Who's In and Who's Out?

By popular demand (okay, three or four commenters requested it), I'm going to go with one last update as to how I think matters stand on the bubble.

First off, the teams that are obviously in as at-larges (plus auto-bid winners in those conferences). Bold indicates an auto-bid winner, italic indicates that they're still alive in the conference tournament. Number of at-large bids is given assuming that, if possible, one of the teams that would clearly be in will take the spot (* indicates that another team is still alive to take it).

ACC (1): Duke, UNC
BIg East (9): UConn, Louisville, Villanova, Notre Dame, Pitt, Georgetown, Syracuse, St. John's, Cincinnati, West Virginia
Big Ten (2*): Ohio State, Wisconsin, Purdue
Big XII (4): Kansas, Texas, Kansas State, Texas A&M, Missouri
Pac-10 (2): Washington, Arizona, UCLA
SEC (3): Florida, Kentucky, Tennessee, Vanderbilt
MWC (2): San Diego State, BYU, UNLV
A-10 (2*): Richmond, Xavier, Temple
Colonial (1): Old Dominion, George Mason

That accounts for 26 bids (plus a possibility of two more if Penn State and Dayton win their conference title games). Contenders for the remaining 11 after the jump (sorted by conference). Data courtesy of Warren Nolan.

For "best wins", I'm taking top three RPI wins (plus the top non-conference win if that is not in the top three or a fourth if it is), and bottom three RPI losses for "worst losses".

There are 19 teams listed above for those 10-11 spots (depending on the Dayton-Richmond game). On a first pass:

• We have a few great wins, a top strength of schedule, and nothing horrible on the low end. While we certainly didn't make it easy on ourselves, we're in.
• Penn State's SOS beats ours out mainly by 1) getting Iowa only once instead of Ohio State, and 2) not playing anyone below #230 in non-conference (while they didn't schedule a murderer's row like we did, they have a lot of middling cupcakes and a decent number of semi-competent teams in the 80-120 range instead of the pond scum they played in past years). This is a case where the numbers (at least, the Stone Age RPI) lie. Still, their double over Wisconsin, good finish, and non-conference devoid of RPI anchors look like enough.
• While Michigan hasn't beaten a true top team, they've absolutely dominated the mid-level/bubble regime and don't have much that's truly embarrassing (the loss to Minnesota looks worse than it is since it was before Joseph and Nolan both found themselves out for the year). This year, that's easily enough (especially since of that 0-8, four of those games are against the RPI top 2).
• Marquette looks like a good bet with 5 top 25 wins. The RPI is a bit low due to the "anchor" phenomenon (four sub-300 RPI non-conference games, plus #280 and #250), but as much as the committee loves to make examples of such teams, wins over four different top 25 teams will see them through. They could end up in the First Four as punishment for it, but I just can't see any way they get left out entirely. (That and the committee has an even better target to whack, as we'll see later.)

That gives us four teams in, 6-7 spots remaining for 15 teams. (Even if Penn State wins tomorrow, since we're fairly sure they're in at this point, that would affect seeding only. Personally, I don't think Penn State's a First Four team as things stand, but if they would be, winning the Ohio State game would remove them from that automatically by virtue of not being an at-large team.) Let's try to trim the other end:

• Sorry, Alabama. An RPI of 80 is not going to get it done. Best match to the profile is Kentucky 2008 (abominable non-conference, good run in conference) but Kentucky did that against a much tougher non-conference schedule, in the better division of the conference, in a stronger year for the SEC. And they had only one bad loss, not four.
• UTEP's profile would be very, very borderline even if I weren't convinced the RPI numbers for CUSA were crazy. If the committee shares that opinion, they're long gone. (And even if not, they're probably only at the edge of the discussion.)
• It takes a ridiculous number of good wins to overcome 6 bad losses (including three sub-200 abominations). USC has a decent number, but not a ridiculous number. Goodbye.
• Colorado, meet Penn State 2009. This is what happens when you schedule a bottom-25 non-conference. Six sub-300 games and another at #294? Three wins over Kansas State won't make up for that. There's always one team who gets made an example of for a terrible non-conference schedule, and you're it.

