Winning and Losingby Aaron Gleeman
September 05, 2005
Gloria Clemente: You see Billy, I've got a different set of rules. Do you want to hear them? Here they go. Sometimes when you win, you really lose. Sometimes when you lose, you really win. Sometimes whether you win or lose, you really tie. And sometimes when you tie, you really win or lose. Winning and losing is all one big organic globule from which one extracts what one needs.
Billy Hoyle: I hate it when you talk like that.
I wrote last week about the lack of support the Twins' horrible offense has provided Johan Santana this season, which resulted in a lot of you e-mailing me to say that Santana's poor run support is nothing compared to the lack of runs both Roger Clemens and Kevin Millwood have had to work with this year. Coincidentally, Santana and Millwood matched up against each other at the Metrodome Saturday night, and the result was your basic unstoppable force/immovable object situation. In other words, neither pitcher got a win.
Santana was brilliant, tossing eight innings of one-run ball while striking out 10 Cleveland hitters and allowing just two hits and zero walks. The Twins were shut out for five innings before finally scoring two runs in the bottom of the sixth, and then closer Joe Nathan blew a 2-1 ninth-inning lead to stick Santana with his ninth no-decision of the year. Meanwhile, Millwood was also very good, allowing just two runs in seven innings of work. The Indians managed just one run while he was in the game, however, and Millwood was taken off the hook for his 12th loss of the season when Nathan blew the save.
The Twins ended up scratching out an unearned run off Bobby Howry in the bottom of the ninth inning to pick up the victory and give Nathan his seventh win of the season. So in a game which featured two starting pitchers tossing eight and seven innings while allowing one and two runs, respectively, the guy who "won" the game had a 9.00 ERA in the one inning he pitched. Anyone with half a brain will agree on that making very little sense, yet it's the same sort of "logic" that is used by fans and media alike every year.
There has been a lot of progress in the way people watch, discuss, and analyze baseball over the last decade. Fewer and fewer people think the "best hitter" is the guy with the highest batting average, on-base percentages are gaining ground on RBIs when it comes to stats spouted in MVP discussions, and even the mainstream media is beginning to see the value in things like OPS. However, the one area where very little progress has been made is in how we view the relationship between pitching performances and wins.
Since the beginning of August, Santana has tossed 53.1 innings with a 1.35 ERA over seven starts, and has just three wins to show for it. How? Well, the Twins scored him a pathetic 2.3 runs per game and managed just one measly run in four of the seven games. Given even four runs per game to work with, Santana would have gone 7-0 over that span and 10-0 since the All-Star break, and he'd have 17 wins on the season. Instead, because he pitches for a team with an awful offense, he's stuck on 13 wins and his chances of winning a second straight Cy Young award have all but vanished along with the Twins' playoff hopes.
While most people would agree that Nathan "winning" Saturday's Twins-Indians game is silly, a lot fewer people would recognize that the same flaws exist in Santana's overall record this season. With 197.2 innings pitched and a 3.07 ERA in 28 starts, along with a 205-to-36 strikeout-to-walk ratio and .212 opponent's batting average, you wouldn't think that the quality of Minnesota's offense should have anything to do with the quality of Santana's season. But when it comes time to cast a vote for who the "best" pitcher in the AL is, you can be sure that his mediocre win total will come into play.
And as the many e-mailers pointed out to me last week, even Santana doesn't have a whole lot to complain about when compared to Millwood. Despite ranking fourth in the league with a 3.15 ERA in 25 starts, Millwood has just seven wins all year. Far too often this season he has pitched very well—like he did against the Twins Saturday—and come away with nothing because the Indians failed to score more than one or two runs.
While Minnesota's offense has been bad regardless of what pitcher is on the mound, the thing that makes Millwood's lack of run support unique is that the Indians actually have a pretty good offense. Or at least they do when he's not pitching. In Millwood's 25 starts, the Indians have averaged 3.2 runs per game (which would rank dead last in the AL, 24% behind the lowly Royals). When Millwood hasn't started, the Indians have averaged 5.2 runs per game. That's how Millwood is 7-11 with a 3.15 ERA while Cliff Lee is 14-4 with a 3.90 ERA and C.C. Sabathia is 13-9 with 4.54 ERA for the same team.
