February 7, 2010

The day is here.

Many of you know of this niche sport called American football. It's quite popular in some regions and creates quite an underground buzz during it's championship weekend. This year, two teams face off, and those two teams are the Colts and the Saints.

According to Vegas, and every analyst on ESPN, CBS, and FOX, the Colts should win. This is all fine and dandy. You can choose your picks on the basis of how good Peyton Manning is on SNL or how disheveled Pierre Garcon is about his homeless ancestors. Go ahead and do that. You just leave the statistical analysis to me.


Now, I'm not a math major, or statistics major, but to become a member of the business school at the University of Kansas, you have to take a statistics class. So this is why I have some accreditation. Normally here at the Perm, we refer to the National Association of Madeup Stats to get our information. This is actually not the case for once. Basically what I did was take the games where the Colts and Saints played the same opponents and broke down the score differential between the two teams, because they won every one. This is what I came up with.
OOOOO, numbers. I look so distinguished. The only thing you need to focus on is the P(T<=t) one-tall number. It tells you whether the difference in scores is statistically significant towards the Saints. For it to be statistically significant, this would need to be somewhere around .05. You can see this is no where near, and because the correlation is so far from 1, that means that this game is basically a coin flip. Also, you'll notice the variance on the Colts is extremely high compared to the Saints. That's because the Colts had a ridiculous 36 point win against the Rams, where the Saints had only a one point win versus St. Louis. Because those two numbers are so far out of the spectrum of the two teams averages, I decided to throw the Ram game out like a used condom. This is what resulted.
If you notice, the P (T<=t) one-tall has dropped significantly. This means, with the Rams game thrown out, that the Saints were significantly better than the Colts when playing the same opponents, which would also mean the Saints should be favored over the Colts.

Take what you will from this, because I'm an amateur at this, but statistically on this level, the Saints should be favored. This in no way considers the playoffs, or how well Pey-Pey is playing, but at the same time, it is statistically significant. Still, the original analysis brought up a coin flip so that's what I expect really. Not to mention, the Football Outsiders agree with me. All in all, I will be cheering on Dat Saints. WHO DAT? DREW BREES DAT!

No comments:

Post a Comment