June 30, 2005
Left Wing Conspiracy
Towards the end of last season, I was digging up some data involving platoon splits and noticed that back in the 1980s and early 1990s, the number of lefty-lefty matchups was a lot higher than it is now. That led to an article showing that left-handed pitchers are pitching a significantly smaller percentage of the available innings (or PAs) than they were just 10 to 15 years ago. In 1991 nearly 34 percent of PAs were against LHPs; in 2002, it was under 24 percent.
The article prompted interest from a few people in baseball who were doing studies with regards to handedness and performance. We talked about why there are only "crafty" left-handed pitchers and not righties (unless you count Greg Maddux) and why baseball's most aesthetic swings (including Will Clark, Ken Griffey Jr., and Ben Grieve) were all performed by southpaws. But mostly, we talked about whether or not there's an inherent advantage provided to one hand or the other when it comes to baseball players.
For example, check out the top 30 batters in career VORP, 1901-2004:
Rank Name VORP Bats Throws ---- ------------ ------ ---- ------ 1 Babe Ruth 1564.4 L L 2 Barry Bonds 1406.1 L L 3 Ty Cobb 1396.3 L R 4 Stan Musial 1277.0 L L 5 Lou Gehrig 1246.3 L L 6 Rogers Hornsby 1234.2 R R 7 Ted Williams 1205.0 L R 8 Hank Aaron 1192.1 R R 9 Willie Mays 1110.2 R R 10 Tris Speaker 1105.0 L L 11 Eddie Collins 1080.2 L R 12 Jimmie Foxx 1071.0 R R 13 Honus Wagner 1037.9 R R 14 Mel Ott 934.1 L R 15 Mickey Mantle 911.1 S R 16 Frank Robinson 907.1 R R 17 Charlie Gehringer 892.0 L R 18 Joe Morgan 876.7 L R 19 Frank Thomas 806.4 R R 20 Rickey Henderson 800.4 R L 21 Jeff Bagwell 787.6 R R 22 Cal Ripken Jr. 776.3 R R 23 Pete Rose 769.5 S R 24 Mike Schmidt 768.2 R R 25 Wade Boggs 757.2 L R 26 George Brett 757.1 L R 27 Rafael Palmeiro 751.9 L L 28 Eddie Mathews 748.8 L R 29 Harry Heilmann 746.0 R R 30 Carl Yastrzemski 722.9 L R
If you're looking at their batting hand, that's 16 lefties (53 percent), 12 righties (40 percent), and two switch-hitters (seven percent). By throwing hand, that's seven southpaws (23 percent) and 23 people like me (77 percent). (It's clear that there's a difference between players who are truly left-handed and those who simply hit from the left side. For the purposes of this article, we'll refer to players who throw left-handed as "true left-handers" as opposed to left-handed or switch hitting batters.) Depending on who you talk to, the worldwide percentage of people who are left-handed is between eight and 14 percent, meaning that of the top 30 batters of all time, two to three times as many are left-handed than the general populace.
Of course, baseball is a game that seeks out players with an advantage; southpaws have long captivated scouts and executives because of their rarity. Thus, baseball boasts a much higher percentage of both "true left-handers" and batters who learned to bat from the left side. In 2005--excluding switch hitters--35 percent of the PAs have gone to left-handed batters. The percentage of regulars (players accumulating at least 450 PAs in a season) who are lefthanded has hovered between 36 and 38 percent for the past 15 seasons. And yet 53 percent of the top 30 hitters of all time are left-handed. At the very extreme, there appear to be a disproportionate number of elite left-handed hitters.
So left-handers are better than right-handers at batting, so what? Everybody knows that, right? They have an inherent platoon advantage since the majority of innings are thrown by right-handed pitchers. But is that platoon advantage so strong that it propels otherwise equal players significantly higher in the game's elite? Or are left-handed batters more adept at the skills required in baseball?
To find out, we have to somehow isolate the performance from the platoon. Not unlike park factors, we're trying to extrapolate how batters would compare to other batters if they all batted from the same side of the plate. To do this, rather than looking at players as left- and right-handed, they'll be considered by their comparison to the pitcher on the mound. Thus we can break each batter's PAs into two groups, one in which they faced a pitcher of the same hand as they bat, and one in which they face the opposite. Switch hitters will be excluded for now.
