Win Probability Added leaders for 2010
Posted by Andy on September 10, 2010
We've been talking about WPA on another thread and I just threw together a quick list.
Click through for guys whose 2010 WPA total is at least 1% of their plate appearance total--in other words, guys who not only helped win games, but at the best rate relative to their number of plate appearances.
These 20 guys have all done it so far in 2010:
| Rk | Player | WPA | PA | Tm | G | AB | R | H | 2B | 3B | HR | RBI | BB | SO | Pos | |||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Miguel Cabrera | 5.929 | 582 | DET | 135 | 490 | 97 | 163 | 42 | 1 | 33 | 110 | 83 | 82 | .333 | .428 | .624 | *3/D |
| 2 | Josh Hamilton | 5.888 | 559 | TEX | 130 | 507 | 94 | 183 | 40 | 3 | 31 | 97 | 43 | 95 | .361 | .414 | .635 | *78D |
| 3 | Logan Morrison | 1.893 | 181 | FLA | 39 | 151 | 30 | 48 | 16 | 4 | 1 | 13 | 29 | 31 | .318 | .431 | .497 | *7 |
| 4 | Brooks Conrad | 1.724 | 148 | ATL | 90 | 128 | 26 | 31 | 10 | 0 | 7 | 27 | 15 | 39 | .242 | .326 | .484 | 5/4 |
| 5 | Donnie Murphy | 1.466 | 47 | FLA | 29 | 44 | 9 | 14 | 6 | 1 | 3 | 16 | 2 | 19 | .318 | .348 | .705 | /654 |
| 6 | Chris Nelson | 0.271 | 8 | COL | 7 | 7 | 5 | 4 | 0 | 0 | 0 | 0 | 0 | 0 | .571 | .571 | .571 | /45 |
| 7 | Darren Ford | 0.252 | 0 | SFG | 3 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | ||||
| 8 | Brandon Allen | 0.198 | 14 | ARI | 6 | 14 | 1 | 4 | 1 | 0 | 1 | 4 | 0 | 5 | .286 | .286 | .571 | /*7 |
| 9 | Matt Belisle | 0.160 | 7 | COL | 66 | 4 | 0 | 1 | 1 | 0 | 0 | 2 | 0 | 2 | .250 | .400 | .500 | *1 |
| 10 | Daisuke Matsuzaka | 0.114 | 5 | BOS | 21 | 4 | 0 | 2 | 0 | 0 | 0 | 1 | 0 | 0 | .500 | .500 | .500 | *1 |
| 11 | Gustavo Chacin | 0.093 | 1 | HOU | 38 | 1 | 1 | 1 | 0 | 0 | 1 | 1 | 0 | 0 | 1.000 | 1.000 | 4.000 | *1 |
| 12 | Joel Pineiro | 0.082 | 5 | LAA | 20 | 3 | 3 | 0 | 0 | 0 | 0 | 0 | 2 | 2 | .000 | .400 | .000 | *1 |
| 13 | Alexi Ogando | 0.076 | 1 | TEX | 32 | 1 | 1 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 1.000 | 1.000 | 1.000 | *1 |
| 14 | Michael Dunn | 0.057 | 2 | ATL | 15 | 2 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 1 | .000 | .000 | .000 | *1 |
| 15 | Kenley Jansen | 0.049 | 2 | LAD | 18 | 1 | 1 | 1 | 0 | 0 | 0 | 0 | 1 | 0 | 1.000 | 1.000 | 1.000 | *1 |
| 16 | Dillon Gee | 0.045 | 3 | NYM | 1 | 3 | 0 | 1 | 0 | 0 | 0 | 1 | 0 | 1 | .333 | .333 | .333 | /*1 |
| 17 | Tyler Clippard | 0.036 | 2 | WSN | 68 | 2 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 1 | .500 | .500 | .500 | *1 |
| 18 | Jarrod Dyson | 0.034 | 1 | KCR | 2 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 1 | 0 | 1.000 | |||
| 19 | Alejandro de Aza | 0.011 | 0 | CHW | 4 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 0 | /*D | |||
| 20 | Clay Buchholz | 0.011 | 1 | BOS | 24 | 1 | 0 | 1 | 0 | 0 | 0 | 0 | 0 | 0 | 1.000 | 1.000 | 1.000 | *1 |
Basically this is saying that by using this method, Miguel Cabrera and Josh Hamilton come out on top as far as regular players go.
Obviously, the sample sizes are extremely small for most of the guys on this list.
I can't tell you how glad I am to see Logan Morrison get off to a hot start. If Logan reads this, I was the guy in spring training yelling a lot of encouraging things at your while when you were playing first base 🙂
If you're curious about who did this in 2009, click here.
