Another Phillies Myth Busted

After a disappointing 8-7 loss to the Washington Nationals on Thursday night, manager Charlie Manuel accused his Phillies team of playing down to their level of competition. Per David Murphy:

Afterward, Manuel lamented those struggles, saying the Phillies were guilty of playing down to their competition while falling to 79-59 and seeing their NL East lead drop to five games over second-place Florida.

So, I threw the Phillies’ stats in Excel and ran some numbers. The following two graphs show the correlation between opponent winning percentage and runs scored and runs allowed per game. If Charlie’s hypothesis is correct, we’d expect that as an opponent’s winning percentage increased, so to would run-scoring (first graph), and runs allowed (second graph) would decrease.

There is just about zero correlation in both graphs. The Phillies don’t play down to their opponents, and nor do they play up to them.

Maybe there’s a correlation in specific offensive categories like OBP and SLG?

  • OBP: .0001 r-square to opponent winning percentage
  • SLG: .0064 r-square to opponent winning percentage

For Phillies’ pitchers…

  • Opponent OBP: .0004 r-square to opponent winning percentage
  • Opponent SLG: .00002 r-square to opponent winning percentage

The final nail in the coffin is that there’s only a .0005 r-square between the Phillies’ winning percentage against teams this year and those teams’ real winning percentage.

Leave a Reply

*

2 comments

  1. Matt Swartz

    September 12, 2009 10:40 AM

    Crash, this is a cool thing to check out, but it sounds like he’s totally right. Playing down to their competition doesn’t mean performing worse against them– it means playing no better against them and achieving similar results. This seems like he hit the nail on the head! Is there any other team with such a low correlation?

  2. Bill Baer

    September 12, 2009 12:08 PM

    Playing down to their competition doesn’t mean performing worse against them– it means playing no better against them and achieving similar results.

    I don’t know, I guess I have a different definition of “playing down”. The variance in performance from game-to-game, I would presume, would have more to do with “playing down” than an actual mental complex being described.

    Is there any other team with such a low correlation?

    It’s way too time consuming to check the other teams, but we would definitely notice any statistically significant correlation in this regard.

    It’s also kind of hard to discern because we are using actual results. That means that a team with, say, a true talent level at a .550 winning percentage that also plays down to the level of its opponents might actually be a team with a .500 winning percentage. Maybe third-order wins would help in this regard?

Next ArticleHasty Judgment