Saturday, July 22, 2006

Overwhelming Stats

"We hope a big one," said Giants general manager Brian Sabean of Hillenbrand's expected impact on the club. "There's an overwhelming stat that I wasn't aware of until a week ago -- when we score four or more runs, we're a good team, and we need to do that in this ballpark.
Accardo for Hillenbrand and Chulk is a pretty fair deal, and it helps both teams; given the Giants' position in the race, paying $2m and change for the ~10-15 run improvement they can expect from the upgrade over Niekro is just fine, and the downgrade in the bullpen isn't a big deal. But for Pete's Sake, Sabean considers it an overwhelming stat that he has a "good team" when they score four or more runs? And until a week ago, he didn't realize that scoring more than four runs is a strong indicator of success?

What's the Giants' distribution of runs and wins? The stat Sabean refers to would show the Giants at 9-31 (.225) when scoring fewer than 4 and 41-16 (.719) when scoring 4 or more. Here's the numbers beyond that dichotomy:
Runs    Record   %
0-3 9-31 .225
4 4-7 .364
5 6-4 .600
6 11-2 .846
7+ 20-3 .870
I can't think of anything remarkable about that distribution for a team that allows 4.68 runs per game. They tend to win when they score more than they allow on average, and they tend to do poorly when not scoring much. You might expect a little more symmetry, but the Giants have outscored their opponents by seven runs and the deviation is miles from extraordinary given the sample size. And it is odd that Sabean would refer to their record scoring four runs or more when for just pennies more he could have boasted their .804 winning percentage when scoring five or more.

I'm not disturbed by the trade, nor am I disturbed that Sabean knows the teams' record with 4 or more runs. I'm merely disturbed by the following apparent implications of the quote:
  • that someone was responsible for providing this information to Sabean and didn't make any effort to explain to Sabean why this was without import
  • that Sabean was overwhelmed by a figure which could be completely explained literally by elementary arithmetic
  • that - unless the author took these quotations out of their contexts by grouping different parts of the interview into one paragraph - the Giants' record when scoring four plus runs was offered as a unique justification for upgrading at a position where they've got no talent.
On the bright side - for the Giants, anyway - Sabean's approach to numbers only made him look foolish (and perhaps just to me) in an article, instead of causing him to do something like give a ton of PA to Neifi again.


Post a Comment

<< Home