Wednesday, July 18, 2007

Storm clouds

I love statistics. I'm definitely a Bill James kind of guy. So I'm going to rain on our parade just a bit. Pythagorean wins is a stat developed off of the true statistical concept that generally speaking points for and against is a more useful predictive strategy year to year than wins and losses. In other words, if you out scored your opponents by a wide margin, but lost a lot of close games, the next year, your team should perform more on line with the scoring difference than the record. Don't argue this. It's a fact. It doesn't ALWAYS happen, but it does usually happen. PF and PA are more useful for figuring records for 2007 than just looking at 2006 wins.
Check out this article exploring adjustments to the formula. It explains the whole formula (thanks to the for the original link):

If you read the article and look at the stats, you'll notice that the Colts 'expected' win total was 9.6. That means that they should have been 10-6 just based on PF and PA. In fact, the Colts had the lowest Pythagorean win total of any champion in history. All of this should temper our expectations somewhat.

Of course, these stats aren't a rule. This doesn't mean the Colts will regress. I think there are good reasons why the Jags looked better on paper (more P wins), but couldn't put together wins. They are coached by a horrible game manager and have a dicey kicker. Of course the Colts clearly would not have won the Super Bowl had they played at the same level they did during the regular season. Defensively, they found a 'higher plane of existence' so to speak. I think we all know that the 07 Colts will not have a repeat if they finish 32nd against the run again. I personally expect them to improve dramatically on defensive. That is to say, they will have a middle of the pack D against the pass and the run. I think this will lead to larger margins of victory (if not more actual wins...12-4 will be hard to top with the schedule this year). Larger margins of victory in the pros DOES mean that a team is better.

All this to say, I'm optimistic, but do recognize that the numbers are somewhat stacked against us. We won a lot of games last year just because Peyton is king (Denver, Jets, Tennessee). I expect that to continue, but know the numbers have to improve. I don't think we can make history twice in a row.

Also, the FO site ranked all the wideout corps in the NFL. The Colts were 2nd behind Dallas. Before you get up in arms, recognize 2 things: 1. Their rankings are based on their statistical system which is very reliable and 2. It's impossible to appropriately measure Gonzalez's value until he plays. Unlike most internet hacks (hey!), these guys don't operate just on random opinion, but use real data and facts. As they point out in the rankings, the numbers for the WR on Dallas improved dramatically once Romo started QBing. Love him or hate him, Romo was a vast improvement over Bledsoe, even if he doesn't prove to be the second coming of Roger Staubach.

1 comment:

coltsfanawalt said...

I thought I was fairly smart until I read this article..