The nice thing about the NCAA Tournament selection process is that it is not a secret. You can review the guidelines for the process here. The simplest part of that document is the section that refers to the selection of at-large teams:
The committee shall select the 34 best teams to fill the at-large berths, regardless of conference affiliation.
Even with the last four words being superfluous, it's a very simple statement of the committee's task. Most of the speculation from the media in the next three weeks will involve using data from the RPI to tell us who should be in the tournament. However, it's worth remembering that ultimately, committee members must select which teams they feel are best. That's the only thing that matters when it comes to filling out the field.
There isn't a formula in existence that can do that for them. If there was, or if there were some sort of objective blueprint for the entire process, there would be no need for a committee. One person could simply look at a checklist and pick the best teams based on pre-established criteria. However, there is room for subjectivity--there has to be. Among even the most knowledgeable group of basketball fans, there will be differing opinions on which two or three teams should be the last in the field.
Of course, objective data has a place in the process, and I would hope for it to have a prominent place. It appears it does, based on this provision in the section labeled "General Principles for Selection, Seeding and Bracketing":
Among the resources available to the committee are complete box scores, game summaries and notes, various computer rankings, head-to-head results, chronological results, Division I results, non-conference results, home and away results, results in the last 12 games, rankings, polls and the NABC regional advisory committee rankings.
That's a lot of data, and none of it is useful on its own--every bit of it has context and is open to interpretation. Contrary to what will be implied over the next three weeks, a team's worth is not defined by its RPI, or its SOS (which is merely a subset of the overall RPI calculation), or even its record against the top 25 of the RPI. It all comes down to whether a committee member feels that a particular team is among the best 34. That's a problem that the RPI can't solve on its own.
This brings us to the case of Davidson. Based on what everyone has done to date, I think Davidson would be one of the 34 best at-large teams. What this really means is that they're one of the 45-50 best teams in the country, since automatic bids will be occupied by some of the teams in that group. I can't use the RPI to prove it, however.
The Wildcats stand at 21-6 after Friday's win at Winthrop. Among their losses are games against UCLA, North Carolina and Duke, all of whom are expected to find themselves on the top two seed lines. As expected, Davidson went 0-3 against those teams. If the committee is looking for the best teams, however, Davidson's record in those three games should take a back seat to their performance in them.
The UNC game wasn't decided until the final minute, and while the UCLA game was not as close down the stretch, Davidson managed to put UCLA as far down (18 points) as they've been all season during the game. The Duke game wasn't as suspenseful, with the Blue Devils going into clock-milking mode after taking a 12-point lead with five minutes to go. Where Davidson's case loses steam is its three other non-conference challenges, losses to Western Michigan, Charlotte and NC State. Each of those teams are respectable in the sense that they are better than Davidson's Southern Conference brethren, and each game was on the road, but each is also inferior to an at-large quality team. Overall, the Wildcats went 1-6 against D-I non-conference teams in November and December, recording their only win in a home game against D-I newcomer North Carolina Central.
At some point, a team needs wins to prove their worth. Davidson has wins--they may well close the regular season with 21 consecutive against SoCon competition plus the road win against Winthrop. There's not a quality win in the bunch, but collectively, that group of wins is an achievement worth considering. We can't know how bubble teams from bigger conferences would fare with such a slate, but UCLA, Tennessee and Kansas have shown that being much better than the rest of your conference does not mean you can roll through your conference schedule, even when it's just 16 or 18 games. There's enough baggage in each team's portfolio to suggest that what Davidson has done shouldn't be ignored.
I often get frustrated that people think 27-30 games are enough of a sample to distinguish among the last three or four at-large teams. In fact, we know it's not, because every season there are a couple teams that cry bloody murder after the bids are revealed. If a full college season were a big enough sample to judge teams by, everybody's guess at the bracket would agree. So imagine the issues in judging a team on what is essentially a six-game sample. The fact that the sample includes all losses will no doubt disqualify Davidson from consideration for an at-large should they need it. However, were I somehow serving on the committee, I could not be so sure the Wildcats weren't deserving. Based on how they have played, and based on non-RPI data, they would be one of the best teams left out of the field.
The good news for Davidson is that the RPI is less relevant than it's ever been. Over the past two seasons, five teams ranked in the top 30 of the RPI failed to make the NCAA Tournament. In the ten seasons prior to that, no member of the top 30 of the RPI was left out. The other piece of good news for Davidson fans is that if they really are as good as I believe, they won't need the committee to tell them so. They'll win the remainder of their games against inferior Southern Conference competition and get an automatic bid.
Ken Pomeroy is an author of Basketball Prospectus.
You can contact Ken by clicking here or click here to see Ken's other articles.