Add new comment

1

...it's not random to be inconsistent.  That's a spectacularly inaccurate use of statistical terminology and you shuold know better.  Random variation has a consistent pattern of spread when enough samples come in.  The argument was that the specific model the Mariners use to build a pitching staff (stars and scrubs) is more likely to produce a series of run allowed figures that are more skewed-right than the league average bias would be given the Mariners' average RA/G.  No other AL team allowed as few as the 4-ish runs per game that the Mariners allowed.  The variance of RA is directly propotional the mean of runs allowed.  So you need to adjust the data to account for the change in variance expected when you decrease RA from the league average to the Mariners' average.  Of course the Mariners' pitchers didn't allow more than 12 runs while other teams did...other teams had worse team defenses than did the Mariners.  Frickin' DUH, Sandy.  But if you check the Mariners' standard deviation of run allowing vs. the all time major league average standard deviation for run allowing for all teams who allowed a similar number of total runs per game, you're GOING to find that Seattle was less consistent than the average 4 RA/G team.

 inserted to fix formatting...
This is what happens when people start looking at the surface data willy-nilly with zero real understand of the statstical context.  Your 30,000 foot view approach is a good starting point and you are eminently logical in your thinking, but this type of tactic will fail EVERY. SINGLE. TIME. if you don't understand the model.  Which you apparently don't.

format fix again
To your larger points...the Mariner team offense was unusually consistent...that is a major component for beating pythag.  Studies HAVE INDEED SHOWN that consistent offenses tend to stay consistent when the personnel don't change a lot.  Do some background research before you go mouthing off about there being zero evidence.  Thank you.

 format
On the defensive side, the assumption implicit in Pythag is that the standard deviation of performance will be a constant that is proportional to the mean of performance.  NOT that the standard deviation will be constant for all run scoring contexts.  Your attepmt at analysis would be right if we expected the Mariners to have the same variance in RA as the Rangers...but we DON'T...and neither does Pythag.

 format
I'll post more on this at a later time.  You have a point about one-run games...and that compoent isn't likely to persist (but if you correct our 1-run game performance we drop from 35-20 to 29-26 (Baseball Prospectus had a study showing that efficient defensive clubs beat the expected 1-run performance consistently)...that's 6 games...we'd still be near .500...and we'd still be beating pythag).

Filtered HTML

  • Web page addresses and e-mail addresses turn into links automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <blockquote> <code> <ul> <ol> <li> <dl> <dt> <dd><p><br>
  • Lines and paragraphs break automatically.

Plain text

  • No HTML tags allowed.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.

shout_filter

  • Allowed HTML tags: <a> <em> <strong> <cite> <blockquote> <code> <ul> <ol> <li> <dl> <dt> <dd>
  • Lines and paragraphs break automatically.
  • Web page addresses and e-mail addresses turn into links automatically.