Pitching by the Numbers: No-win situation

If not for bad luck, Jeff Samardzija has had no luck at all in '14. (USAT)
If not for bad luck, Jeff Samardzija has had no luck at all in '14. (USAT)

Wins are fantasy baseball's most frustrating statistic.
“Quality starts” is no solution. But finding some other benchmark isn't much better and seems downright arbitrary.
Wins defenders argue that the answer is to pick pitchers on good teams and, if you don't, that's your fault. It's a clever rationalization, but the correlation between the quality of the team and the degree to which a pitcher gets screwed out of wins is weak. Also weak is the degree to which we can reasonably assess which teams are going to be good in March. Plus, when your ERA is 4.00 or less, you should win the majority of games given the league average in runs scored.
Here are the pitchers with ERAs under 4.00 with at least 10 no-decisions plus losses.
Thanks to the Baseball-Reference's indispensable play index for the stats (stats through Monday).

Hernandez, Samardzija and Kuroda have better ERAs when they lose than when they win. Kuroda (4.71 ERA in six wins) is actually WAY better. That perfectly encapsulates the argument against wins as a category to measure individual performance in our fake game. And are you really going to argue that fantasy baseball isn’t a proxy for projecting individual performance?

Now let’s look at Samardzija having 17 non-wins at that level of performance — out of just 20 total starts. That’s mind boggling. If you’re going to say that “Samardzija owners should have known better backing a Cub,” fine. But note that his teammate and trade-mate to the A’s, Jason Hammel was 8-5 with the same team, pitching slightly worse.

And maybe that’s not unfair. Maybe it’s just too bad, i.e., part of the randomness that makes our game fun. But we’re already trying to conquer randomness by picking individual performers, I believe.

The other pitchers above aren’t necessarily horribly victimized but they are pitching well enough to get wins in more than half their starts, as opposed to zeros. At least most of us don’t play in crazy leagues that count winning percentage as a category, which is far, far worse than just counting wins (at least the losses in standard scoring are zeros).

The flip side of this is the pitchers who pitch the worst in wins. Interestingly, there are only five players with at last seven wins who have ERAs over 3.00 in these wins: Marco Estrada (seven wins, 4.67 ERA in those wins), Justin Verlander (nine, 3.86), Wei-Yin Chen (10, 3.12), Gerrit Cole (seven, 3.07), Jarred Cosart (9, 3.00).

Just for fun, best ERAs with at least seven wins: Clayton Kershaw (10, 0.67), Adam Wainwright (12, 0.78), Kyle Gibson (eight, 0.83), Tim Hudson (8, 0.91), Tyson Ross (8, 1.28), Johnny Cueto (10, 1.29), Edinson Volquez (8, 1.31). Volquez is obviously a shocker. In losses, he’s 8.29. Yikes.

So it’s fair to say that wins are generally earned. And you can make a case that losses are earned, too. But the exceptions are frustrating and often game changing. We can do better. I just have no idea how. Feel free to submit your suggestions in the comments below.

What to Read Next