VT/Nebraska Statistical Comparison

It's still early in 2009 to read much into performance to date, but to settle our anxiety let's take a look at some statistical comparisons of the two teams.  While we all are familiar with basics such as yards of offense per game or yards allowed per game, these fail to take into consideration how good the opponent was - while Nebraska may have put up better numbers thus far, their opponents were relatively weak.

As football fans we like to lump everyone who is not a top team into one bucket, but by breaking this habit we can start to discern which team has actually played better to this point in the season...after all, you wouldn't argue that Florida and Kansas are equal teams so why do you allow yourself to believe that New Mexico State and Troy are?

Fortunately for us, statisticians have great ways to separate all teams by their strength.  The BCS <insert your own sarcastic/cynical/frustrated BCS acronym here> uses computers as part of it's rankings, but unfortunately refuses to take into account score differences even though it is trivial for a statistician to insert diminishing returns to make running up the score of little consequence.  There are a huge number of rankings out there that fit the BCS requirements and those that do not - without listing them, let's just say that they are for the most part consistent which is a testament to their accuracy (keeping in mind they are only as accurate as the result of games and do not account for unusual occurences such as a bad call by a ref that results in an undeserved win).

On to this week's game...

Quick trivia question - Where did Kenneth Massey, developer of one of the BCS computer models, attend graduate school?  That's right, VPI&SU.

So while I don't like the BCS ban on using winning margin, it is natural that we should first compare the two teams using Massey's Ratings (http://www.masseyratings.com/).  Fortunately Massey has his own rating system as well as his BCS, and only his non-BCS rating system will be listed here as the BCS system is essentially worthless with so few games played (unless you believe Division III Otterbein is the 16th best team in the country in which case I hope you enjoyed your four years at Otterbein).  Essentially here a higher number means a better team although conceptually there is no easy interpretation (he does provide details on making prediction on the website).

Next up are two rating from Jeff Sagarin (http://www.kiva.net/~jsagarin/sports/cfsend.htm), perhaps the most popular of all statistical rankings.  The first is his ELO ranking, which is used for the BCS and fits their criteria; the second is PREDICTOR which takes into account scoring margin and is much more accurate for determining how good teams are.  Even better, Jeff Sagarin is so curteous as to make a natural interpretation for us - the difference in ratings is the estimated score difference on a neutral field (lookin' your way, Ken Massey).  Don't be confused though and think he is predicting that many points for the team - only the score difference.

For now that's all we will use, but as the season progresses some more ratings become available that we can look at included Rothman's FACT, which was part of the BCS until the unfortunate rule against using scoring margin.

Without further ado, here is how we stack up:

Without pressing for details on the calculations, Massey has us winning by 3.2 and Sagarin has us by 3.9 (both taking into account home field advantage).  Suffice to say these are tight margins and this is anyone's game.

As this is my first Gobbler Country post, let me know if this is interesting and should potentially become a weekly column!