Most rants about the silliness of the pitching wins statistic focus on its application for starting pitchers. Rarely does the light shine on its even more arbitrary application when it comes to relievers, but on Thursday night in Baltimore, things got very weird. The result was a legal but almost never used official scoring decision that, given the similar context, hadn't been seen in baseball in several years.
The Yankees led the Orioles 5-2 through seven innings when New York manager Joe Girardi called upon David Robertson to pitch the eighth. Robertson is generally one of the game's top setup men but recently the victim of a bout of shoulder tendonitis, which shut him down for five days. He retired the first two Baltimore hitters he faced, Manny Machado and Chris Davis, but proceeded to yield singles to Adam Jones and Nick Markakis, then served up a game-tying three-run homer to Danny Valencia and a double to J.J. Hardy before striking out Matt Wieters. Fortunately for the Yankees, they scratched out a go-ahead run in the ninth against struggling Orioles closer Jim Johnson, so on came Mariano Rivera, who retired the O's 1-2-3 to preserve the win.
By official scoring custom, Robertson should have been charged with a blown save (just his second of the year) but credited with the win, because he was still the pitcher of record when the winning run scored. In turn, Rivera should have been awarded the save, which would have been his 44th of the year, pushing him past Johnson for the AL lead, and the 652nd of his career, padding his all-time record total. Instead, the official scorer in Baltimore invoked rule 10.17(c), which states the following:
The official scorer shall not credit as the winning pitcher a relief pitcher who is ineffective in a brief appearance, when at least one succeeding relief pitcher pitches effectively in helping his team maintain its lead. In such a case, the official scorer shall credit as the winning pitcher the succeeding relief pitcher who was most effective, in the judgment of the official scorer.
Rule 10.17(c) Comment: The official scorer generally should, but is not required to, consider the appearance of a relief pitcher to be ineffective and brief if such relief pitcher pitches less than one inning and allows two or more earned runs to score (even if such runs are charged to a previous pitcher). Rule 10.17(b) Comment provides guidance on choosing the winning pitcher from among several succeeding relief pitchers.
The scorer judged Robertson to be "ineffective in a brief appearance" and thus took away his win and awarded it to the only succeeding reliever, Rivera. What's strange, though, is the inconsistent application of this rule. A quick look at the Baseball-Reference.com Play Index shows that over the past 20 seasons, a total of 69 relievers were awarded wins despite having the equivalent outing Robertson had on Thursday: one-inning appearances in which they allowed three or more runs, including three pitchers this year. All but 10 of them also included the pitcher being charged with a blown save; if he wasn't it's because he either entered with a lead of more than three runs, entered a tie game or entered before the starter had completed five innings. If we cull the list to exclude those 10 as well as relievers who finished the game (since Robertson did not), 42 pitchers in that set found themselves in similar territory, all but three of whom pitched in games decided by three or fewer runs, i.e., games where a subsequent pitcher who finished up could have been credited with a save under normal circumstances.
In each of those remaining 39 games, in fact, saves were awarded, and 22 times, they were awarded to the pitcher who immediately followed the guy who allowed three or more runs in a full inning of work, the penultimate inning of the game (eighth inning, or ninth in a 10-inning game). In other words, in those analogous situations, it was scored as you'd expect, with the struggling pitcher being awarded the win, and the closer throwing the cleaner inning getting the save. Trevor Hoffman, who held the all-time saves record for a short time before Rivera broke it, was one such beneficiary, on July 18, 2007:
|Heath Bell, H (17)||0.2||0||0||0||0||0|
|Scott Linebrink, BS (5), W (3-2)||1||1||3||3||2||0|
|Trevor Hoffman, S (26)||1||0||0||0||0||1|
For comparison, here's the relevant portion of Thursday night's Yankees box score:
|Adam Warren, H (1)||0.2||1||0||0||1||0|
|Cesar Cabral, H (1)||0.1||0||0||0||0||0|
|David Robertson, BS (2)||1||4||3||3||0||2|
|Mariano Rivera, W (6-2)||1||0||0||0||0||0|
Robertson allowed four baserunners, one more than Linebrink, but half of the pitchers in question allowed four or more baserunners and still got the win. Looking over recent history, it's difficult to reach any other conclusion but that Thursday's official scorer deviated from custom in applying rule 10.17(c). Why he did so is unclear, but the decision sticks out like a sore thumb nonetheless. Via ESPNNewYork.com, which cited research provided by the Elias Bureau, three other times in the past 25 seasons relievers have entered with the chance to get a save but instead were awarded wins. Of those three, the one that parallels Thursday night's situation is a July 21, 2001 game between the Angels and Orioles. That night, Angels closer Troy Percival entered to start the bottom of the ninth with a 5-2 lead and gave up three runs. The Angels scored in the top of the 10th to go ahead 6-5. Shigetoshi Hasegawa replaced Percival to start the bottom of the 10th and pitched a 1-2-3 inning and was awarded the win instead of the save.
As with retroactive decisions to change the official scoring on certain plays regarding hits/errors, earned/unearned runs and RBI/no RBI, the ruling on Rivera is presumably something that could be reviewed by the league office and changed. Of course, none of this matters in the grand scheme beyond the fact that as far as the standings are concerned, the Yankees won and the Orioles lost. Nobody should lose any sleep over whether Rivera is deprived of his save, because in all likelihood, nobody in our lifetime is going to catch up with his total any more than Robertson is going to use his missing win to overtake Cy Young.
This is yet another example of why the win statistic for pitchers is an increasingly outmoded one. It has at least as much to do with circumstances beyond a pitcher's control -- the support he receives from his offense, defense and bullpen -- as it does with his own performance. On Thursday, it was also determined by one person's opinion that was far out of line with contemporary precedent.
All of this goes to show how silly the application of this particular statistic can get. I'm not arguing we should — to use MLB Network host Brian Kenny's term — kill the win, but it seems appropriate to stick it in the junk drawer with the souvenir matchbooks, rusty can-openers, keys for changed locks and random foreign coins. It's not much more useful than that. NOTE: This story has been updated to reflect additional research and information.