Happy Holidays from Baseball Prospectus! We'll have new content on December 26th.
October 27, 2005
One of the few differences--other than wins and losses--between the Astros and the White Sox in the World Series is that the Astros pitching talent tends to be concentrated in their best pitchers. There's been no shortage of coverage of the performance of Roger Clemens, Andy Pettitte, and Roy Oswalt, but additionally the Astros' bullpen leans heavily on Brad Lidge, Chad Qualls, and Dan Wheeler before dropping off steeply into the depths. By contrast, the White Sox show little difference between their top four starters, and the drop to Brandon McCarthy or Orlando Hernandez is slight. While the raw difference between Cliff Politte and Dustin Hermanson or Neal Cotts is large, that's mostly because of Politte's ridiculous season. As displayed Tuesday night, the White Sox's pen can continue to roll out quality relievers long after Houston is forced to call in the likes of Ezequiel Astacio.
Normally, the postseason and World Series format would appear to tip the scales in favor of the Astros, but when Clemens went down on Saturday and the extra innings piled on Tuesday night, the scales favored the White Sox. Over the course of a season, we tend to assume that those kinds of differences even out, at least from a batter's perspective. White Sox hitters would see their fair share of pitchers the likes of Clemens, Pettitte, and Oswalt, but they'd also get a few at bats against the Astacios of the world to even things out. That's not entirely true--which is why we provide the Batter's Quality of Pitchers Faced report--but the range in difference between the best and worst array of pitchers seen by qualified batters is very small. In 2005, as measured by OPS, the Rockies' Luis Gonzalez faced the weakest pitching in the majors (.261/.341/.419, due in no small part to his home park) while the Royals' Mike Sweeney (.249/.318/.392) was pitted against the stingiest. While the 50 points of OPS difference is significant, that's the difference between the two extremes. Between most batters, the difference is very small.
The overall difference may be slight, but that doesn't prevent some batters from developing reputations for feasting on bad pitchers while struggling terribly against the games' better hurlers. These players, it is said, aren't as good as their overall performance suggests because they can be shut down in key game situations by top quality pitchers. Of course, most batters are going to do worse against better pitchers, but some are rumored to do worse than we would expect. This argument is similar to the one Alex Rodriguez apparently wears around his neck; while Rodriguez supposedly pads his stats in garbage time--games that have long since been decided--other batters are rumored to only perform well against poor pitchers. In either case, the batter is overvalued by their stats because they don't have the same effect on the outcome of the game as other players with similar stats.
In particular, Adam Dunn seems to have this reputation. To find out if this is true in Dunn's particular case, let's take a look at his numbers this year broken down by the quality of the opposing pitcher. Among pitchers with at least 30 IP, the average ERA was 4.22 and the first and third quartiles were at 3.32 and 5.03, respectively. Breaking down Dunn's season into plate appearances against pitchers outside of those quartiles, we should expect about 25% to come from below 3.32 and another 25% to come from above 5.03. In reality, it's a little bit less than that, probably because most of the outliers are relievers and pitchers who don't pitch as many innings. As such, let's set the lower bound at 3.40 and the upper at 4.90, giving us slightly larger samples of good and bad pitchers.
The other important consideration to make before looking at the numbers is the handedness of the opposing pitcher. It may be the base that Dunn is facing a disproportionate number of quality southpaws since teams are trying to target his weaknesses against them. If that is the case, an inability to hit quality pitching may simply be the result of a poor platoon split. As such, let's look at Dunn's numbers against six groups: three strata of pitchers (Excellent, Average, and Bad) broken down by pitcher handedness:
BATTER YEAR THROWS QUALITY PA AVG OBP SLG Adam Dunn 2005 L Excellent 56 .063 .196 .083 Adam Dunn 2005 L Average 107 .204 .308 .699 Adam Dunn 2005 L Bad 67 .278 .403 .630 Adam Dunn 2005 R Excellent 144 .263 .396 .559 Adam Dunn 2005 R Average 235 .303 .438 .766 Adam Dunn 2005 R Bad 114 .273 .439 .545
First, check out those numbers against Excellent LHPs: .063/.196/.083. Yikes. Sure, it's only 56 PA, but that's one ugly line. What's more, against top quality righties, Dunn managed a .263/.396/.559 line; that's not the line from a batter feasting on the low-hanging fruit. His .955 OPS against that group isn't far off the .984 he managed against the worst pitchers in the league.
Expanding this idea to the entire league is a little tricky. It's easy to look at Dunn's line against quality LHPs and see that that's less than we would expect him to hit. But what would we expect him to hit against that group? Overall, lefties with an ERA under 3.40 allowed left-handed batters to hit .227/.290/.358 compared to .254/.320/.415 for all LHPs, a net of -.027/-.030/-.057. (As an interesting side not, pitchers in the Excellent group show smaller platoon splits than the other two groups.) Meanwhile, Dunn hit .190/.309/.528 against lefties overall, so if we adjust those numbers for the elite group of LHPs, Dunn "should" hit .163/.279/.471 against that elite group of southpaws. Thus, in this case, Dunn underperformed against that elite group by -.100/-.083/-.388.
In order to determine if Dunn is the kind of batter who underperforms against good pitching and overperforms against the game's have-nots, we can compare the line we found above to the same line for Dunn against terrible southpaws. As expected, Dunn performs much better against bad pitchers--to the tune of .165/.153/.422--than against good ones even when we account for the quality of the pitchers involved. However, in the vast majority of his PAs, Dunn faced RHPs and in that case he actually performed much better against the games best pitchers when accounting for expected performance. Dunn's line of -.069/-.039/-.195 indicates he significantly underperformed against the worst pitchers in the league compared to his performance against good ones. Thus, Dunn's reputation for feasting on bad pitching is undeserved, at least in 2005.
How does the aforementioned Rodriguez do in this? The following is his performance against bad pitchers compared to his performance against good pitchers, adjusted for the quality of the pitcher. Negative numbers indicate that he did worse against bad pitching than we would expect; positive that he instead took advantage of bad pitching:
BATTER THROWS AVG OBP SLG Alex Rodriguez L -.097 -.179 -.139 Alex Rodriguez R -.092 -.081 -.188
Interestingly, Rodriguez performs significantly better against the best the league has to offer than the lesser pitchers of the league.
Of the newly crowned World Champion Chicago White Sox, only Paul Konerko and Joe Crede show much of a tendency to perform better than expected against the cream of the crop. The rest of the squad, to varying degrees, had seasons in which they took a greater advantage of the league's worst pitchers. This isn't to take anything away from their incredible run through the playoffs or their championship, but rather to point out just how improbable their victory over the league's best was.