Tony Sipp is one lucky bastard. Monday night at baseball’s finest ravine, Sipp, a pitcher for the Arizona Diamondbacks, earned himself a Major League win. What did he do? Sipp came into the game with two out in the eighth inning to face Andre Ethier of the Dodgers. Either popped out weakly to center to end the inning. Arizona went on to score four runs in the ninth against LA closer Brandon “minor” League. Sipp, for his troubles in getting all of one out, was awarded the win.
Don’t tell Mike Francesa, but pitcher wins are stupid. It’s not just that wins are sometimes unfair, they are often misleading. Just look at the example above. The only two people on earth that believe Tony Sipp deserves an honest win for retiring one batter are Tony Sipp’s mother and Tony Sipp’s agent.
What separates the pitching win from the game-winning RBI, another pointless number that became a stat in 1980 and was subsequently revoked when people saw how arbitrary and dumb it was? Nothing much. Here’s what you need to do to get a game-winning RBI: bat in the first run of the game that gives your team the lead. This could be a home run in the first inning of a 9-6 game or a scoring fly ball in the ninth to break a 2-2 tie. It’s all the same. You see how meaningless this information is.
Now consider the pitching win. If you’re the starter, you must complete five innings and you must either finish the game or leave with a lead that is never relinquished. If you’re a reliever, you need only be the pitcher on record when your team takes the lead for good. Far too much of this is arbitrary. For instance, what if your team doesn’t score?
Ask Felix Hernandez of the Mariners about this. In the 2010 season when he won the Cy Young Award, Hernandez averaged 3.07 runs of support per nine innings. In his 34 starts, Hernandez allowed the opposition to score four runs or more just six times.* That means he was good enough to win the game — the average offense scores a bit more than four runs per game — 28 times. He finished the season with 13 wins. Five times he pitched at least seven innings with one or no runs allowed and walked away with nothing.
And what about those bullpen “wins” where the pitcher really didn’t do much except come in at an opportune time. You might be thinking that’s not really much of a big deal, but in reality, distorted wins like this occur in about 27% of games. The below table shows all games for the past five seasons along with those in which the winning pitcher recorded six outs or fewer:
| G | BSW | % | |
|---|---|---|---|
| 2012 | 2,430 | 642 | 26.4% |
| 2011 | 2,429 | 664 | 27.3% |
| 2010 | 2,430 | 636 | 26.2% |
| 2009 | 2,430 | 656 | 27.0% |
| 2008 | 2,428 | 682 | 28.1% |
| Total | 12,147 | 3,280 | 27.0% |
These are indicated as BSW or BS wins, which is what they are. So far in 2013, there have been 274 such games out of a possible 945, or 29%.
Worse, so far this season there have been 53 “wins” awarded to pitchers who recorded just one out. This is the pinnacle of absurdity.
The great thing about baseball, as opposed to other team sports, is the ability to track everything as a series of discrete events. Mark Reynolds struck out 223 times. Derek Jeter reached base 37.9% of the time leading off an inning. The Red Sox have seen 10,920 pitches, the most of any team. Yet fans and analysts seem to accept the pitching win alongside these legitimate statistics despite the fact that much of what goes into the win itself is completely out of the pitcher’s control.
That people in the media, and particularly radio talk show hosts born before 1970, continue to god-up wins is doing a disservice to players and fans. Broadcasters chatter all the time about will there be another 300-game winner? Sportscenter anchors crow about the first guy in the league to get to 15 wins. And talking heads of an earlier generation protested when Zack Greinke and Felix Hernandez won Cy Young awards despite having “only” 16 and 13 wins. This is a bunch of crap.
Lots of “wins” doesn’t mean a pitcher is good. Look at 2012 Phil Hughes. He “won” 16 games. Pretty much any idiot could have done that when the Yankees scored seven runs or more in 10 of his starts. (His record: 7 wins, 1 loss, 2 no-decisions.) They scored five or more 18 times behind Hughes in 32 starts. But old people like to tell you he was a 16-game winner.
The best solution to this is to eliminate the win altogether. The traditional earned-run average has flaws for sure, but it at least tells you to some degree who is good and who is not. Couple that with some meaningful information, such as how many times was a starter gave his team a chance to win. One way to do this currently is with quality starts, defined as six innings or more while allowing three earned runs or fewer. That’s a 4.50 ERA. Given the average team scores that many runs per game, anything better from the pitcher means his team is more likely than not to win. It’s not perfect, but it’s good enough. So instead of displaying arbitrary wins and losses, you should see James Shields, 2.79 ERA, 12/14 QS. (Shields is tied with Hisashi Iwakuma of Seattle for the best quality start ratio in the AL, 86%, while he has but a 2-6 record. And you see why wins are dumb.)
At least a generation of people is going to have to die off before this would ever happen. In the meantime, why not at least eliminate the BS win. Make wins the exclusive provenance of the starting pitcher. So what if a game ends without a “winning” pitcher. Is that worse than awarding a win to someone who had minimal impact, something that now happens more than a quarter of the time? Or, to give deference to long relievers, allow awarding of wins only to those who pitch a majority of the innings, so if the starter comes out in the third and another guy finishes the game, he can be the winner. Fine.
*If you’re scoring at home, he allowed four or more earned runs only four times. In an August game at Cleveland, the Indians scored six unearned runs against Hernandez, sending him to his 10th loss of the season. In a July start in Chicago, only two of the four runs he yielded the White Sox were earned. That would make his 2010 Quality Start Factor 30/34 (88.2%).








