The Crashing the Dance model of the at-large selection part of the committee's process gives each team a confidence rating between 0% to 100%. This is a measure of how well the CTD oracle thinks the team's profile matches that of an at-large selection based on comparisons to the at-large selections from the 2000-2007 seasons. 100% confidence means we have no doubt that team will be an at-large selection and 0% confidence means NIT.

Of course that's the easy part. A ham sandwich could identify the teams with 100% confidence. It's those pesky bubble teams that give everyone trouble. Some parts of their profiles look like Sweet 16 material while other parts look like NJIT. (Well, not that bad.) That's what makes it difficult.

As outside observers, we obviously have no way to know which factors the committee values more, and whether they even apply their philosophies consistently. (Usually, but not always, is my guess.) What we try to do is look at teams in the past with similar profiles and see how the committee treated them. That's where the confidence value comes from.

We rank each team by the confidence value and take the first 34 non-conference-champs as the at-large selections. The Type column in the selection report shows whether each team is an automatic qualifier (Q) or at-large selection (L). Hovering the mouse over the Type column will show the confidence value for the team.

In a perfect world, we'd have the 34 at-large selections with 100% confidence, and everyone else with 0%. What ends up happening is a bunch (20ish) of them with 100% followed by a transition down to 0%, crossing the bubble somewhere in between 0% and 100%.

Last year, we crossed the bubble between the 75% and 45% confidence levels. This means that the teams on the good side of our bubble made strong cases (even if all did not make it in), and the our bubble was approximately a 50-50 shot.

With less than two weeks to go to Selection Sunday, that's not the case this year. Through Tuesday's games, the bubble is split at 3% and 0.5%. In other words, there will be some teams that make the field with weak arguments for being there.

Let's take a look at some of these bubble teams with arguments for and against (all results through Tuesday's games).

TeamSparklineConfidenceProsCons
Arizona State 82.5%
#28 of 34 at-large		 4 W over RPI Top 25
5 W over RPI Top 50		 9-10 vs. RPI Top 200
L 8 of 12
New Mexico 43.7%
#29 of 34 at-large		 14-7 vs. RPI Top 200 1-4 vs. RPI Top 50
3-5 vs. RPI Top 100
Davidson 10.8%
#32* of 34 at-large		 W 19 straight
Top 30 coaches poll		 #125 non-conf RPI
0-5 vs. RPI Top 100
Ohio State 0.4%
3rd team out		 #29 non-conf RPI
#39 Sagarin rank		 1-7 vs. RPI Top 50
L 9 of 15
Syracuse 0.1%
7th team out		 #33 non-conf RPI
6 W over RPI Top 100		 #92 conference RPI
1-8 vs. RPI Top 50
Kentucky 0.1%
8th team out		 #5 conf RPI
W over Tenn, Vandy		 #208 non-conf RPI
10-10 vs. RPI Top 200

* Note: Davidson is their current conference leader; they would be the 32nd at-large selection otherwise.

Of course this may all change as conference tourneys commence. The good thing (for those with no vested interest, anyway) with a weak, fluid bubble is that, well, it's fluid. And that means exciting. One of those teams (cf. Arkansas 2007) with a weak argument could go on a nice run, knock off a few highly ranked teams, and boost that confidence. Not to mention their own confidence.