September 10th, 2010 at 4:12 pm
Interesting to see Votto on the list for 2009. Makes those counting stats a bit less unimpressive and his 2010 seem all-the-more like it doesn't necessarily have to be his career year.
September 10th, 2010 at 4:32 pm
Chacin? Really? I did *NOT* expect to see him and his 1.789 WHIP on this list.
September 10th, 2010 at 4:33 pm
Ha... just realized this was for offense. Durp.
September 10th, 2010 at 5:09 pm
Andy-
What is the scale for WPA? Like, is Cabrera's 5.9 mean he gave his team a 5.9% better chance of winning? Does that mean if he were on a team that otherwise had a perfect 50/50 shot of winning a game, they'd have a 55.9% chance of winning? Or would it be something like 53%, making a 53/47 split for the 5.9% difference? Just trying to wrap my head around exactly what the number itself means.
September 10th, 2010 at 5:34 pm
What is the scale for WPA? Does Cabrera's 5.9 mean he would turn an otherwise 50/50 game to his team's favor 56/44? Or would it go to 53/47? I love the idea of the stat, I'm just trying to make sense of it.
September 10th, 2010 at 6:56 pm
Andy, you needed to set a minimun number of PA for the list to get interesting.
So why are the NL MVP candidates not on this list? Has Pujols been not getting a lot of key at-bats?
September 10th, 2010 at 7:09 pm
Neil, I listed all players because there were so few. Note that in the original post I mentioned that many of the performances are not statistically significant.
September 10th, 2010 at 7:29 pm
@4 and 5,
WPA's scale is wins. It's just the sum of every plate appearance's WPA throughout the year.
September 10th, 2010 at 7:54 pm
this has been the greatest year for wpa ever. NOT!! why is cabby like 4 away from what prince did in 09. prince was surrounded by good hitters, while cabster has to pull his whole team and should be more valuable.
September 10th, 2010 at 8:00 pm
@7
Andy, so you did. However, I would still like to see where some of the "big" current Yankees (Rodriguez, Cano, Jeter, Ortiz) and other apparent offensive stars appear on the list. This is not possible when you include fringe batters.
September 10th, 2010 at 8:01 pm
Sorry... of corse Ortiz, Sox!
September 10th, 2010 at 10:07 pm
Lee-
Isn't that WAR?
September 11th, 2010 at 1:45 am
BSK,
Take the probability of your team winning before each play and after. The difference is the Win Probability Added, (plus or minus), for that play, for the batter or baserunner. A home run in a high leverage situation could add .3 or more. A double play in the same situation could have a negative as large. Add all them up. So each positive integer represents 100% of a win added. So Cabrera's positive actions minus his negatives have come to a net plus of 5.9 wins for Detroit. In every game at the end the winner's WPA is +.5 and the loser's -.5. (Pitching included.)
September 11th, 2010 at 3:04 am
As a Braves fan, I can tell you that the numbers back up what my eyes have told me about Brooks "Cajones" Conrad. That dude has done some major damage all year, although nobody's quite sure how. It's been fun to watch (and root for!)
September 11th, 2010 at 6:17 am
KDS-
Got ya. I knew the basic idea of the stat, but didn't know exactly what the numbers meant. So we're still working on a true percentage with 1 = 100% and 5 = 500%. Thanks.
September 11th, 2010 at 10:48 am
Half these guys or more are so unknown to me they might as well be in witness protection. I think that means that WPA when taken on small sample size is purely luck.
I wonder if career WPA would substantiate people's thoughts about one player or another being truly clutch? Frankly, I've never really paid much attention to the stat, so I would have no idea.
September 11th, 2010 at 1:34 pm
@13
BSK, got a thick skull here. Just trying to understand. So the pitchers' total WPA for the winning team is the negative of the WPA for the losing team's batters and vice versa?
That way winning team WPA = 0.500 and losing team WPA = -0.500, including BOTH pitchers and batters?
Why doesn't somebody put the relationships between these stats up somewhere that is easy to find?
September 11th, 2010 at 2:34 pm
Neil L.,
Yes, you got it.
I recommend looking at box scores here.
September 14th, 2010 at 11:34 am
Great addition - please add BA-RISP column to results set to provide the user context for clutch & WPA.
September 15th, 2010 at 1:26 am
it would be nice, if we could search the data base and sort the results by the formula we want to include in the restrictions. In this particular case the search should have been minimum 500 PA sorted by (WPA*100)/PA. This would result in a list led by the same 2 guys Hamilton and Cabrera but followed not by a bunch of small samples instead the ones that came close to the club(of 1%).