Don’t tell anyone, but I really enjoy watching Randall Simon hit. The loose, goofy motion in his stance as the ball approaches the plate; the flyswatter swing; the big-stepping follow through, his blubber, after half a second in gelatin-like suspension, mimicking the motion of his bat. It’s a lot of fun to watch, especially when Simon manages to make contact, which happens more often than you’d ever expect.

I’ve had the occasion, however, to watch Simon against Kerry Wood a couple of times this year, and from Randall’s point of view, the results have been disastrous: zero-for-six with four strikeouts. Not just any kind of strikeouts, mind you, but ugly, pirouetting, breeze-generating, no-chance-in-hell strikeouts, the sort that make you think that Simon could face Wood 500 times and go oh-fer.

I didn’t mind this, really; Wood is one of my favorite pitchers. But this particular matchup was interesting to watch because Simon and Wood are such an odd couple: Simon swings at everything, and never draws any walks, but by virtue of his superior hand-eye coordination, manages to keep his strikeout rate very low. Wood, on the other hand, is one of the toughest pitchers in the league to make contact against–though sometimes that’s because he isn’t throwing the ball anywhere near the strike zone. In any event, Simon’s performance against Wood looked so bad than I began to wonder whether the batter isn’t at some sort of systematic disadvantage in pairings of these types of players.

To study the question, I’ll leverage from a technique that Gary Huckabay and I introduced last month in a 6-4-3 column, comparing the actual performance observed when certain types of batter-pitcher pairings occur against the results predicted by Bill James’ log5 formula. Instead of dividing players up based on groundball and flyball rates, this time we’ll look at a quick-and-dirty index of plate discipline. Ladies and gentlemen, introducing PDQ:

PDQ = SQRT(BBRate * KRate)

PDQ, or plate discipline quotient, is the geometric mean of a player’s walk rate and his strikeout rate. I have some hesitancy about grouping strikeouts and walks together under the same metric–the research I’ve done for PECOTA suggests that the characteristics operate somewhat independently of one another: an ‘impatient’ hitter can find himself with either a low strikeout rate (think Nomar Garciaparra) or a high one (think Cory Snyder), depending on the other skills that he brings to the table. Nevertheless, this metric has the virtue of convenience, and using a geometric mean instead of an arithmetic one ensures that we identify players who both strike out and walk a lot (or a little), rather than being exceptional in one category but average in the other.

I went about dividing up hitters and pitchers based on their 2002 PDQ’s as follows:

  • ‘Finesse’ player: PDQ of .10 or less

  • Neutral player: PDQ of between .10 and .14

  • ‘Power’ player: PDQ of .14 or greater

These designations are generally consistent with conventional usage. In various permutations, the ‘Power’ and ‘Finesse’ labels have been used to classify pitchers for years. They work pretty well for hitters, too; ‘Power’ hitters do in fact hit substantially more home runs, while ‘Finesse’ hitters manage higher averages and more stolen bases. The cut-offs used here were chosen such that a ‘Power’ player’s PDQ is roughly one standard deviation greater than the league average, and a ‘Finesse’ player’s PDQ roughly one standard deviation less than it. (OK, so Simon gets classified here as a ‘Finesse’ player. No system of nomenclature is perfect; if it helps, try and think of a graceful, curvaceous Randall taking the field in a black-and-gold tutu).

There seems to be no determinate advantage to either ‘style’ of pitching–here’s how Power and Finesse pitchers compared in 2002:

Pitcher      Power      Finesse    All Pitchers
BA            .241        .276            .261
OBP           .339        .327            .333
SLG           .387        .434            .417
KRate         21.9%       13.3%           17.1%

By virtue of their higher strikeout rates, the Power pitchers allow fewer hits and lower batting averages, which is enough to make up for their inconsistent command. Neither Power nor Finesse pitchers showed a marked tendency to serve up the long ball more often.

