Saturday, April 23, 2011
Pimp my wOBAPosted by JT Jordan
And with that, I have fulfilled my pop culture reference quota for the year.
Steve Slowinski of FanGraphs wrote a nice piece on wOBA the other day, comparing it to OPS. In the comments section, reader Andy posted the following question:
Why are strikeouts not included in wOBA? i would imagine that they have more of a negative impact than other outs (as a whole). I understand grounders could lead to double plays, flies to sac runs. Also are there different weights for different types of outs?
Good question. The creator of wOBA, Tom Tango, responded with:
Because the run value of a strikeout, overall, is similar to other outs. K with runner on 3B and less than 2 outs are deadly. Horrible. But, a K with a runner on 1B and less than 2 outs are preferably to groundball outs. Basically, they cancel out mostly.
And, of course, he’s right. Looking at the run values for batted ball outs and strikeouts in each base/out situation (2001-2009), we get the following situations in which the strikeout is worse than a regular batted ball out:
There are 24 different base/out states, and the strikeout is worse in ten of them. There are only a few situations in which strikeouts are, as Tango puts it, “deadly”—the aforementioned runner on second and third with one out, and a man on third with one out. Runners on the corner with one out is pretty bad, and then it becomes less damaging.
When all is said and done, these situations account for roughly 11% of a player’s total plate appearances, so these are relatively rare occurrences anyway.
But if you’re like me, you like to go after as many of the little things as you can for the sake of completeness. Let’s say you want to incorporate strikeouts into your wOBA—how do you go about doing it?
It’s actually quite simple.
I’ve extracted the following run values for 2010, as an example:
Home Run: 1.414
Reach on Error: .494
Unintentional Walk: .312
Hit By Pitch: .339
Non Strikeout Out: -.277
Since wOBA is a rate, just like on-base percentage, the value of the out in the numerator is zero. So, add the run value of the out (.277) to each event. You get this:
Home Run: 1.691
Reach on Error: .771
Unintentional Walk: .589
Hit By Pitch: .616
Multiply by the frequency of the event, add them all together, and then divide by plate appearances sans intentional walks and sacrifice hits. Call this rate X. Divide your desired league rate (for me, it’s .330) by X, and you get your scale. In this case, it’s 1.19. Multiply all of your run values by this number, and you get:
wOBA = (.90*1B + 1.25*2B + 1.60*3B + 2.01*HR + .92*ROE + .70*NIBB + .73*HBP - .02*K) / (PA - IBB - SH)
And now you have your “pimped out” version of wOBA. That’s all there is to it. At the most, you’ll see about a three point difference in the player’s wOBA, and about one run of difference. Seriously, that’s it.
And, of course, here's a nifty little calculator that'll take care of everything for you. As you'll see here, the difference is incredibly small. So really, it's not a big deal to ignore strikeouts when using a context-neutral method like linear weights and wOBA. But it can be done. When all is said and done, we're talking about a run or two of difference.
Also, it's worth noting that we're not limited just to the addition of strikeouts or even to the OBP scale. We can even make a weighted Batting Average Against for pitchers, using things like balks and wild pitches...really, any event that we wish to include can be included rather easily.
JT Jordan is an avid Giants fan and runs a blog at Triples Alley. He can be reached at email@example.com.