It didn't take too many ringing line drives careening around a 80% empty Rogers Centre before I dove headlong into something interesting: win probability added. WPA is a wonderful thing, one I wish would gain more traction outside the Nerd Herd. Unfortunately, if ever a stat was in need of a rebrand, it's WPA. Click here if you'd like a full definition of Win Probability Added. Old school types should eat it up for its ability to elevate the clutch while de-emphasizing the usual counting stat nonsense.
Again, a rebrand is in order. Drop the tweedy "probability" from the name and go with something that pops like "grititude" or "total awesomesness." Total Awesomeness Added is a stat I can get behind. Perhaps take it in a more bonered direction; like Overall Engorgement Factor or EDHM (Equivalent Dry Hump Minutes.)
Johnny Mac's big home run the other day got me thinking and looking at individual WPA contributions this year. That context is the main driver behind WPA cuts both ways. It can undermine an otherwise dominant performance (as you'll see below) because the game just isn't close enough. But this is a "team game" after all and winning is everything. So just which hitter made the biggest single contribution to a Jays win thus far in 2009?
|Aaron Hill||0.728||May 2, 5-4 (11) v. Baltimore|
|Adam Lind||0.721||May 5, 10-6 v. Cleveland|
|Still the GBOAT||0.630||June 16, 8-3 (10) v. Philadelphia|
|Aaron Hill||0.611||April 7, 5-4 v. Detroit|
|Lyle Overbay||0.475||April 18, 4-2 (12) v. Oakland|
|Rod Barajas||0.473||August 26 3-2 v. Tampa Bay|
|Aaron Hill||0.413||August 1, 6-5 v. Oakland|
|Aaron Hill(!)||0.396||May 27, 10-12 v. Baltimore|
|Travis Snider||0.390||April 13, 8-6 v. Minnesota|
Thoughts or concerns? Lots of extra innings as the added leverage really jacks up the win probability. Also, Aaron Hill is apparently vastly underpaid. 5 of the offense's top 10 games this season, including one loss? Shocking. I am shocked. Barajas accomplished his 0.473 of a win in just one at bat, a pinch hit, game tying affair with the Rays. Remember when Travis Snider hit those two home runs that one game? Yeah, that was awesome. How about the pitchers?
|Roy Halladay||0.458||May 22, 0-1 v. Atlanta|
|Ricky Romero||0.443||April 19, 1-0 v. Oakland|
|Roy Halladay||0.430||July 19, 3-1 v. Boston|
|Ricky Romero||0.423||July 1, 5-0 v. Tampa Bay|
|Brett Cecil||0.421||July 21, 1-2 v. Cleveland|
|Brett Cecil||0.414||May 10, 5-0 v. Oakland|
|Roy Halladay||0.398||July 24, 2-4 (10) v. Tampa Bay|
|Brett Cecil||0.385||July 10, 2-0 v. Baltimore|
|Brian Tallet||0.385||August 8, 3-2 (10) v. Baltimore|
Holy shit, Brett Cecil! Interesting to see Cecil & Romero up here so many times. I guess pitching yourself into and straight out of jams will keep your games tight and your WPA high. 3 losses among the top 10 games, which really says a lot about the amount of context given to WPA. Pitch to the score my ass. Brian Tallet pitching 5 innings of scoreless relief is enough to let the big man amble into the top 10.
Do you notice anything when comparing the contributions of pitchers relative to hitters? The hitting performances carry much, much more weight than solid starts. Again, context is king. Creating a positive event weights that much heavily than preempting negative events before they occur. In other words, it takes a lot to build a large WPA pitching while for hitters it comes in bunches. Again, extra innings helps a lot. Consider Brandon League's three innings of ol'time killin in extra innings against the Yanks ranks ahead of Roy Halladay's one hit masterpiece (which I gave a pitch f/xing earlier this week). Two early runs and the Doc's complete domination worked against him, as it relates to WPA.
Check the list as I compiled it here. I basically took the best performances from each guy this year to a limit. I've also included something else interesting: how hugely a bad pitching performance impacts a win. Seems obvious but I was still surprised. Sadly, Brandon League has 2 of the biggest negative WPA outings of the year. A severe bed-shitting by Tallet in which he repeatedly gave away leads the offense handed him is the worst of the lot, totaling -0.681 WPA.
Thanks to Fangraphs for the WPA info and Baseball Reference for the game logs.