Among batters, however, Power players are at a demonstrable advantage:

Batter        Power       Finesse    All Batters
BA              .257       .265            .261
OBP             .361       .312            .333
SLG             .455       .390            .417
KRate           22.4%      13.6%           17.1%

There’s quite a strong relationship–some of it causal, some of it circumstantial–between taking more pitches and generating more power. ‘Power’ hitters hit for a somewhat lower batting average, but are otherwise considerably more productive than their Finesse counterparts.

The plate discipline horse has been whipped more often than Funny Cide, however, so we’ll set it aside for now. We’re more interested in what happens when freak show matchups like Wood v Simon occur. In 2002, there were more than 13,000 plate appearances between Power pitchers and Finesse batters, as we’ve classified them above. Here’s how those plate appearances played out, as compared with the results predicted by the log5 method.

Power Pitcher v Finesse Batter, 2002

		Actual       Expected
BA              .244            .244
OBP             .311            .318
SLG             .362            .361
KRate           18.0%           17.7%

Finesse batters fared poorly when matched up with Power pitchers, managing a line of just .244/.311/.362 against them. However, they didn’t fare much more poorly than we would have expected them to based on the overall statistics of the two groups–a slight decrease in OBP was the only discrepancy. The log5 model, in fact, predicted their performance quite accurately.

What about the opposite pairings–Finesse pitchers against Power batters? How would Adam Dunn fare against, say, Rick Reed?

Finesse Pitcher v Power Batter, 2002

	        Actual      Expected
BA              .278          .271
OBP             .356          .355
SLG             .487          .473
KRate           17.6%         17.7%

There appears to be a slight advantage accruing to the offensive player here. Although he takes somewhat fewer walks than predicted–look at the difference between BA and OBP–his batting average and isolated power are a notch above expectations.

Though we’ve conceived of them as opposites, Power batters and Finesse pitchers share a propensity to work the count–the Finesse pitcher trying to get ahead of the hitter, the Power batter waiting for that juicy slider on 3-1. It may be that a Power hitter is a little bit less susceptible to the pitcher’s trickery than the other way around. Then again, the advantage is slight, and probably not meaningful enough to make lineup decisions.

What about the other two permutations?

Power Pitcher v Power Batter, 2002

		Actual       Expected
BA              .239          .236
OBP             .371          .367
SLG             .429          .425
KRate           27.6%         28.2%

Contrary to popular belief, there aren’t a disproportionate amount of home runs generated in power-on-power matchups Rather, the defining characteristics of these encounters are a low batting average and a high walk rate–but not any lower or higher than was predicted by log5.

Completing the circle:

Finesse Pitcher v Finesse Batter, 2002

	       Actual        Expected
BA              .280           .279
OBP             .313           .306
SLG             .407           .407
KRate           10.2%          10.5%

Finesse batters recorded a slightly higher walk rate than would have been expected when paired up against Finesse pitchers, but otherwise the predictions are dead-on.

You’ll hear managers and announcers talk a lot about matchups; a 1-for-7 track record against a given pitcher is supposed to spell certain trouble for the hitter, while a 2-7 tally means he hangs in there just fine. Well, I’m exaggerating: I don’t mean to suggest that this sort of analysis is meaningless. For one thing, there are some pressure points where the particular characteristics of the batter and pitcher do seem to make a difference–disciplined hitters hit for slightly more power than expected against finesse pitchers, flyball hitters enjoy, probably, a slight advantage against groundball pitchers. For another thing, the fact that these sorts of trends are weak in the aggregate doesn’t mean that individual matchups don’t have statistically meaningful idiosyncrasies. Testing whether there’s more variation in the results of particular batter-pitcher matchups than would be expected based on chance alone is a tricky sort of question to study (and, by the way, a great opportunity for a submission from an enterprising reader).

One of the things that’s wonderful about baseball is its symmetry. There may be some irregularities in batter-pitcher matchups that are worth exploring, or even exploiting. But for the most part, the matchups play out pretty much like you’d expect them to: good players succeed more often than not, and bad players don’t.