Now we're left with 11 teams for 6-7 spots: the four ACC bubble teams, Georgia, Illinois, VCU, St. Mary's, Missouri State, Harvard, and UAB.

First, let's try to get the order of the ACC teams. Virginia Tech and Florida State have the marquee wins the other two lack; Virginia Tech has a second solid win (PSU) and head-to-head wins but more bad losses and a worse overall record. I wouldn't argue too much with someone who put VT tops in that group, but I'll go with Florida State as the top and Virginia Tech second. On overall body of work, Boston College might have an argument over Clemson, but with Clemson blowing them out in the ACC tournament (and going 4-1 against the three ACC bubble teams), I think Clemson's likely to get the nod.

Georgia has a resume with almost nothing of note in either direction - one great win, a couple decent wins, two mediocre losses (Alabama), and no disastrous ones. I think the factor that's really going to kill them, if they're left out, is that they had two bubble battles in their last three games and lost both, blowing a huge lead in one of them. Winning one of those would have put them in good shape. Winning both would have made them a lock. Instead, with a chance to play their way in, they choked. Thinking they're ahead of Boston College but behind Clemson.

Illinois would look like a fairly solid contender if not for a late slump. The win over North Carolina came over the early-season version, which may make it discounted a little bit (North Carolina slumped a bit early before a personnel change caught them on fire). Still, a bunch of top 50 wins and few bad losses (although one of them's a real doozy) ... unless the committee puts a lot of weight on recent performance I can't see any of the ACC teams ahead of them.

VCU would be an absolute shoo-in if not for an ugly losing streak to end the season (lost their last four Colonial games, interrupted by a BracketBusters win over Wichita State). And part of that is scheduling, as they played four of the top six in those games. With five decent (RPI 60 or better) non-conference games and two wins in there, I feel like they're on par with Virginia Tech or slightly behind.

St. Mary's non-conference is about on par with VCU's (slightly better top-end, but only four top 60) but they didn't get the advantage of playing Old Dominion and George Mason five times in conference. I have them just ahead of Boston College but behind Clemson.

Missouri State: winning the Valley usually counts for quite a bit. But the Valley isn't usually this bad. The conference's 0-fer on BracketBusters may be the final nail in their candidacy. Looks to me like they're behind all four of the ACC group.

Harvard: Yes, Harvard is somehow on the bubble. There's not a whole lot of depth to their resume, but a reasonably aggressive non-conference schedule which includes two bubble-level wins and only one real bad loss (to hated rival Yale). Majors get a lot more opportunities to impress; Harvard didn't get a ton but made the most of the ones they had. I think they're clearly ahead of Boston College, but probably not Clemson.

UAB: This candidacy rests entirely on how much faith you put in CUSA's RPI numbers. Looks to me like it's mostly built on Central Florida's crazy run in non-conference then fall-apart in conference. Even if you believe the numbers, there's not much heft to this profile. Chalk up another "ahead of BC but behind Clemson" vote.

Now we just need to sort out all the teams between Clemson and BC. St. Mary's goes ahead of Georgia on the basis of Georgia failing to take advantage of late opportunities; one abominable loss is not a disqualifier. UAB also goes behind St. Mary's, but I think they're ahead of Georgia, whose conference record is built on 7 wins against the awful half of the SEC (South Carolina and the non-Alabama part of the West). Finally, Harvard looks like the best of that bunch, just by virtue of taking advantage of the few opportunities they had. Profile looks a lot like St. Mary's but without the bad losses.

So that looks like an order of:

Last three byes
Illinois
Florida State
Virginia Tech (unless Dayton steals a bid, in which case they're in the First Four)

In the First Four
VCU
Clemson
Harvard

Praying for a Richmond victory (last team in, assuming Dayton does not steal a bid)
St. Mary's

NIT-bound
UAB
Georgia
Boston College
Missouri State

It will be interesting to see how this stacks up when the official bracket comes out. The teams I would be least surprised to be wrong about are Georgia, USC, Colorado (who I've predicted as out) and VCU, Harvard, Illinois (in).