The perfect use of WPA
ByIn baseball’s endless search for new ways to measure the game, we’ve recently seen the development of WPA. Using historical data, it tracks the odds a team has of winning a game at any given time, based on run differential, outs, and base situation. It has its flaws, but it does serve one purpose: it captures the essence of a game. Tom Tango examines this using a painful example, Ichiro’s walk-off. With two outs in the bottom of the ninth, WPA gave the Yanks an over-95-percent chance to win the game. Mike Sweeny’s double reduced that to 85.8 percent, and Ichiro made up all the difference. Tango also notes the general fan reaction, which makes sense after such an improbably turn around.
This feeds in nicely with the discussion in the Dominance Factors thread. The purpose of WPA is not to evaluate a player’s performance. It is to show the ebbs and flows of a game. Used for this purpose, it’s a wonderful stat. Used for others, it can mislead.
By request, the graph from that night:
From FanGraphs.
22 Comments»
RAB Info
The Founders
Mike Axisa, Ben Kabak, Joe Pawlikowski
Editor-In-Chief
Mike Axisa Archives | Email
Managing Editors
Ben Kabak Archives | E-mail
Joe Pawlikowski Archives | E-mail
Contributing Editors
Larry Koestler (Archives)
Moshe Mandel (Archives)
Stephen Rhoads (Archives)
Weekend Writers
Hannah Ehrlich (Archives)
Matthew Warden (Archives)
Special Contributor
J.R. O'Grady (Archives)
Subscribe to RAB by RSS and Twitter
Follow @RiverAveBlues Search RAB
Spring Training Countdown
Just 5 days 'til pitchers and catchers report.Banner Design By
Do me a favor if you don’t mind, Joe Pow:
Embed the image of the WPA graph from that Mariano blown save to illustrate why I singled that game out in my earlier “Had we not lost that game, we’d be 13-6 instead of 12-7 and up 6 games on Boston and most of this caterwauling would be moot”.
A picture tells a thousand words.
Thanks bro.
That’s a game we’re judging our September run off of. That line is almost vertical. Against the greatest closer in the history of baseball.
Yeesh.
I blame Mariano River and Ichiro for the unnecessarily panicking masses of Yankee fans. If we reversed the WPA graph from Friday, it would provide a nice cliff off of which everyone could jump.
Heh.
http://www.fangraphs.com/tgrap.....ners_0.png (embed pls? kthxbai)
+42
i would never judge a player’s ability based off something like this, but i think it’d be interesting if the changes in win-probability were calculated after every at-bat, and then compiled for each player over the course of the year, as a way of showing which players had the greatest all-around impact on their teams, and which players were “lucky,” and stuff like that.
And if there is already a stat like this, i’m sorry. I’m a newb. But i’d like to say that i did think of it independently, so shut up
WPA works exactly that way.
Mariano’s career WPA is +47.92. By comparison, in about 200 MORE innings, Lee Smith’s WPA is +23.97, so, um, yeah, Mariano’s really good.
See also: http://en.wikipedia.org/wiki/Win_Probability_Added
I was going to point out that they played in different eras, but I remembered that I don’t have as much of a grasp on eras as a certain other person, so now I’m not sure.
I think I speak for everybody when I say holy fucking shit.
their stats were wrong–i saw it with my own eyes.
ONE IN A 130,000,000!!!
Talk about lame stat
I’ll start off by giving you the benefit of the doubt-why?
Yeah, it’s so LAME to know how much a given AB added to or reduced a team’s chances of winning.
RRD and whozat:
This is what I was talking about in the previous thread. Why even acknowledge this comment?
I wanted to give this guy the benefit of the doubt to see if he had reasoning to back up this claim.
Unlikely, but I’m feeling generous.
If he has something to back it up, he’ll bring it on his own accord. If not, well, my suggestion is to leave it alone.
All right.
Well, that was productive.
Of all the games to pick to illustrate the awesomeness of WPA, he picks this one.
::writhes on the ground in agony::
Unfortunately, the most painful game is also the most illustrative.
I was at the Orioles/Red Sox game this week, and it was 11-3 Sox going into the bottom of the ninth. We were thinking about leaving, so I fangraphed the WPA at they gave me basically 100% chance of the O’s losing. Oriole magic must not be factored into WPA…weird. So we trusted the stats and we just out of Camden Yards, when we watched the back to back HR’s on the TV, and then a walk. Needless to say, we threw WPA out the window in search of historic comebacks. 5 mins, and a game ending DP later, we left with the crowd and traffic jam.