHP and LHP would have been a better choice by far. While this is just one case study and it does not show anything on an aggregated level, you can see that this strategy might not be as useful in shallow leagues and would require a lot of attention. You also wouldn’t have had the foresight to know Baddoo and Hampson would be a good platoon combo at the time of your draft. While utilizing platoon situations can certainly be advantageous in deeper leagues, and possibly in small samples, perhaps towards the end of the season, this is one strategy I don’t feel I’m missing out on.