This diary is a thread that outgrew its holding pen. Here is the original comment on the Sep 10 game v Texas with Street's ultra-clutch 5-out save under some very trying conditions. Read the first Winexp articles to see what this is about.
Street Has a Nice Butt
Baby WinExp (v31) says (recall this doesn't account for defense, like Swish's nice plays; the software definitely has bugs, no wagering please, etc.)
- The three stars of the game are Huston Street (by a long shot), Michael Young and Hank Blalock.
- Interestingly, it didn't even rate DJ's dinger as the biggest play of the game. I believe the idea is that once you've got the bases loaded with no one out, you are going to usually score some runs, so DJ did great, but he had a lot of help.
- The A's offensive star was... Hatteberg? Huh? Basically, it loved his walk to load the bases before DJ's dinger, and it loved his double to score Payton and remain in scoring position with 1 out in the 6th increasing our lead from 2 to 3.
- It also thought Huston's biggest out was the second out of the 9th, with a runner at 2nd and 1 out, even bigger than his two big outs in the 8th.
Sep 10, 2005 TEX Michael Young +0.267 4
Sep 10, 2005 TEX Hank Blalock +0.211 4
Sep 10, 2005 TEX C.J. Wilson* +0.087 6
Sep 10, 2005 TEX Mark Teixeira +0.054 4
Sep 10, 2005 TEX Mark DeRosa +0.000 1
Sep 10, 2005 TEX Adrian Gonzalez -0.035 4
Sep 10, 2005 TEX G Matthews Jr. -0.129 4
Sep 10, 2005 TEX Gerald Laird -0.155 4
Sep 10, 2005 TEX David Dellucci -0.157 5
Sep 10, 2005 TEX Kevin Mench -0.161 4
Sep 10, 2005 TEX Kameron Loe* -0.262 30
Sep 10, 2005 TEX Alfonso Soriano -0.262 4
Sep 10, 2005 OAK Huston Street* +0.437 7
Sep 10, 2005 OAK Scott Hatteberg +0.155 4
Sep 10, 2005 OAK Dan Johnson +0.136 4
Sep 10, 2005 OAK Jay Payton +0.093 4
Sep 10, 2005 OAK Jay Witasick* +0.011 4
Sep 10, 2005 OAK Jason Kendall +0.010 4
Sep 10, 2005 OAK Eric Chavez -0.003 4
Sep 10, 2005 OAK Ricardo Rincon* -0.009 2
Sep 10, 2005 OAK Marco Scutaro -0.019 4
Sep 10, 2005 OAK Mark Ellis -0.056 4
Sep 10, 2005 OAK Nick Swisher -0.062 4
Sep 10, 2005 OAK Joe Blanton* -0.072 25
Sep 10, 2005 OAK Mark Kotsay -0.079 4
Sep 10, 2005. OAK v TEX. Bot 6.0. R:-3. B:1. M. Teixeira homered to deep right center, M. Young scored (0.402, +0.193)
Sep 10, 2005. OAK v TEX. Top 2.0. R:-1. B:123. D. Johnson homered to deep right, J. Payton, S. Hatteberg and E. Chavez scored (0.738, +0.177)
Sep 10, 2005. OAK v TEX. Bot 9.1. R:-1. B:2. G. Laird grounded out to third (0.114, -0.164)
Sep 10, 2005. OAK v TEX. Bot 8.2. R:-1. B:12. K. Mench struck out swinging (0.151, -0.133)
Sep 10, 2005. OAK v TEX. Bot 9.0. R:-1. B: . A. Gonzalez singled to right (0.315, +0.126)
Sep 10, 2005. OAK v TEX. Bot 8.1. R:-1. B:12. A. Soriano flied out to right center (0.284, -0.115)
Larry E Has a Nice Butt
Larry E wrote: Its too bad we don't have more enough data to do these stats based solely on games in Ameriquest Field. I think it would value pitchers getting outs even more. So Street's performance (even at .427)would be even more off the charts if the field of play was taken into consideration.
