Baseball Musings: Run Estimation

February 07, 2005

Run Estimation

Curveblog has a couple of post up on David Eckstein's range and what it's costing in terms of runs. I have to disagree with something:

re the cost of dave eckstein’s poor range, a poster calling him/herself "anolis" asks an intelligent question:
"If I read this correctly, Eckstein made 44 less outs than might have been expected, and the analysis suggests that this translated into allowing 33 more runs that expected. Can this be correct? Since the outs that Eckstein missed almost surely turned into singles, is the expected value of turning an out into a single 3/4 of a run? Seems high to me."

That seems real high to me, also. Unfortunately, I can't think too well right now (as I write this I find I wish I hadn't started, it's very difficult to concentrate). But 44 outs means that Eckstein is adding less than two extra games of offense to his opponents. If the opponents score about 4.5 runs per game, than Eckstein should cost his team 4.5*(44/27) runs, or a little over 7 runs.

Bill James showed over 20 years ago that the difference between the best and worst fielding shortstops couldn't be more than 25 runs a season. So I don't think David's fielding cost his team anywhere near 33 runs.

Posted by David Pinto at 11:50 AM | Defense | TrackBack (0)

Comments

Could you also estimate the net result based on the linear weight value for a single? (I believe that's around .41 runs per single). That way, it would end up 18.04 runs.

Posted by: Mike at February 7, 2005 02:05 PM

Let me clarify: I don't think the difference is 44 additional outs for Eckstein's opponents; it's 44 additional outs in which they have a man on first base.

Posted by: Mike at February 7, 2005 02:08 PM

44 outs *does* translate into 30-something runs. Here's how you do it:

===============
Suppose a team with Ozzie at SS gives up on average 12 non-HR hits, and 2.6 walks every game (which of course is 27 outs). Applying .50 runs per non-HR hit (I know it should be closer to .55, but I just want to keep it basic), and .30 runs per BB, and -.10 runs per out, and we get 4.08 runs scored per game. And per game, we see that Ozzie's team faces 41.6 batters (again, let's not worry about DPs, etc).

Now, let's say Ozzie was traded for Spike, and let's say for every 41.6 batters faced, there is one ball that Ozzie gets to that Spike doesn't. So, for those 41.6 batters, Spike's team records 13 non-HR hits (1 more than Oz), 2.6 walks, and 26 outs (1 less than Oz). However, there's still one more out to go! Since Spike's team gives up 13 non-HR hits / 26 outs, we can estimate that this team will give up 13.5 non-HR hits, 2.7 walks, and 27 outs per game ( a total of 43.2 batters, a remarkable 1.6 MORE batters than Oz). Anyway, applying our LW constants, and we see that Spike's team gives up 4.86 runs per game.

This number is .78 runs MORE than Ozzie. This is the result of Ozzie getting to one more hit than Spike. .50 runs for the hit, and about .30 runs for the out gives you the .80 runs.

Posted by: tangotiger at February 7, 2005 04:07 PM

It can be shown in a multitude of ways that the range in true talent level for fielding is +/- 25 (or so) runs per 162 games.

However, in any given sample, if the true rate is +/-25, then we should be observing larger ranges, just by luck.

Posted by: tangotiger at February 7, 2005 04:21 PM

I'm getting 16.97 runs based on what Chronicles is doing.

Posted by: Richard at February 7, 2005 11:12 PM

Right, the 16.97 was based on the average of the players in 2004 only. It's in the 30s against the 2002-2004 average, the "predicted" figure David has been using.

As for the run value, the other way of looking at it, which was how it was introduced to me, is that an out not recorded by a SS is worth ~.47 (the linear weight for a single) PLUS ~.28 (the linear weight value for an out), which is ~.75.

Posted by: Black Hawk Waterloo at February 8, 2005 01:09 AM