Early in the season, the RPI is very much in flux. As a result, I like to avoid using parts of a team's profile that depend on it. Unfortunately, that's almost everything.
From the beginning of the season to this point, the model has been using a very small set of data (Sagarin ratings and coaches poll only) in the results. The RPI is settling down and many of the major conferences are close to one-third of the way through their conference schedules. Therefore, starting with tomorrow's results, I have reintroduced into the model most of the profile attributes that depend on RPI. Since the RPI is only a "blunt instrument" as it is, I'm still not using the team's RPI ranking itself, but it is good (and is used by the committee) for comparing how teams perform against different levels of competition.
Before I make that change to the model, I thought we'd take an early look at some teams that may challenge the committee come Selection Sunday. All predictions are based only on the results from the coaches poll and Sagarin ratings and are through the games of January 25.
Xavier (17-2, 5-0, #7 Sagarin, #4 RPI) and Butler (17-1, 9-0, #16 Sagarin, #8 RPI)
Will these two get some mid-major love from the committee this year? If last year is any indication, probably not. Neither has a signature win (does one beating the other count?) and their numbers can only go down after conference play. I was afraid of the model over-reaching with Butler and Drake last season, and missing Butler by 4 seed lines showed those fears were justified. Xavier seems to be in better shape than Butler this year.
UCLA (15-4, 5-2, #22 Sagarin, #42 RPI)
In his first Bubble Watch column last week, SI.com's Andy Glockner listed UCLA (ranked #11 in the coaches poll at the time) not as a lock but "should be in." This proved greatly insulting to Bruins everywhere ("I don’t know where Glockner is from, but this smells to me like typical East Coast bias" wrote one commenter). Barring collapse, the Bruins should be in the field, but Glockner is correct in saying that they haven't yet had a noteworthy win; on the other hand, they are #24 in the Sagarin ratings and don't have any bad losses either , and Pac-10 play should give them the opportunity to snag a few good wins. CTD has them as a #5 seed as of Sunday.
West Virginia (14-5, 3-3, #13 Sagarin, #21 RPI) and Notre Dame (11-6, 3-4, #44 Sagarin, #77 RPI)
One of these teams is ranked #22 in today's coaches poll and the other received 1 vote. If you were to look only at the numbers in their profiles and not the name on the front of the jerseys, you would think the Mountaineers were the ranked team. And I have no idea why. Notre Dame has not looked impressive lately - aside from the home win over Georgetown there isn't much to speak of. West Virginia has not had a bad loss and beat the same Georgetown team on the road. By 17. The good thing about the Big East this year is that there are still plenty of opportunities for each to strengthen their profile.
North Carolina (16-2, 3-2, #6 Sagarin, #7 RPI)
Despite not winning the championship last season, the Tar Heels came into 2009 with hype reminiscent of the 2007 Florida Gators. Much like the Heels, those Gators had a propensity to coast at times, especially on defense. Lost in the Gators' dominant postseason run was their 2-3 close to the regular season. Many (including Crashing the Dance) thought that those Gators didn't deserve to be on the #1 seed line, much less the #1 overall seed. As I wrote in reviewing the 2007 bracket:
Not much about this one, just that Florida makes absolutely no sense as the #1 seed overall, and the mainstream and non-mainstream media agree. Not to mention the #2 seed given by the Web consensus. I think the committee violated their principle that neither previous season performance nor potential should enter into selection or seeding.
Of course, Florida went on to repeat, so it's a moot point. Which brings me to the 2008-09 Heels. If the Heels lose a few more along the way but win the ACC, I wouldn't be surprised to see them sneak in to the #1 seed line based on the perception of their potential. And I certainly wouldn't be surprised to see them go on a similar run through the tournament. (Oddly enough, with a much improved Duke and Wake Forest, Carolina arguably would not even be in the Greensboro pod as of today.)
Next Monday I will start the 2009 version of Best Week Ever, the weekly look at teams making big moves up (or down) the bracket over the previous week. While you're waiting for this year's version, take a look at some of last year's.