In addition to averaging 38% fewer runs when Millwood is on the mound, the Indians have been shut out four times in his 25 starts and have scored three runs or fewer in 17 of the 25 games. And while Cleveland has had plenty of offensive explosions while averaging 4.8 runs per game overall this season, the Indians have scored more than five runs in just four of Millwood's 25 starts. Millwood has been one of the best pitchers in the AL this season, but I'm guessing regardless of how well he pitches down the stretch he won't receive a single vote in the Cy Young balloting.
Clemens' situation is like a cross between Millwood's and Santana's. Like Millwood, he has received remarkably little offensive support from his team. And like Santana, his lack of wins is hiding the fact that he has likely been the best pitcher in the National League so far this season (although Chris Carpenter has amazingly pitched well enough to make it a close call).
Not only does Clemens' 1.57 ERA lead the league by 31% over Carpenter's 2.28, it would be the third-best ERA by a starting pitcher since Bob Gibson's miniscule 1.12 in 1968. Only Dwight Gooden (1.53 ERA in 1985) and Greg Maddux (1.56 ERA in 1994) have posted lower ERAs since then, which is pretty special when you consider Clemens is a) pitching in a much more hitter-friendly era than Gooden, and b) pitching in a far more hitter-friendly ballpark than both Gooden and Maddux. Throw in the fact that he's doing it at the age of 43 and it's incredible.
Yet despite a league-leading ERA that is historically low and 189.1 innings pitched (which ranks seventh in the league), Clemens has a grand total of just 11 wins on the season. If you take wins and losses out of the picture it's clear that what he's doing is awesome, but when you look at his 11-6 record compared to Carpenter at 20-4 or Dontrelle Willis at 19-8, it suddenly doesn't look all that spectacular. The problem, just like with Nathan, Santana, and Millwood, is that much of Clemens' low win total is not his fault or even under his control.
Clemens has started 28 games this season and the Astros have failed to score a single run in eight of them. That means, regardless of how well Clemens has pitched, he had zero chance to win a game in 29% of his starts. There's nothing a pitcher can do when his offense literally gives him zero support, although with a 1.57 ERA Clemens is certainly trying to test the limits of that theory. Counting the eight shutouts, Houston's lineup has scored three runs or fewer with Clemens on the mound 18 times, or in 64% of his starts.
Their overall average of 3.2 runs per game in Clemens' 28 starts is bad enough (Carpenter has received an average of 5.3 runs per game from the Cardinals' lineup), but even that number is inflated a bit by a 14-run outburst the Astros had against the Nationals back on July 22. Those 14 runs helped Clemens, of course, but since he tossed six shutout innings it was probably overkill. Take that one offensive explosion out and Houston has provided Clemens with an average of 2.9 runs per game in his other 27 starts this season.
Clemens has essentially had zero margin for error this year. In fact, he has allowed zero, one or two earned runs in 10 of his 11 wins this season. But even when he's been nearly flawless, it hasn't been anywhere close to a guaranteed victory. In the 17 starts he has made without picking up a win, Clemens is 0-6 with a 1.78 ERA in 115.1 innings of work. Wins and losses are hugely important, and no one will ever claim otherwise. However, to pin them on an individual pitcher is deeply flawed for the reasons discussed above and many more.
Why depend so much on and place so much value in a system where guys like Clemens, Santana, and Millwood can have outstanding outings turned into "losses" and "no-decisions" because the hitters on their team struggle to score runs and relievers like Nathan can record "wins" in appearances where they don't even pitch effectively for one inning? Much like buying Jessica Simpson albums and going to Jennifer Lopez movies (or buying Jennifer Lopez albums and going Jessica Simpson movies, I suppose), it makes absolutely zero sense whatsoever and yet we continue to do it year after year.
Aaron Gleeman is a freelance writer whose work can also be found regularly at AaronGleeman.com, Fox Sports, Rotoworld, and Insider Baseball. He welcomes comments, questions, and suggestions via e-mail.