This pitching stat is meaningless
Last season, Zack Greinke and Tim Lincecum claimed, respectively, the AL and NL Cy Young Awards, and they did so despite combining for just 31 wins.
That voters would reward a pair of pitchers who together tallied as many wins as, say, Denny McLain in 1968, was to some a grave corruption of the way things ought to be. Others, though, saw it for what it is: progress.
Indeed, it’s long past time for fans, analysts, baseball people — anyone having anything to do with the game — to stop paying attention to pitcher wins and losses. Doing so would enhance our understanding of the sport and lead to better decisions by both organizations and those charged with handing out hardware.
The most common and most predictable objection will be something like this: wins are the most important stat, so why shouldn’t they be counted? The simple answer is that wins and losses are team stats, and they’re counted — with utmost importance — in the standings.
The idea that we should assign wins and losses to individual players is an absurdity unique to baseball. This is, of course, a team sport, and given the determinative roles the offense and bullpen play in the matter of which team wins and which team loses, it’s inexplicable that we assign glory or blame to an individual pitcher. Such lazy thinking wouldn’t be tolerated in a third-grade classroom, yet it persists in baseball at even the highest levels.
Why is that? It’s a force of habit and history, to be sure. Still, habits and histories can be overcome, and in the service of doing so, let’s remind ourselves where the rules for doling out wins and losses can go stupidly wrong:
• As we all know, a starting pitcher gets a win for working at least five innings and giving his team a lead that’s never relinquished. Sounds harmless enough, but in practice a pitcher can — by any reasonable standard — stink up the joint and still get a shiny "W" next to his name. On such occasions, it’s of course the offense that contributed the most to the victory and not the starting pitcher whose good fortune it was to be lousy on a day in which his counterpart was just a bit lousier. In fact, you can hemorrhage 12 hits and nine runs in 5.0 innings and still "earn" the victory.
• By the same token, a pitcher can "lose" a game in which he was utterly dominant. The defense can let him down, or the bullpen can fritter away his lead, or he can get no run support. If he’s especially fortunate on such days, he’ll escape with a no-decision despite giving his team ample chance to win. In fact, you can throw a no-hitter and still be slapped with a loss; you can retire the first 36(!) batters you face, give up one unearned run in 12.2 innings and still somehow "lose"; and you can win an ERA title while going 8-16.
• Let us only briefly wander into the thicket of idiocy that are the rules governing reliever wins and losses. Some variation of the following happens all the time: Closer blows save at home, offense pushes a run across in the bottom half, closer gets win. If starting pitcher wins and losses are mostly useless, then reliever wins and losses are actively harmful.
There’s precious little evidence that pitchers, with any consistency, can "pitch to the score" or otherwise do anything beyond keep runners off base and keep runs off the board (and even those elements of the game aren’t fully under their control).
In other words, the idea that wins and losses are an accurate reflection of pitcher performance (or, more inanely, a pitcher’s desire/will to win/moral fiber/innermost fires/testicular fortitude) is based on myth and received wisdom.
Like almost any flawed statistic, win-loss records can be vaguely useful at the margins (i.e., a guy who goes 23-5 or some such in a given season was at least good, but it doesn’t mean he was better than the guy who went 17-10). However, in the main, they’re treated as something more than that.
So, how should we be evaluating pitchers? If you’re a quick-and-dirty traditionalist, innings and ERA will suffice. If you want a higher level of precision, advanced measures like Fielding-Independent Pitching, Wins Above Replacement and Leverage Index will provide a fuller and truer picture of a pitcher’s value.
As for wins and losses, it would be a good and belated thing if we started treating them like, say, the records of quarterbacks and goalies — largely pointless afterthoughts — or, better yet, if we didn’t track them in the first place.
At the very least, though, the rules for designating pitchers of record need to be restructured to reflect — novelty of novelties — pitching performances that are notably good or notably bad. Indeed, it’s time we take the lessons of Greinke and Lincecum and carry them to their logical conclusion: wins and losses are team statistics, not pitcher statistics.