Look Past the Hype To Find NL Cy Young
This article is from the archive of The New York Sun before the launch of its new website in 2022. The Sun has neither altered nor updated such articles but will seek to correct any errors, mis-categorizations or other problems introduced during transfer.
Most of the column inches around the country concerning “postseason hardware” have been devoted to hashing out the American League MVP debate, but there’s also an engaging one over in the senior circuit: the Cy Young race.
The candidates most often bandied about are Houston’s Roger Clemens and his 1.87 ERA, St. Louis’s Chris Carpenter and his 21 wins, and Florida’s Dontrelle Willis and his league-leading 22 wins.
If the recent statistical revolution has taught us anything about evaluating pitchers, though, it’s that ERA is a flawed indicator of effectiveness, and wins and losses are almost completely useless. Instead, let’s view the NL Cy Young contenders through the prism of the more evocative statistics. Here are the top five NL pitchers this season ranked according to Value Over Replacement Pitcher (VORP), a Baseball Prospectus metric that measures, in runs, what a pitcher provides over a readily available replacement hurler (e.g., the waiver claim, the minor league veteran, the back-of-the-roster mop-up man). Also included are innings (IP), park-adjusted runs-per-game (PK-RA), strikeout-to-walk ratio (K/BB), and home runs per nine innings (HR/9).
You may not hearing the name “Andy Pettitte” in these debates, but according to the meaningful numbers, he’s manifestly a better candidate than the two favorites, Carpenter and Willis.
As you can see, VORP likes Clemens as the best pitcher in the NL, and it’s not a very close call. A thorough analysis of these statistics should give us our NL Cy Young winner.
* Carpenter bests Clemens by more than 30 innings this season, but Clemens’s PK-RA is better than Carpenter’s by almost a full run. The issue is how much value is tied up in those additional innings of Carpenter’s. One way to determine this is to ask what kind of numbers would Clemens have to put up over those 30.1 innings to equal Carpenter’s inferior runs-per-game mark. The answer is that Clemens would have to give up 30 runs – almost a run per inning – over those final 30.1 frames to increase his PK-RA of 2.17 to Carpenter’s mark of 3.08. In other words, those 30.1 extra innings of Carpenter’s had very little value insofar as they relate to Clemens.
* What will no doubt hurt Clemens in the balloting is his modest win total of 13 (Johan Santana, with 16 wins, is similarly handicapped in the AL race, despite being the best pitcher in the circuit). After all, no starting pitcher has ever won the Cy Young in a non-strike-shortened season with fewer than 16 wins. But Clemens’s deficiency is a result of paltry run support; this season, the Astros gave their ace only 3.58 runs of support per start, a figure that ranks 88th among the 93 qualifying starting pitchers in all of baseball this season.
* Quality of opposing hitter must also be considered. You don’t find a great deal of variance among starting pitchers, but nonetheless, here’s how they rank (with average opposing batter’s on-base percentage plus slugging percentage – in parentheses): 1.Carpenter (735); 2.Pettitte (732); 3.Martinez (731); 4.Willis (726); 5.Clemens (724).This is certainly a point for Carpenter and against Clemens, but again, the differences aren’t enough to nullify Clemens’s other advantages.
* What kind of defenses did these pitchers play in front of? According to defensive efficiency ratings, which measures how often a defense converts a ball in play into an out, the Astros rank first in the NL, the Cardinals rank third, the Mets rank fourth, and the Marlins 14th. Consider this to be a point in defense of Willis, who toiled in front of the third worst defense in the NL.
* Among all qualifiers, Willis and Clemens rank first and second in home run rate this season. However, once environment is taken into account, Clemens’s figure becomes far more impressive. Consider that Dolphins Stadium, Willis’s home digs, depressed homer rates by 31.0% this season and by an average of 16.0% from 2002-04. In contrast, Minute Maid Park, where Clemens pitches his home games, increased homer rates by 11.2% this season and by 3.0% from 2002-04.
* Speaking of park tendencies, no Cy Young contender had a rougher go of it than Pettitte. While Minute Maid is a good stadium for all hitters, it’s a dream environment for right-handers: Over the last three seasons, Minute Maid has inflated right-handed home run rates by 16.0% and right-handed batting averages by 4.0% (by contrast, over that same span, it’s decreased lefthanded home run rates by 16.0%). In Pettitte’s case, he faced right-handed batters 78.3% of the time this season, and worked 53.4% of his innings at home. In other words, Pettitte pitched in what was easily the most hostile environment of any NL Cy Young contender. And he thrived in that environment.
* Another point for Pettitte: In six starts against the Phillies, Mets, and Marlins – Houston’s primary wild-card competitors – he logged a spectacular runs-per-game mark of 1.54.
All factors considered, Pettitte should be the NL Cy Young winner. Clemens holds an 8.1-run edge in VORP, but Pettitte has faced tougher opposing hitters, and Minute Maid is a much, much tougher stadium on left-handed starters. Clemens bests Pettitte in PK-RA, but that statistic isn’t adjusted to reflect the differences in how right-handed and left-handed batters fare in a given park. As previously explained, such adjustments would squarely benefit Pettitte. Pettitte, meanwhile, comfortably tops Clemens in innings and K/BB ratio. It’s a close call, but the other former Yankee has been the NL’s best pitcher this season.
Mr. Perry is a writer for Baseball Prospectus. For more stateof-the-art commentary visit www.baseballprospectus.com.