Wednesday, March 12, 2014

From 351 to 68: We Begin Again

Welcome to my annual trip into college basketball coverage.

I started trying to suss out who would make the tournament all the way back in college, when there was a lot less data and a lot more guesswork. I've been a huge college basketball fan since grade school, when the St. John's team led by Chris Mullin became an obsession. These days, about the only fandom that still really gets my blood boiling is USC basketball, which is exactly as sad a notion as it sounds.

A few years back, after launching Basketball Prospectus, I ported the work I'd been doing privately for years to the public, joining approximately 7.2 million people in the act. I don't do mock brackets and I don't seed the field. All I do is try and figure out who will make the cut, and why. If you're looking to know whether North Carolina will stay close to home or if Oregon will be wearing white, you want a different Joe. I do this because I really enjoy it, and I publish it because some vocal subset of people who read it ask for it each year. I don't pretend for a second I'm the best or most thorough at it.

One important note before I throw a lot of names at you: every year, it seems, guys like me complain about the soft bubble, where there are too many teams under consideration who haven't done anything. If anyone uses that term this year, ignore them. There are plenty of teams under consideration who would, in other years, be in great shape for a bid. There are few teams in the mix for an at-large bid as weak as some of the teams who made it right down to the end of the discussion in previous years. There are teams with flawed profiles, to be sure, but deep into the pool of teams under discussion are ones with good wins, good road wins and good metrics. This is not, relative to other recent years, a weak bubble.

There are 68 slots in this year's tournament. 32 will be awarded to the champions of their respective leagues, with 31 of those being determined by the winner of the conference tournament. Bless you, Ancient Eight. Twelve of those have already been determined.

In (12): Mercer (Atlantic Sun), Coastal Carolina (Big South), Delaware (Colonial), Wisconsin-Milwaukee (Horizon), Harvard (Ivy), Manhattan (Metro Atlantic), Wichita State (Missouri Valley), Mount St. Mary's (Northeast), Eastern Kentucky (Ohio Valley), Wofford (Southern), North Dakota State (Summit League), Gonzaga (West Coast)

Note: of those 12, just five won their league's regular-season crown. I'm not terribly opposed to conferences who send their tournament champion to the dance, but for those who are, and who make strong arguments against it, this year is an excellent case study. Davidson, Wisconsin-Green Bay and Vermont went 44-4 in conference, and none of them even reached their conference final.

On the other hand, as someone who roots for NIT chaos, I like that seven bids in that event are already committed, as regular-season league champs are guaranteed NIT bids if they don't reach the NCAA tournament.

There are 36 at-large slots this year, down one due to the Big East schism. Oddly, due to the weakness of conferences like the West Coast and Missouri Valley, none of those bids have yet been claimed -- all at-large candidates have either finished their seasons on the bubble (BYU, Green Bay) or are still playing with a chance to win an automatic bid. Per my analysis, 33 teams are locks to make the tournament:

Locks (33): Duke, Syracuse, Virginia, North Carolina; Cincinnati, Louisville, Connecticut, Memphis, Southern Methodist; Massachusetts, Saint Louis, Virginia Commonwealth, George Washington; Villanova, Creighton; Wisconsin, Michigan, Michigan State, Ohio State, Iowa; Kansas, Iowa State, Oklahoma, Texas, Kansas State; San Diego State, New Mexico; Arizona, UCLA, Oregon; Florida, Kentucky, Tennessee.

The closest calls here were George Washington, Kansas State and Tennessee, and we'll get into all of them over the next few days. I suspect the last of those causes the most raised eyebrows. I am, to some extent, projecting the committee to be using more tools than just the RPI. For all the data we do have, we're really just guessing as to how the committee members will integrate modern rating systems into their work.

Those 33 teams come from nine conferences, and will use somewhere between 24 and 33 at-large bids, leaving three to 12 slots for bubble teams. Here's my spreadsheet bubble, listed in RPI order. Keep in mind two things: I am very conservative about moving teams from the bubble to the at-large board, and I try to include everyone at the start who might play their way into the field this week.

Bubble (29): Colorado, Southern Mississippi, Baylor, Brigham Young, Arizona State, Dayton, Nebraska, Oklahoma State, St. Joseph's, Pittsburgh, Stanford, Xavier, Minnesota, Missouri, California, Georgetown, Providence, Florida State,, Arkansas, Middle Tennessee, St. John's, Wisconsin-Green Bay, Illinois, Louisiana Tech, Stephen F. Austin, Louisiana State, Georgia, Clemson, Utah

The most interesting case on this board in Pittsburgh. Despite "feeling" like a lock, they have a shockingly empty resumé, having beaten no teams in the tournament field and just one team likely to make the tournament (Stanford, back on November 26). It is possible that they could end up having beaten zero tournament teams, pending what a few teams do this week. You wouldn't think you could go 23-8 as an ACC team with this soft a slate, but they went 1-6 against the RPI top 50 and played just five games total against the four good teams in the league. Pitt plays the Notre Dame/Wake Forest winner tomorrow, and cannot afford to lose that game.

I don't think all the Conference USA teams are at-large contenders, but I do think a tournament final between two of them would open the door to an at-large bid for the loser if the bubble fell apart.

We'll go deeper on all these teams as the week goes on. I think, right now, the Big 12 teams, the Pac-12 teams and BYU are all in pretty good shape.

There is some bubble action today in some of the bigger conferences. Colorado, Baylor, Georgetown and Stanford are playing don't-you-dare-lose games against the dregs of their leagues. Utah has a slightly tougher assignment, Washington, but can't afford a loss. Oklahoma State has a competitive game against Texas Tech; they wouldn't fall off the bubble with a loss or move into the field with a win, but with a 5-10 mark against the RPI top 100, they could use another game against a good team.

Also, Oregon plays Oregon State, but Oregon is in no matter what they do in Las Vegas.