In short, WPA gives the pitchers points for a good play and takes away points for bad plays. The amount of points depends on how much of an impact that play had on the Rays winning the game (for an intro to WPA, please check yesterday’s post).
Here are the pitchers so far this season…
James Shields has been the Rays most important pitcher, by far. Of course, that is no surprise.
An interesting thing to look at here is the differences between relievers and starters. We always work under the assumption that starting pitchers are more important than relievers because they work far more innings.
However, the outs later in the game can be far more important than the outs early in the game. And if a relief pitcher is given regular work in high-leverage situations, they can have just as much of an impact as some starters.
For example, Kyle Farnsworth has been the Rays second most important pitcher so far this season. One thing that helps Farnsworth is Joe Maddon’s trust. Unlike Rafael Soriano, Farnsworth has come into several games this season in the 8th inning with runners on base. Those outs tend to be far more valuable than the leadoff hitters in the 9th inning with a 2- or 3-run lead.
Of course, if the reliever sucks, those negative points can add up. And we see that with Andy Sonnanstine (although some of that was as a starter), JP Howell, Adam Russell, Jake McGee, and Cesar Ramos, the Rays five pitchers that have hurt the team the most.
We can also scale these values to get a better sense of who is making the most of their opportunities. Here is the WPA per 9 innings…
And now we can really see how important those late inning outs can be, with the good (Farnsworth) and the bad (McGee).
And we also see which relievers should be getting the ball in critical situations (Farnsworth, Juan Cruz) and who shouldn’t (Howell, Ramos, Russell). Then again, most of you guys probably already knew that. But at least we have numbers now to prove it.