Sunday, June 24, 2007

What Price An Out?

Unlike all the cool blogs, I haven't done a whole lot with Win Probability Added or Win Expectancy this year, other than get into an extended debate with some of its biggest proponents when I had an aside in an unrelated post...

I think it's interesting, but, for me, it's just not something I want to or need to look at every day. To each their own. But in cases like Saturday night's game, it's kinda fun to look at.

It's a pretty intuitive measure of what's happening on the field. And, yes, if you're one of those people who hates when baseball is reduced to petty numbers, you're going to loathe this.

The stat is a way of measuring the probability of the actual events on the field. If a player hits a solo homer in the first inning, it doesn't mean nearly as much as if he hits it in the bottom of the 9th with his team trailing a run. Sure, they both count as a run, but one has a significantly larger impact on the team's fortunes that day. This is just a way of measuring how large that impact is, based on how things have played out in the billions of baseball games played since the beginning of the time.

For the purposes of this post, I'm using the numbers fount at Walk Off Balk. These differ from the ones found at FanGraphs, which is probably the largest pusher of this stat. The WOB site uses real-life game results from 1977 to 2006.

The Indians started the 9th inning yesterday down by 2 runs. According to WOB, visiting teams have won the game just 5.5% of the time in that situation. Of course, the typical team doesn't have a flat-fastball-throwing Chad Cordero on the mound. After two singles put runners at the corners, the Indians (as you'd intuitively expect), had greatly improved their chances of winning, but the Nats still had 'em by the balls. With runners on the corners and nobody out, visiting teams trailing by two have won 25% of the games they've played. Those two singles, upped the chances by 20%, a pretty sizable move.

Of course, Victor Martinez' blast did even more damage. His three-run homer erased the two-run deficit and gave the Indians a lead, dramatically increasing their odds of winning. They went from a likely loss to a likely win, as the Nats' chances of winning dropped all the way to 15%.

So Martinez single-handedly increased the Indians chance of winning by 60% (25% before, 85% after). It's hard to have more of an impact than that.

As the Nats entered the bottom of the ninth, it wasn't impossible for them to win.

Home teams trailing by 1 in the bottom of the 9th start out with about a 19% chance of victory. When the Indians intentionally walked Cristian Guzman to load the bases with one out, the Nats had actually shifted the odds in their favor, and now had a 53% chance of winning. Then came the play.

Felipe Lopez grounded weakly to the pitcher for the force at home and Nook Logan got caught rounding third too far for the game-ending double play. That 53% chance of winning, of course, dropped all the way to zero.

Lopez deserves some of the scorn, but how much should we heap on Logan himself?

Had Logan done the right thing and held at third, the team would've still had the bases loaded, just with two outs. In that exact situation, home teams still win 25% of the time. Remember, too, that that's the average home team, and that Ryan Zimmerman would've been at the plate, who certainly feels like he's better than the 'average' batter in those situations.

One base-running play killed the team's chances. Oh well. Those losses could come in handy come draft time next year. ;)

  • Here's the FanGraphs play log of the game. Unfortunately, it doesn't break the Logan/Lopez play down into its component parts, instead saddling Lopez with the blame for the wacky GIDP.

    Or if you're a more visual person, here's the actual graph, which looks like the roller coaster from hell.

  • 2 Comments:

    Post a Comment

    << Home