After looking at 90+ Game Scores the other day, I thought I'd look at the other end of the spectrum. I began by looking at the anti-Nolan Ryan's, the guys who threw the most games that scored ten or less. Kenny Rogers is the leader with eleven. Jamie Navarro, one of the few players I truly loathed, was second with 10. Navarro is the career leader in games that scored zero or negative. He was somehow allowed to hurl three of those.
As I was looking through these awful, awful games, I saw that every now and then a pitcher would actually win one of these. The one that stood out to me the most was a game in 1928 between the Reds and the Braves.
The Reds Pete Donohue was allowed to throw batting practice to the Braves even after the game started. He faced 33 batters, walking none of them, striking out none of them. He did hit a batter, the only one who didn't get wood on the ball. Fourteen hits - six singles, four doubles, a triple and three home runs. Eleven Braves that faced Donohue crossed home plate. Donohue's Game Score was 1.
Donohue got the win.
Despite Donohue's efforts, his Reds teammates kept putting up runs of their own. The Reds scored in each of the first six innings and when Donohue was finally yanked in the seventh inning, he left with a 15-11 lead. Donohue's 11 runs allowed ties him for the most surrendered in a game by a winning starting pitcher.
But if that wasn't bad enough, Carl Mays was brought in to relieve Donohue. He got out of the inning, finished the game while surrendering another home run (Les Bell's third of the game) and ended up with a save. The Reds kept adding on runs so the final score was 20-12. I'm trying to determine if the eight run margin is the largest for a game in which a save was "earned".
Yes, it's only one game and an unusual one at that, but I think it exemplifies pretty well why the win is not an adequate measure of pitcher performance and why a save isn't an adequate measure of, well, much of anything.