For example, through Tuesday, Hank Blalock has faced lefties 85 times, righties 240, hitting .284/.318/.457 against the former, .285/.363/.495 against the latter. However, if we weighted Blalock's PAs and ABs to the league-average playing time against same- and different-handed pitchers, instead of hitting .285/.351/.485, Blalock would be hitting .284/.338/.474. It's not an outstanding difference, but still, it appears that some of Blalock's advantage comes from facing pitchers of the opposite hand a greater percentage of the time.
Taking the process league wide for 2005, currently lefties are hitting .266/.341/.422 while righties are managing .261/.322/.417. Adjusting all PAs so that players face the same number of same- and different-handed opposing pitchers, LHBs as a group now hit .260/.331/.403 while righties bat .262/.326/.423. Given that left-handed batters are suddenly hitting worse than right-handed batters, perhaps their superiority is merely tied up in their platoon advantage.
There is one small bump in the road of conclusions with analysis, though: we're assuming that left-handed batters would continue to hit as well against a larger sample of left-handed pitchers as they do currently. For example, this season, left-handed batters are hitting .252/.317/.381 against fellow southpaws while knocking around righties to the tune of .271/.349/.435 for a platoon split of .019/.032/.054. By contrast, righties hit .257/.315/.406 and .271/.339/.444 for a split of .014/.023/.038. As a result of left-handed batters' larger platoon split, any analysis forcing them to hit more often against fellow southpaws is going to reduce their advantage more than if they showed a smaller platoon split.
Thus, the question becomes: Why do left-handed batters consistently show a larger platoon split than right-handed batters from year to year? This question is one that's very difficult to answer, but generally a theory with two points has been voiced: Left-handed batters, like all batters, face an inherent disadvantage when batting against same-handed pitchers. However, because there are so many more right-handed pitchers than left-handed pitchers, particularly in the minors, left-handed batters never learn to hit against same-handed pitching quite as well as right-handed batters, so their advantage appears greater than it actually is.
If this theory is true and left-handed batters would see a reduced platoon split based on more experience against same-handed pitching, then the analysis above is unfairly biased against them (of course, considering how hypothetical the analysis is, it's anybody's guess at this point). So instead of using how left-handed batters actually did against lefties and righties, let's instead assume that lefties will show the same platoon split as righties, centered on their yearly averages.
Again using Blalock as an example, the Texas third baseman hit .285/.351/.485, with platoon splits of .284/.318/.457 and .285/.363/.495 for a difference of .001/.045/.038. Now, assume that Blalock shows the typical right-handed platoon split from 2005, .014/.023/.038. Our hypothetical Blalock will now hit .289/.357/.495 against righties and .275/.334/.457 against lefties. This new Blalock is our estimation of his performance if he was a right-handed batter with typical platoon splits with the identical distribution of his PAs and ABs against same- and different-handed pitchers. Of course, now we can adjust his PAs and ABs to see how he does facing many more same-handed pitchers than he does in real life. Doing so yields a line of .281/.344/.474 (as opposed to the initial adjustment of .284/.338/.474). This is our expected performance for Blalock given typical right-handed platoon splits and a PA and AB distribution close to league averages based on same- and different-handed pitching.
Running a similar analysis on the entire league, right-handed batters now come out with a composite line of .274/.339/.438 for the season while lefties manage .273/.349/.431; the two sides post nearly identical OPS: 777 to 780. Interestingly, the groups come out with very similar numbers with right-handers managing a slight edge in power, but lefties holding an edge in OBP. From this analysis, it appears that most of the "edge" gained by the left-handed batters has disappeared, but remember that this analysis is based on the assumption that left-handers don't face an inherently wider platoon split as a result of ability but rather as a result of the normal left/right distribution in the general populace. It's quite possible that there is something about being left-handed (and consequently, right-brained) that grants people a slight edge when it comes to hitting a baseball. However, it appears that if left-handed batters had to face similar playing time as right-handed batters against same-handed pitchers, they wouldn't see the same consistent advantage they now enjoy.