Baseball Run-Scoring Equations

[owlcroft]

I have now had the gift of a modest but useful database of individual innings, which covers the period 2000 - 2008 inclusive. I have posted an update at the page on Tom Tango's site, where the ongoing "discussion" is taking place. But the short form is this (I wish I knew how to get monospace in these posts):

Innings: 389,042
BR Average Error: 0.232548157 -------
TOP Average Error: 0.241989297 +0.00944114
XR Average Error: 0.261624709 +0.02907655

Besides the point that .009 of a run per inning is not exactly a killer difference is my belief, noted on Tango's page, that over a fuller period (ideally the 55 years I used for team-season-level evaluation), the BaseRuns advantage, trivial as it is, would shrink and possibly even reverse, because the optimized coefficient in BR looks more sensitive to the run environment than do the coefficients for TOP. Maybe someone will send me a larger DB.

This is not the first time I have had less than civil responses over there (and recall, I didn't make any initial post there: someone over there brought the matter up, in this case following up an article I had done for The Baseball Analysts). I had an even worse go-round over the Steroids issue and my associated site. Some of the suggestions put forth to handwave away the rock-solid, indubitable fact that player power has not gone up during the so-called "steroid era" would make a cat laugh (like, the baseball was being gradually de-juiced on a graduated annual basis to hide the effects--oh, me, oh my).

It is always hard to make nice judgements when one's own ego is involved, but I get the very definite feeling over there of it being a sort of "Tubby's Klub House", with "outsiders"--or their ideas--not welcome: if it wasn't invented here, it is The Work of Satan. And, apparently, BaseRuns, because Tango likes it and has praised it, is the fair-haired, blue-eyed princess, and anyone suggesting that something else is as good or better is Ming the Merciless come to ravage the Princess. Just a feeling I get . . . .

(If you think otherwise, don't stand on ceremony to not say so. I have what Nero Wolfe refers to as "a robust ego".)

[DavidTate]

Sorry for the delayed response, Eric -- life intervenes. I did read the original thread, though, and I'll try to catch up on more recent entries and post a comment or two here. I regret to say that I think tangotiger et alia had some legitimate beef, though I think you caught the fallout (in terms of tone) from past frustrations. I'll try to expand on that maddening teaser by the end of the week.

[DavidTate]

...and here I sit, 3 years later, having never followed up on that promise. Mea culpa. Probably best to let it lie, at this point.

[owlcroft]

I guess.

Here, for tangential interest, are the results from the 2012 season:

"Error" Type (calculated vs. actual) Absolute size of "Error" Percentage size of "Error" Cumulative "Error" (+ and - cancel)
Batting Runs, from TOP.........................14.6 runs.........................2.1%.......................6 runs
Pitching Runs, from TPP........................20 runs...........................2.8%.......................6.7 runs
Games Won from R/OR..........................2.8 wins....................................................-0.2 wins
Games Won from TPP/TOP......................3.27 wins...................................................-0.07 wins

(Sory about the dots--it doesn't format otherwise.)

An average error of 2.45% is about as good as it will ever get with any formulae. (Every year the pitching runs has a higher error, so it is probably something systematic, and equally probably something small and silly that I am overlooking, but frankly, my dear, I no longer give a darn (or heck).

[jessicasmith]

very nice forum