More to starting pitching than wins, losses


People love a winner, and there's nothing wrong with that. People also love to assign credit for winning. There's nothing wrong with that, either, so long as we recognize its limitations.

Take a starting pitcher's win-loss record.

It tells us what happened, but not how or why. If we use wins and losses to judge the quality of a pitcher's performance, we risk reaching faulty conclusions.

Felix Hernandez is a recent example fresh in everyone's minds. He turned the American League into his personal playground in 2010, but because he played for a Seattle Mariners team that refused to score runs, he finished with a pedestrian 13-12 record. This led to talk in some circles that Hernandez might not deserve the Cy Young Award.

Beyond Hernandez, others suffered similar fates this past season. And still others succeeded despite themselves.

A while back, I repurposed an old Bill James study to compare the 2003 seasons of Jarrod Washburn and Ramon Ortiz. I've since done the same with the 2008 seasons of Jake Peavy and Tim Redding. In each case, one pitcher was more effective than successful, while the opposite held true for the other.

You can find these pairs of pitchers in any season. There were several in 2010, the most extreme being Atlanta's Tommy Hanson and Milwaukee's Chris Narveson.

Here are their overall numbers as a starter (Narveson began the season in the bullpen, going 1-0 in nine relief appearances):

Did Narveson have the better year? As measured by wins and losses, yes. As measured by run prevention, which typically leads to wins, no, and it's not close.

The obvious first question is what kind of run support each pitcher received. In this case, the answer doesn't help -- the Braves averaged 4.29 runs in Hanson's starts, while the Brewers averaged 4.25 when Narveson toed the slab.

So we dig deeper and discover something curious. When their teams scored three runs or more, Hanson and Narveson were indistinguishable:

Hanson was slightly more effective, Narveson slightly more successful, but those two lines are close. What isn't close is how they fared when their teams scored two runs or fewer:

Even accounting for the fact that Hanson allowed several unearned runs, this is a huge discrepancy. It doesn't explain everything, but does provide insight into why there is a gap between these two pitchers' overall effectiveness (as measured by ERA) and their success (as measured by wins and losses).

There are other angles to examine as well. One is how each pitcher fared in their wins:

And in other starts (losses and no-decisions):

Is it reasonable to expect someone with a 4.19 ERA to sport a record of 0-13? Well, Phil Hughes had the same ERA in 2010 and went 18-8. Anyway, this is a baseball column, not a calculus class so we'll stop here. If you're curious and motivated, here are some follow-up questions about Hanson and Narveson's starts worth considering:

• How well did the opposing starter pitch?

• How effective was the bullpen?

I'm sure you can think of others, but the point is that without proper context, it is impossible to make an accurate judgment about a player.

Wins and losses tell a story about a pitcher, but not the whole story. When we miss that point, we find ourselves dismissing Hernandez as a Cy Young candidate or arguing that Narveson had a better season than Hanson.

When we ignore available evidence, we look like idiots. Nobody wants that.

Geoff Young writes Ducksnorts, a blog about the San Diego Padres. Follow him on Twitter.