BaseballGeeks
User: Guest39055
Posts: 0
Read Mode: Topics
Status: Guest
  BGEEKS Stats ~ Sign-Up As New User ~ Message Areas ~ FAQ

BaseballGeeks.com Home : BaseballGeeks : Atlanta Braves : Reading "Win Probability Added " Thread

Message #9 of 26  *NEW*
To:  All
From:  
Newey  
Subject:  Win Probability Added
Date:  5/21/06, 10:00pm (Last Edited: 5/21/06, 11:18pm)
graphic
Welcome to the wonderful world of Win Probability. Not only a fan of alliteration, I happen to enjoy the endless amount of numbers and metrics that baseball brings as well. Finishing up a math minor as well as my mechanical engineering degree in a few weeks, I was drawn to an article from Dave Studeman of The Hardball Times. It's an interesting way to look at the ups and downs of a game in progress.

http://www.hardballtimes.com/main/article/the-one-a
bout-win-probability/

Using historical data, a model has been developed to create the probability that a team will win for each and every base/out/inning /score situation. I'll save the rest of the details to the above article. I'm sure it's a better explanation than I could muster at this point.

But anyways...on with the show. Studeman has also been kind enough to share an Excel workbook to track games as they happen. It has a very easy user-interface and I recently decided to follow along on Friday night while my Braves took on the Diamondbacks.

The excel workbook is at ftp://ftp.baseballgraphs.com/wpa (you may need to enter the address manually instead of this link b/c of the FTP instead of HTTP address)

And I have to admit my spirits went as far up and down as the peaks and valleys on this graph. Between applauding the efforts of Chipper Jones and Adam LaRoche and yelling less than complimentary things at Chris Reitsma, I watched the Braves lose a back-and-forth battle to open up their nine game road trip.

But therein lies another beauty in Win Probability Added. Not only can you chart the team's potential level of ultimate success, you can also determine each player's level of contribution to the win/loss. So all the ire and objects you can throw at the television at your less-than-favorite closer after blowing another game can be justified with numbers. I always knew there was a reason I liked math.

The first climb on the chart resulted from the Braves beating up Miguel Batista on 3-days rest in the first inning. Adam LaRoche's first inning grand slam made the winning percentage for the Braves rise from .604 to .840, a +0.236 increase. The Braves kept up their winning percentage while the Diamondbacks chipped away until the 6th inning. And that's when John Thomson decided to let it slip away.

He retired the first two batters and then, following a walk and a single, he gave up a 3-run bomb to Jeff DaVanon. D-backs 7, Braves 5.

The Braves would tease me in the top of the 7th as they managed to load the bases with no one out (the upward spike between the 7 and 8 in the graph). But Brian McCann traded one run for two outs with his infield grounder. And Jeff Francoeur grounded out as well and the threat was gone with only putting one on the board.

The Braves and D-Backs would trade runs against mediocre middle relief in the 7th and 8th, with Chipper Jones' 2-run double knotting the score at 8. And after 8 innings of West Coast excitement, with a total of 16 runs scored, our probability graph is back where we began. It really was a whole new ballgame. Which is probably why LaRoche's RBI single in the 9th made the eventual loss all the more painful. Because with a win expectancy of .817, someone was going to have to do something devastating to make that number fall to 0.

Enter Chris Reitsma. I've never been one that's particularly fond of Mr. Reitsma, regardless of how many times my roommate here in Cincy tells me that Reitsma was once "his boy" as a Red. He's just never been able to shut down hitters with the kind of Rivera-esque quality. He didn't work last year and we brought in Farnsworth. I only hope that Schuerholz has some ace up his sleeve again this year. Because with Atlanta relievers blowing 11 of their first 19 save opportunities (only 3 by Reitsma, to be fair), we're going to need some better arms yet to make it to October.

But back to the game at hand. Luis Gonzalez leads off with a double. Conor Jackson flies out and we're now two outs away from victory. But Shawn Green doubles home Gonzo and we're deadlocked at 9. Former Brave catcher Estrada gets intentionally walked and the soft-hitting Orlando Hudson flies out. So with two outs and pinch-hitter Tony Clark up, what does Reitsma decide to do? He hits Clark with the 0-2 pitch. So now the bases are loaded and he's in even more trouble. Because up next...He-Man.

Craig Counsell, holding his Louisville Slugger above his head like the Sword of Power, aptly earned this nickname amongst my roommates during the 2001 World Series. The guy continues to look ridiculous and continues to find a way to come up with big hits. So, He-Man did what He-Man does best and got a base hit to end the game.

And so what began as a quick note about Win Probability I'm sure turned into a longer, rambling post than I first thought. I wanted to introduce some of our readers to this new tool. Check out the Excel file if you want or if you're too lazy to do that (which I admittedly am at times), you can find the win probability graph as well as a table for individual contributions for every game at the excellent FanGraphs website.
http://www.fangraphs.com/wins.aspx?date=2006-05-19&
team=Braves&dh=0

The following is the Win Probability Chart for the game I mentioned earlier, courtesy of FanGraphs.


Previous Topic
Next Topic
Reply to Msg #9
Start New Topic
Back to Topics

3034 reads

RRBBS© Software, designed by Robert Reed.
Copyright, November, 2002. All Rights Reserved.

Site Meter