October 25, 2002
More on aggressive baserunning

I wrote that last post in a bit of a hurry, so this morning I'd like to spend a bit more time with the subject of aggressive baserunning and risk/reward ratios. Of the three aggressive baserunning decisions the Angels made last night when the game was still within their reach (Palmeiro legging out a double, Eckstein going from first to third on Salmon's single, Eckstein scoring on the wild pitch), the most defensible one is Eckstein going from first to third on the single. There are advantages to having that runner on third with less than two outs. One out is generally considered the optimal time to try for third on a single - the adage is "never make the first or last out at third base".

What all of these situations have in common, though, is the risk of an extra out for an extra base. That's true even for the wild pitch where Eckstein scored. Had he stayed put, the Angels would still have a man on third with less than two outs and their 3 and 4 batters coming up. They'd even be out of the doubleplay sitation, since Salmon advanced to second on the play. You've got to like youir chances to score there.

It's the risk of the extra out that requires a high probability of success. Take a look at the run expectation chart in this ESPN article by Michael Wolverton of the Baseball Prospectus. In 2002, the expected value of runners on second and third with one out is 1.358 runs. The expected value of a runner on second with two outs - the result if Eckstein is thrown out - is 0.322 runs, less than one fourth the value of staying put. That means that Eckstein needs a better than 80% chance of scoring in order to break even.

Aggressiveness has a much better case for it when you're not risking an extra out. Two situations from the 2000 playoffs illustrate this. In the divisional series between the Mets and Giants, the Giants had Armando Rios on second with one out. The batter hit the ball into the hole between third and short. Rios ran, shortstop Mike Bordick threw to third, and Rios was nailed. The Mets went on to win the game, and Rios was roundly criticized.

Problem is, by any reasonable definition, Rios made a rational decision. The math works out such that he needed only a 15% chance of being safe for it to be worthwhile. The reason for the shift in the odds is that if Rios stays put, Bordick throws to first and the out is recorded anyway. Rios is risking a base, but not at the cost of an out. Joe Sheehan explains it as follows:

I know some of you may disagree, but I don't think there's any way that Rios had less than a 1-in-6 chance of being safe. I'll concede that there were two good defensive players involved, and that the situation could possibly have called for a conservative approach. Still, if you look at the big picture, it's not a bad play.

If anything, Mike Bordick made the riskier decision. If he goes to first base, no one says anything. But if he one-hops the throw or hits Rios with it, he's got a shot to be Bill Buckner. OK, with knees.

And I guarantee this: had Bordick hit Rios with the throw, or Rios's foot kicked the ball out of Robin Ventura's glove, leaving the Giants with first and third and one out, all we would have heard was about "putting pressure on the defense" and Dusty Baker's aggressive attitude rubbing off on his players. Those things that sportswriters love to talk about, speed and hustle and forcing teams to make plays? That's what this was. The Mets made the play, won the game, and now get to be Bird feed.

The key is that now it's the defense that's risking an out in order to prevent an extra base. Playing with the house's money, as Rios was doing here, is always a good bet. The fact that this particular gamble didn't pay off doesn't change the fact that it was still the right move to make.

An almost identical situation came up in the World Series. The Mets had runners on second and third with one out. They were down by a run and the infield was in. The batter hit a ground ball to the second baseman and the runner at third held. Here's Joe Sheehan again:

The Mets had second and third with one out in the ninth when [Timoniel] Perez hit a two-hopper to Jose Vizcaino at second base. Todd Pratt held at third base, Perez was out at first base and the Mets didn't score in the inning.

This is ground well covered, but even granting that Pratt moves slower than campaign finance reform and Vizcaino was in on the grass, the contact play has to be on in that situation. There's virtually no downside--the worst-case scenario is first and third with two outs--and the upside is a run and the potential for more.

Holding Pratt was a needlessly conservative move. A nod to Tim McCarver, who was all over this one.

At least this time the announcers understood the proper strategy.

Baseball is a long-run game. Teams that look for and exploit small advantages win more games over the course of the season. (It's a lot like bridge in that regard.) Outs and bases are the currency, and needlessly risking one for the other is almost always a bad idea. It's amazing how often in a short series this gets forgotten, especially when the result is against the odds. David Eckstein will get lauded for his hustle while Armando Rios got criticized for his. Maybe results are all that matters. I'm just saying that going with the odds is usually the best way to get those results.

Posted by Charles Kuffner on October 25, 2002 to Baseball | TrackBack