Batted Ball Types and Handedness Matchups, in General

Last month, I did a two-part analysis that showed what happens — strike out-wise — when, say, a pitcher who strikes out 15% of batters faces a batter who strikes out 20% of the time. As a special bonus for you all, I included a few hundred other K%-matchup types too. I made handedness matchups central to the study, as I think it’s pretty well-established that you can expect a hitter to strike out more often against same-handed pitchers. That is, if I was trying to give an expected result for a righty batter against a lefty pitcher, I looked only at the hitter’s past performance rates against lefties and the pitcher’s history against righties. Before I moved on to performing a similar analysis on batted ball types (grounders, liners, outfield fly balls, and infield popups), I wanted to see whether handedness matchups mattered to these as well.

For this study, my sample was all non-switch-hitting batters from 2002-2012 with at least 300 PA against lefty pitchers plus at least 300 PA against righties. I’d have gone by number of batted balls, except I’m throwing some non-batted ball stats into the analysis.

Let’s get right to it — the following table shows the chances that handedness really makes no difference to each stat, according to paired t-tests:

Paired t-test results
PU% 2.21E-09
BABIP 5.39E-11
LD% 1.77E-22
HR/FB 2.23E-25
IFFB% 1.76E-25
H/PA 3.48E-31
FB% 1.37E-39
HR/PA 4.18E-41
OFFB% 6.36E-59
HBP/PA 8.31E-62
GB% 6.37E-73
IBB/PA 1.48E-75
wRC+ 8.32E-84
wOBA 2.04E-85
SO/PA 9.56E-86
BB/PA 2.46E-99

“2.21E-09” means 0.00000000221, with “221” starting 9 places after the decimal (also known as 2.21*10^-9).  Specifically, those are the chances that the means of each stat for batters are truly the same against opposite-handed pitching as they are against same-handed pitching.  Interpretation: it’s pretty certain that pitcher handedness makes a difference to the batter for all of these stats; however, the t-test is a lot more sure that handedness makes a difference when it comes to walks per plate appearance than when it comes to popup percentage (PU%, which I define as infield fly balls per batted ball, as opposed to IFFB%, which FanGraphs defines as infield fly balls per fly ball).

That doesn’t exactly answer the question of how consequential the handedness-related differences are, however.  To shed some light on that, here’s the following table:

Stat vs. Opposite Minus vs. Same League Average (O-S)/L Standard Deviation in (O-S)/L Signal-to-Noise Ratio
PU% -0.36% 3.81% -9.36% 33.87% 0.276
BABIP 0.009 0.2973 2.91% 9.56% 0.304
LD% 1.07% 19.96% 5.37% 11.55% 0.465
HR/FB 1.50% 10.43% 14.37% 28.77% 0.500
IFFB% -1.57% 10.54% -14.94% 29.84% 0.501
H/PA 1.36% 23.37% 5.83% 10.32% 0.565
FB% 2.34% 36.12% 6.49% 9.91% 0.654
HR/PA 0.59% 2.68% 22.10% 33.00% 0.670
OFFB% 2.70% 32.31% 8.34% 9.88% 0.845
HBP/PA -0.61% 0.91% -66.57% 76.30% 0.872
GB% -3.39% 43.93% -7.71% 7.89% 0.978
IBB/PA 0.80% 0.69% 115.47% 115.16% 1.003
wRC+ 20.8 100.0 20.8% 19.2% 1.081
wOBA 0.031 0.3256 9.52% 8.69% 1.096
SO/PA -2.73% 17.50% -15.62% 14.21% 1.099
BB/PA 2.31% 8.46% 27.35% 22.26% 1.229

In the “vs. Opposite Minus vs. Same” column, we have the average batter’s results against opposite-handed pitching (e.g., against lefty pitchers for a righty hitter) subtracted by their results against same-handed pitching.  Since the scales between all the stats are very different, I divided these by their MLB averages for 2002-2012 to make this more of an apples-to-apples comparison; those results are in the (O-S)/L column (Opposite minus Same, divided by League average).  Next is the standard deviation of the players’ (O-S)/L values for each stat — a measure of how spread out the values are (higher being more spread).

In the last column, we have the absolute value of the mean (O-S)/L divided by the standard deviation, which is sometimes called the “signal-to-noise ratio.”  The idea there is that the larger the mean difference is in relation to the standard deviation, the more seriously you can take it.  Notice the signal-to-noise ratio (SNR) ranks the stats the same way the t-test does (you’ll see that’s not a coincidence, when you look at the formula for paired t-tests).

Hopefully I won’t muddy the issue any more by throwing correlations into the mix.  The following table will give you an idea of how predictable a player’s stat against a same-handed pitcher is, given their performance against opposite-handed pitchers (or vice versa).  A correlation of 1 would mean it’s perfectly predictable, while a 0 would be completely unpredictable.  I’m allowing you to change the confidence interval on these correlations, but by default, it’s telling you the minimum and maximum possible correlation with a 95% degree of certainty:

This seems to me to line up pretty well with the year-to-year correlations for these stats (although they’re a little higher here).  What we’re seeing here probably has more to do with the nature of the stats themselves than it does with handedness phenomena, it seems to me.


Strikeouts, Walks, wOBA and wRC+ — four of the most critical stats to consider, and the four to which handedness matchups apparently make the most convincing difference.  Not the most difference — that would seem to be intentional walks and hit by pitches — but IBB and HBP rates are too wild (thanks to their rarity, for one thing) to be sure of that.

Let’s break down some of the findings, and we can discuss why they do or don’t make sense:

  • Pitchers intentionally walk more opposite-handed batters (OHB):  in light of the 20-point wRC+ advantage opposite-handed batters in general enjoy, you can see why managers would sometimes prefer the pitcher go after a same-handed batter next in the order.
  • A same-handed pitcher (SHP) is more likely to plunk a batter: I would guess this has to do with the trajectory of SHPs generally being more difficult to judge for batters, making it more difficult for them to get out of the way in time.
  • Batters walk more often and strike out less often against opposite-handed pitchers (OHP): besides the hitter having an easier time seeing the ball and its motion, the pitcher likely shows more caution.  Arguably, the motion of some pitches might be more difficult to make contact with regardless of vision issues.   A lot of the other stats can likely be explained by these factors and their effect on strikeout rates, counts, etc.
  • Batters hit more line drives against OHP (per batted ball): they see the ball better and get into more hitters’ counts, and this apparently translates into more line drives.
  • Batters hit fewer ground balls against OHP (per batted ball): well, for one thing, if they’re hitting more liners, that leaves fewer batted balls going to the other batted ball types.  As you’ll see next…
  • Batters hit more fly balls against OHP (per batted ball): and it’s a more dramatic difference than the line drive difference.  I think batters are swinging for the fences more, which I believe is supported by the HR/FB numbers also being higher.
  • Batters hit fewer popups against OHP (per batted ball): despite hitting more fly balls, fewer of them are hit poorly (too low on the ball, in this case).
  • Batters hit for a higher BABIP against OHP: this in in spite of hitting more fly balls, which tend to have a lower BABIP.  The combination of more liners and fewer popups apparently more than makes up for that.

Hope you enjoyed.  Let me hear your thoughts.

Steve is a robot created for the purpose of writing about baseball statistics. One day, he may become self-aware, and...attempt to make money or something?

Newest Most Voted
Inline Feedbacks
View all comments
10 years ago

“See I don’t overmanage!” – Joe Girardi