Tuesday, October 18, 2005

So, you want to be a bracketologist?

Does Joe Lunardi drive you nuts with his projections? Can't understand the ramblings of the selection committee chair when he starts defending teams with terms like SOS, OOWP, and S-Curve? Bewildered by the seeding process? I have good news. You can answer a lot of these questions (or at least gain some context) by following TBB's Five Steps to Being a Bracketologist. (Enter sales pitch for three easy payments of only $39.99 each)

1. Forget the polls. The final AP poll is a fair predictor (better than the RPI, actually) of what seeds that the top 25 teams will get. Beyond that, the polls are mostly just a weekly list of what big conference teams are winning games. Conversely one can apply the NCAA selection criteria rules at any point during the season. Granted, they may not be any more accurate than the polls in December, but at least every team is evaluated under the same criteria.

2. Join CollegeRPI.com. Right now. Jerry Palm maintains the best RPI and bracketology info site anywhere. The RPI is updated daily during the season and the site has everything that you could ever want to know about past NCAA tourney results, historical team RPI data, past performances, and bracketology. It is well worth the $20.

3. Read and learn the NCAA Selection Criteria. Reading this document is about as entertaining as watching toenails grow for a lot folks, but for bracketologists it is scriptural. It will take awhile to get through, but readers will come away with a clearer understanding of how teams are selected, seeded, and assigned to their respective pods.

4. Consult TBB's "Bracketology/RPI section." It contains an explanation of the true purpose of the RPI, a philosophy of bracketology, and the "why's" of projecting brackets.

5. Now comes the fun part: watch A LOT of basketball. This is easier for some than others, but I teach, so I am off from Dec 10-Jan 23 this year. I travel to WKU games and set up a game watching TV schedule each week. The more hoops that a bracketologist watches, the more informed he/she is about what teams have improved and regressed. After combining all of the statistical data with what your eyes are seeing in games, it becomes much easier to split hairs between those closely bunched Bubble Boys. Of course, the data usually bears out who gets in and who does not, but momentum is a factor for the selection committee. An example, you ask? Here's one from last season.

UAB: 21-10 overall, 10-6 in CUSA, RPI #49
DePaul: 19-10 overall, 10-6 in CUSA, RPI #53

Both teams won one game in the CUSA tourney. Both teams were 5-5 in their last 10 games. Both teams had four losses to sub-100 RPI teams. DePaul actually had two Top 50 RPI wins while UAB had zero. UAB did have a better road record (10-7 to DePaul's 5-8), but that was the only glaring statistical advantage. That may have been offset by DePaul's good wins over Old Dominion and Cincinnati. This would seem to be a virtual dead heat.

The fact of the matter is, every serious bracketologist knew that DePaul was definitely out and that UAB was a classic bubble team (who wound up getting into the Dance). Why? UAB beat DePaul head to head in the CUSA Tourney. Although head-to-head matchups are technically not a part of the selection criteria, it had been proven on the court that UAB was a better team. The raw data says that teams were even, but from actually watching games late in the season, it was clear that UAB was a better team. Contrary to popular belief, common sense DOES play role in the selection process from time to time (although I still believe that Buffalo should have been in over UAB, but I digress).

So take this a charge to watch LOTS of hoops and go to lots of games this season. And the next time that you are watching paint dry or catch yourself mesmerized by a clock pendulum, dig out the NCAA selection criteria. It will come in handy on Selection Sunday.