Baby WinExp replies:
Ooooh, that would be interesting to just look at site-specific data.
There are mathematical models that do adjust for different "run environment". I just haven't gotten around to writing those plug-ins yet.
Street Has The Nicest A's Butt since Zito's August Butt
Some context: it is HARD to rack up that much win expectancy contribution (WXC). Here are Baby Winexp's picks for the Top 10 individual game performances in A's games since Aug 1:
Date Team Name WXC Events
12-Aug-05 MIN Johan Santana* 0.83 31
4-Aug-05 OAK Barry Zito* 0.489 30
5-Sep-05 SEA Fel Hernandez* 0.461 26
3-Aug-05 MIN Michael Cuddyer 0.46 3
21-Aug-05 KC Jimmy Gobble* 0.451 9
1-Sep-05 ANA Ervin Santana* 0.444 31
10-Sep-05 OAK Huston Street* 0.437 7
30-Aug-05 OAK Barry Zito* 0.433 31
3-Sep-05 NYY Aaron Small* 0.432 33
So Street's save last night was the best WXC game by an A's player since Zito's Aug 4 masterpiece in MIN.
Why is Santana's Butt so Nice?
I was surprised by Santana's shockingly highly rated performance. I thought it had to be a bug, but it looks legitimate. Basically, he and Haren (+0.346) had a complete game awesome duel where the batters kept decreasing their chances to win by doing naught at the plate and the pitchers kept shutting down the other side and increasing their chances. Unfortunately, Haren gave up the run that counted.
Date Team Name WXC Events
12-Aug-05 OAK Dan Haren* 0.346 31
12-Aug-05 OAK Mark Kotsay 0.025 4
12-Aug-05 OAK Nick Swisher -0.003 3
12-Aug-05 OAK Bobby Crosby -0.023 4
12-Aug-05 OAK Bobby Kielty -0.052 3
12-Aug-05 OAK Crosby Error -0.058 1
12-Aug-05 OAK Dan Johnson -0.065 3
12-Aug-05 OAK Jay Payton -0.106 3
12-Aug-05 OAK Jason Kendall -0.174 4
12-Aug-05 OAK Mark Ellis -0.175 3
12-Aug-05 OAK Eric Chavez -0.257 4
12-Aug-05 MIN Johan Santana* 0.83 31
12-Aug-05 MIN Lew Ford 0.103 4
12-Aug-05 MIN Michael Cuddyer 0.088 3
12-Aug-05 MIN Justin Morneau -0.021 4
12-Aug-05 MIN Shannon Stewart -0.027 3
12-Aug-05 MIN Joe Mauer -0.058 4
12-Aug-05 MIN Matt LeCroy -0.073 3
12-Aug-05 MIN Michael Ryan -0.077 3
12-Aug-05 MIN L O. Rodriguez -0.118 3
12-Aug-05 MIN Nick Punto -0.163 4
The numbers check out too... the visiting team has about a .460 chance of winning, so you would think there's only .540 WXC to be collected. However, each half inning the Twins didn't score reduced their winning chances by about .05. Over nine innings, that's another .450 WXC or so to be collected by a pitcher at the plate IF your team does extremely poorly at the plate. Lew Ford (+0.103) and Cuddyer (+0.083) got the only hitting credit and Crosby got dinged for an error (+0.058 for MIN) , so Santana picks up the rest of the credit.
ps. If you got a different answer for the historical chances of the visiting team winning at http://walkoffbalk.com/tools/winexp/index.php, they are wrong. The actual data is different and they rounded off the number of games to 2^15 for some reason.
pps. Baby Winexp is cool, but still has bugs as does the play-by-play from sportsline.com. For any one game, my results will be pretty close to right; for 510 games (played league-wide since Aug 1), there are definitely errors which I haven't had time to look over, so take huge aggregate data with an extra three grains of salt.