WPA Followup to Wheeler & Balfour

Following up to the main page post on Dan Wheeler, Grant Balfour, J.P. Howell, and leverage, lets take a deeper look at Win Probability Added in each situation. First some facts:

*Due to the presence of leverage in WPA, magnitudes will be considerably smaller on lower leverage situations. For the sake of comparing across pitchers it can still be useful.

*I am omitting average WPA+ because the magnitudes are far more consistent. That makes sense when the best you can do is get people out, but the damage you can do is limitless. The WPA+ info is included when the average WPA was calculated.

*its not an exact science b/c not all high leverage scenarios are alike but over 1.25 seasons it can provide some useful information.

* For each leverage scenario we will look at each pitcher's % of -WPA performances, the average WPA- performance, and the overall average WPA performance.

Without further ado:

The most telling stat here is Dan Wheeler's exceptionally low 8.0% WPA-. Magnitudes are too low to give us much more information.

Here is where the longball starts to haunt you. Despite having the lowest % of WPA- performances, Wheeler's overall WPA is nearly half of Howell's and Balfour's. This is becuase his average WPA- is almost 3 times Howell and Balfour's negative outings.

Now Wheeler has the highest fail %. His Average WPA- is nail in the coffin type stuff. Multi-run homers in high leverage situations are nearly impossible to overcome.

Much smaller sample sizes. Wheels did not let the ball leave the park in very high leveraged situations.

In summary its not that Wheeler fails more often, its that he fails harder. When the home run becomes costlier, walks become probable because he is extra careful. In lower leverage situations where he can afford to give up the occasional solo blast, he is very useful in keeping runners off the basepath due to his high strike % and high flyball %.