Comparing Conferences: So much movement, so little reason
2009 Random Walker Rankings (RWFL, p=0.75)
Games through Saturday October 3rd
Conference Rankings (Average Per Team):
Only a week ago, the rankings made it look like the ACC was hands down the weakest of the so-called major conferences. Sure, VT and Miami were highly ranked; but on average, the squads in the ACC garnered fewer net RWFL votes per team. But here we are only a week later, and the same methodology puts the ACC very slightly ahead but in essentially a dead heat with the Big 12 and the Big East (varying the bias value p in the plot at the bottom of this post does change things, but not as vigorously as last week).
How is this big change in one week possible?
First, it would seem easy to go to my favorite rationalization: it's still early in the season. The problem with that argument in this case is that last weekend resulted in very little new information about the ACC's strength relative to the other conferences, with 10 of the 12 teams playing against each other.
Okay, so it must be those two interconference games? Sure, Georgia Tech beat Mississippi State, and of course the big win was Miami over Oklahoma. Those two ACC victories over SEC teams certainly move up the ACC rankings, especially the win over a highly rated Oklahoma team (starting QB or no). Such apparent sensitivity of rankings to a few interconference games only highlights the difficulty in ranking teams from the limited information that the BCS Standings allow.
Did anything else happen to cause this change in the rankings? Very possibly. The ACC was likely also helped this week by intraconference outcomes changing the rankings inside the conferences. For instance, Florida State's intraconference loss to Boston College further suppressed their RWFL rating, thereby decreasing the newly-assessed value of South Florida's interconference win over Florida State the week before. Reshuffled comparisons like this are happening all throughout the season, potentially changing the relative rankings of conferences even in the absence of direct matchups. The potential importance of such indirect effects make attempts to rank teams both interesting and maddening.