Friday, September 16, 2005


Why do bracketology?

Viewing teams as possible seeds instead of relying strictly upon poll rankings offers many advantages.

  1. It is an active exercise. Evaluating RPI, overall record, road record, top 50 wins, sub-100 losses, recent record, and other hard data allows bracketologists to gain a better understanding of what a team has accomplished. Looking at polls is a passive exercise. Bracketology uses the actual NCAA selection criteria to give fans an idea of what the bracket might look like week to week.
  2. Engaging in bracketology from Day One gives fans a better idea of a team’s journey through the season. For example, take the 2004-05 Iowa Hawkeyes. I had them at a 3 seed at the end of the non-conference schedule. On Jan 4, they were 12-1 and had beaten Louisville, Texas, Iowa State, and Texas Tech. They proceeded to open 4-9 in Big Ten play and managed to fall completely out of the bracket. Then, they recovered to win their final three regular season games and two games in the Big Ten Tournament before falling to Wisconsin in the semis and earning a bid in the NCAA Tourney. They were absent from the polls for most of their roller coaster journey.
  3. Most fans would rather know where they stand in regards to the NCAA picture than where a panel of coaches (or, more accurately, SID’s or assistant coaches) rank their team in a poll. I know that the Final AP poll is a better predictor of seed than the RPI, but it is not a better predictor than serious bracketologists.
  4. Finally, looking at the seeding motivates one to look at conferences other than the power conferences. Seed lines 13 through 16 come almost exclusively from conferences other than the power conferences. These teams are NOT afterthoughts; in fact, they make up more than one quarter of the national championship bracket! Furthermore, a few of them end a handful of contenders’ (or pretenders’) hopes of the Final Four on the first weekend (Bucknell, Vermont, and UW-Milwaukee spring to mind from the 2005 tourney). Bracketology gives the fan a deeper, more complete understanding of what is happening during the season in all 31 conferences.

Why do bracket projections so early? Isn’t is pointess?
Why have polls so early? I rest my case. It is sports. It is conjecture. It is just for fun. And, it lets the fan track teams’ adventures from November 1 until Selection Sunday. In fact, until about the mid-point of the conference season, some “Bracket Board Bias” does exist in my bracket projections. But, the more games that are played, the more a bracketologist can rely strictly on the teams’ accomplishments. Only on Selection Sunday will it TRULY matter. I think of early season bracketology as practicing for my final bracket projection.

The RPI: It’s all about purpose
The new RPI is no screwier than the old one. It's ALWAYS been screwy...if you look at the RPI as a measurement of who is the best team.

The reason for this stems from the not-so-widely-known true purpose of RPI. It is NOT a tool to rank the best the teams in order like the AP poll. It is a tool to show how a team performed against its schedule. There is a BIG difference.

Louisville Courier-Journal sportswriter Rick Bozich wrote an anti-RPI piece pointing to the fact that Kentucky trailed Gonzaga in the RPI. In his eyes, UK was clearly a better team (I happened to disagree, but I digress). Here is part of my response to Mr. Bozich:

“In reference to Kentucky's relatively poor RPI rating (11, which is excellent) compared to their lofty AP poll position (#3), you wrote that, ‘Yet the Wildcats sit 11th in the latest RPI behind Gonzaga (losses to Missouri, St. Mary's and San Francisco) and Arizona (ugly losses to Virginia and Washington State).’ That stems from the fact that UK played William and Mary, Campbell, Morehead State, and Georgia State in the non-conference. They did play UNC, Kansas and Louisville, but went 1-2 in those games. Gonzaga played a few patsies, too, but also played Illinois, Georgia Tech, Washington, and Oklahoma State in the non-conference, going 3-1 in those games.

Regardless of who is better, Gonzaga has performed as well as (or better than) Kentucky against their given schedule. That is what the RPI intends to measure and does so effectively. If Kentucky or Charlotte wants a higher RPI, they might want to schedule a little tougher in the non-conference than they did this year. In Charlotte's case, they might want to avoid losing to East Carolina and at home to Rutgers, while they are at it.

A lot of the wailing about the new RPI comes from a lack of understanding regarding it's purpose.”

I could not have said it better myself…only because I wrote it.