From the Rafters: Bracketology Edition

It's March 1st, which means only one thing, the Madness is upon us. I couldn't think of a better way to celebrate than by putting together a bracket of my own...From the Rafters

I won't bore you with my methodology for ranking the teams, but this bracket is an objective compilation of data points which are a blend of those I find most important and the ones the committee relies upon heavily. Of most importance is who you beat, as top wins are weighted heavily. Conference affiliation, conference record, and conference standing have no bearing whatsoever, and while I use the RPI as one data point for giving weight to resumes, Ken Pomeroy's ranks are preferred. Of note, in addition to factoring in records against RPI top 50 and top 100 teams, I factored in a team's records in what Ken Pomeroy deems "A" and "B" games. "A" games are those against Top 50 teams after you account for the location of the game (i.e. he uses data to suggest that beating the 106th ranked team on the road is akin to beating the 26th ranked team at home). Likewise "B" games are equivalent to top 100 games once adjusted for location.

Once the teams were ranked, I placed them into regions with every attempt at protecting geography for the top seeds and avoiding conference matchups as long as possible. The teams from one-bid leagues were filled in based on their KenPom rank. Here are the results for games through Monday, February 29th

Midwest Region

Midwest Notes: Kansas is far and away the best team using this method. With 10 KenPom "A" wins (Villanova and Oklahoma have nine) and 12 RPI top 50 wins (next best, Oregon, has 9) it's clear why they are on track to get the tournament's number one overall seed.

West Region

West Notes: Despite it's recent form, Oklahoma's overall resume is still very good. Both KenPom and the RPI like the Big 12 meaning just about every game they play is an "A" game therefore wins are huge, and losses aren't killers. That being said, you could easily flip Oklahoma and Michigan State based on how they are playing. In the composite rank, Michigan State is closer to Oklahoma as the last #1 seed than it is to Miami who is the next #2 seed. Florida State is a surprising inclusion in the field, but they have a decent Ken Pom rank, and the majority of their losses are to quality competition. Admittedly, this method favors teams who play a lot of their games against top 50 level competition.

East Region

East Notes: Villanova is an interesting case as a number one. Their "A" game wins and RPI Top 50 wins are among the highest in the country, but once you account for top 25 wins (in either KenPom or RPI), other teams leapfrog the Wildcats. Villanova has only one win over a top 25 team in either ranking. For comparison, there are 36 teams (KenPom) and 32 teams (RPI) in these rankings that have at least two. This region also highlights big movers. Ten days ago, Dayton was a four seed and Wisconsin wasn't even in the field. This year has been dynamic, but those two have just been ridiculous.

South Region

South notes: Virginia comes in as the third overall number one seed, however they are mere decimal points away from Villanova for second and Oklahoma for fourth. They fall behind both those schools in resume rankings though their body of work really benefits once you consider top 25 wins. This region features a couple teams who benefit from this model. The first is Utah whose resume suggests they are much closer to a two seed, even though KenPom ranks them 24th. Then there is Texas who actually leads the country in KenPom top 25 wins with six. On the flip side, there is the curious case of Wichita State who KenPom ranks as the eighth best team in the country, despite a 2-3 record in "A" games and only having one win over a KenPom top 25 team.

Last Four in: Syracuse, Florida, Florida State, Alabama

First Four out: Virginia Commonwealth, Gonzaga, Temple, Washington

Conference Breakdown

Overall Observations:

- The 8-10 teams at the top of the bracket are fairly set, but the amount of volatility among those teams will make for an exciting couple of weeks. Take for example the ACC where UVA gets the nod as of today for a #1 Seed, but come Selection Sunday you could easily see UNC or Miami replacing (or even joining) the Hoos on the one line.

- The ACC having 8 teams is a huge surprise. I doubt that is the case in two weeks as FSU, Syracuse, and Pittsburgh are squarely on the bubble. For those teams, the last week of the regular season, and the tournament will be a make or break.

- Speaking of the bubble, it gets ugly really quickly. This year more than any, if a team doesn't make the NCAA Tournament, they just aren't very good.

What do you think? Who is left out? Who is ranked too high? Who is ranked too low? Who is the team you don't want to see in Virginia's region (besides Michigan State)? Leave your thoughts in the comments and come back next week when all of this is turned on its head.