Thursday, February 6, 2014

Welcome to Bracket Quest

The purpose of this blog is to accurately predict the teams, seeds, and locations for the 68-team field of the NCAA men's basketball tournament.  There are dozens of blogs like this all over the internet, with various and sometimes seemingly bizarre methods, but my goal is to work through the process much as I think the selection committee does, and produce the most accurate bracket I can based on my assumptions about their process.

Here's my process:
1. Determine the 68-team field (32 automatic bids + 36 at-large).  The trickiest part, and most controversial, is the "Last 4 in/First 4 out" part, which I will do much as I do step 2:

2. Rank those 68 teams using direct comparisons of schedule.  Part of the challenge of constructing these brackets is comparing teams with vastly disparate schedules, both in terms of a lack of common opponents as well as overall schedule strength.  How does one compare Kansas to Wichita State this year?  The Shockers are undefeated, yet have just one opponent in the top 25 of the RPI (St. Louis).  The Jayhawks have five losses, but all are to teams that would be in the tournament at the moment, and they have about seven wins against current tournament teams, including Iowa State (twice) and Duke.  Who has the better resume?  I think Kansas, but both might be #1 seeds right now.  This is especially important in determing seeding, which brings us to....

3. Place the 68-team field in four regionals, each with four 4-team pods.  I think a lot of people assume that the committee has an agenda where they try to create a "buzz" with interesting matchups, like pitting Michigan against former coach Tommy Amaker and Harvard or against John Beilein's first D-I school, Canisius.  In my mind, with a tournament in which 67 games are played, some of those schools are going to have unusual connections or previous histories with each other, like a Duke-Kentucky matchup for example.  I don't think the committee necessarily cares about that.  Instead, I think they have a defined set of principles that you can find here.  I'll summarize my methods in my next post.

4.  Adjustments. This is the fun and subjective part.  The selection committee has some discretion about where teams are seeded in order to avoid early intraconference matchups.  Teams can switch regions after they are placed, or even be bumped up or down a seed line.  This would most likely occur if a conference has more than four teams that would all be in one half of their bracket; for example, if the Big Ten this year had Michigan State as a #2, Michigan as a #3, Iowa and Wisconsin both as #6 seeds, and Minnesota as a #11, you would have five teams all in the bottom halves of their brackets, guaranteeing a possible intraconference matchup in the round of 16, which is contrary to the NCAA guidelines.  Instead, one of these teams would get bumped up or down a seed line; I think generally the entire matchup would get moved, so maybe Iowa's 6-11 matchup becomes a 5-12.  This sort of thing also happened with BYU the year they were in one of the play-in games, so that they could be in a pod that didn't play on a Sunday.

For each post (once or twice a week until the week before the selection, and then (hopefully) daily until Selection Sunday), I'll take a look at some of the more interesting seed adjustments, or placements that might seem unusual.

Full disclosure: I am a fan of both Michigan and the Big Ten, so that tends to be where my knowledge base is strongest.

No comments:

Post a Comment