A run saved is a run scored?

www.hardballtimes.com — Does a run saved equal a run scored? Maybe not. Click the title to read more. Order the Hardball Times Annual 2009 today !

MLB

vote down

The Hardball Times found this 12/4/2008

4 Comments

studes

+1

Depending on how you look at it, a run scored isn't equal to a run allowed even in the "pythagorean" sense.
If you take a .500 team (runs scored = runs allowed) and add a run to its RS, you'll find that the impact is less than if you subtract a run from its RA. Or use 10 or 100 runs to more easily see the difference. This is the essence of David Gassko's Pitching Runs Created.
This isn't really related to your analysis, but I thought I'd point it out. I think what you're saying is that the impact is even greater than that.

Posted 12/4/2008 respond (flag)
pizzacutter

+1

Colin, good work on this one. I've also found that Pythag residuals are mostly correlated with the average margin of winning and losing. I looked a little further and found that it doesn't seem to be a reliable trait of the manager to win close games.

Posted 12/4/2008 respond (flag)
JinAZ

+1

Nice little study. A while back I took a different look at much the same thing.
I ran a regression comparing runs scored and runs allowed to wins on 1996-2004 data. I found that each run scored correlated to an extra 0.099 wins, while each run allowed correlated to 0.101 fewer wins. I treated them as identical, but the direction is the same as you found: a run allowed has a slightly larger effect on wins than a run scored.
Ballhype isn't letting me post a link, but it's in the first article of my player value series. -j

Posted 12/5/2008 respond (flag)
angelsdust

+1

"Doing this provided me with over 213,000 matched pairs, a very robust sample for testing. When the won-loss records are tabulated, the offensive group of teams won at just below a .5016 clip. The defensive group of teams won at a .5021 clip.... We’re talking roughly one game won every 12 seasons between our defensive-minded teams and our offensive-minded teams...Using the Student’s t-test, though, we can see that while the difference is not substantial, it is statistically significant—that is to say that we don’t think it’s simply an accident of small sample size."
I'm afraid your argument is not valid. If you have a large number of observations, you are bound to get a significant t-Test and 213,000 is very large. In such a case, you have to look at the coefficient and decide whether it is relevant or not and one win in 12 years is not in my opinion. This does not necessarily mean that your findings are wrong, of course.

Posted 12/10/2008 respond (flag)