It’s the bottom of the 7th inning, and with two outs, Brandon Phillips has just singled home a run off Cardinals’ righty Matt Belisle. With the score now 4-3 for the Reds, trivially, Mike Matheny, the Cardinals’ manager wants to limit the damage with the lead runner now standing on third base. Jay Bruce is on deck and he is envisioning ripping a breaking ball down the line of the soft-throwing righty. But he knows he will never get the chance. Almost without hesitation, the left-handed Kevin Siegrist replaces Belisle, and Bruce goes down swinging.

Platoons are not only increasingly common in today’s game, but they are now instinctive. Teams will utilize different relief pitchers or bench players to capitalize off the opposite handedness of the opponent. They will platoon a pair of left fielders who bat from the opposite side of the plate; starting the righty against the lefty pitcher and vice-versa. Every team does this to some degree – it’s not only the Moneyball A’s, Joe Maddon’s ex-Rays, and C. Montgomery Burns’s Springfield Nuclear Power Plant softball team. Splits are interesting, and it would be foolish to not get the most out of matchups.

As a fan of the game, I find myself playing with split functions on Fangraphs and Baseball Reference constantly. As a fan of the Jays, it is essential to understand why Danny Valencia should really only ever hit against a left-handed pitcher. And fortunately, this is easy figure out. On Baseball Reference, there are split functions that had me posing questions. Aside from the inherent handedness-platooning, are there not additional ways to exploit matchups?

Right off the bat (pun intended), we can eliminate rather goofy split and potential platoon ideas. No manager who wants to keep his job is sitting out Jason Kipnis because he is a “first half” guy, nor are they waiting to put in Troy Tulowitzki because he is “really good in September”. Even the idea of some players being significantly better playing at night or during the day is probably just noise. The first of the two I found as worthwhile exploitable splits is hitting against Ground Ball pitchers versus Fly Ball pitchers.

Per Baseball Reference:

Fly Ball pitchers are in the top third of the league in ratio of fly ball outs to ground ball outs. Ground Ball are in the bottom third of the league in the ratio of fly ball outs to ground ball outs. Stats are based on the three years before and after (when available), and the season for when the split is computed. A split in 1994 would consider years 1991-1997 when classifying a pitcher.

Thus by definition, pitchers can be grouped into three parts: the Groundballers, the Flyballers, and the neutral group (avg. F/G). Not only is this easy to understand, but the 33% of pitchers in each group leads to a significant sample size of Plate Appearances each year. Essentially what this means is that even with younger players, it will not take a tremendous amount of time to get a large sample of Plate Appearances. I am not explicitly saying that these ground ball/fly ball splits have incredibly strong predictive power, but the samples are certainly sizeable.

I immediately cut off the players who did not have 600 Plate Appearances against Fly Ball pitchers or Ground Ball pitchers, to not cloud the results. Using the Play Index, I then sorted the data by the difference in their OPS against the split compared to their career OPS overall. To my chagrin, the results did not show any enormous differences, though on the other hand, intuitively this type of split or platooning would have less effect than our standard righty/lefty one anyway. Here are the top 5 players who hit better against Ground Ball pitchers:

Rk Player Split PA PAtot OPS OPStot Diff
1 Chris Davis vs. GrndBall 631 2863 0.963 0.815 0.148
2 Dan Uggla vs. GrndBall 1389 5386 0.906 0.786 0.120
3 David Murphy vs. GrndBall 834 3466 0.881 0.768 0.113
4 Freddie Freeman vs. GrndBall 710 2641 0.923 0.830 0.093
5 Chris Young vs. GrndBall 1114 4341 0.815 0.742 0.073


As you can see, the differences in OPS are not incredibly high, with only 3 players have more than a 100 point difference in the stat. I also left out these players’ BABIPs, which, while they were higher than their career norms, ground balls generate notably more hits than fly balls do, making this a reasonable explanation. However, the names also draw some confusion. Chris Davis is a power hitter. Dan Uggla is a power hitter. Nonetheless, Chris Davis, for instance, has proven to hit much better against the ground ball specialists, with a career tOPS+ (OPS of this split against career OPS, with 100 being average) of 134 against them versus an 86 tOPS+ against fly ball pitchers. Although he represents the type of player good enough to play every day, perhaps the best times to give him rest days are not against lefties, where he has a 91 tOPS+, but against fly ball pitchers. For Uggla, who at this point in his career has proven incapable of being an everyday player, could be better suited lining up against ground ball pitchers more often than not.

Here are the players that have pounded Fly Ball pitchers relative to their norm:

Rk Player Split PA PAtot OPS OPStot Diff
1 Yunel Escobar vs. Fly Ball 1572 4444 0.814 0.728 0.086
2 Buster Posey vs. Fly Ball 740 2485 0.939 0.862 0.077
3 Kyle Seager vs. Fly Ball 940 2225 0.830 0.754 0.076
4 Troy Tulowitzki vs. Fly Ball 1371 4088 0.966 0.892 0.074
5 Casey McGehee vs. Fly Ball 874 2754 0.794 0.724 0.070


Likewise, the OPS differences are not astronomical as we go down the leader board, but they still are helpful. I want to note that I did not plan these results, nor did I tamper with them. I had no intention or belief that the Washington Nationals second base situation (with Rendon healthy) resides near the top of each leaderboard. Regardless, Yunel Escobar has a 123 tOPS+ against Fly Ball pitchers, while having a putrid 66 tOPS+ against Ground Ball pitchers. Not only that, but Escobar’s sample against the Fly Ball group is quite large, in fact it is the largest among these 5 players. Like Chris Davis, Posey, Seager and Tulowitzki are all regular players at the All-Star level, so this likely only helps in deciding when to rest them, but it is interesting to observe.

Most of the names that came up in this analysis were not extremely helpful, and in many cases, including power hitters being better against Ground Ball pitchers, baffling. However, the Yunel Escobar and Dan Uggla situation is very promising. Both have a good enough sample size against either type of pitcher and pair perfectly in terms of this type of split. Not only that, but they both hit right-handed, making the idea of a handedness platoon near impossible, especially if you don’t fully buy into reverse platoon splits. Perhaps I am getting ahead of myself, but when Anthony Rendon returns to take third base back from Yunel, I would love to have Escobar hit the Fly Ball guys and Uggla hit the Ground Ball guys while playing second base.

As mentioned previously, I am uncertain how much predictive power this type of hitting has, year over year, especially since in season there is not enough of a sample size to draw results. Although, like we have seen, with enough PAs conclusions can be drawn and interesting platoon proposals exist. In the second part of this analysis, I will look at a different type of split; another way to exploit hitter/pitcher matchups.

Next post:
Previous post:

Leave a Reply

Your email address will not be published. Required fields are marked *