Selection Sunday is over, and as everyone except Jerry Palm anticipated, Northwestern didn't make it. This was a perfectly reasonable decision by the committee, as Northwestern had plenty of opportunities to earn good wins and didn't take advantage of them.
However, the committee confounded many observers by selecting Iona for an at-large bid. The Gaels failed to record even one RPI top 50 win, and suffered two bad losses to opponents ranked below 200th in the RPI. After they lost to Fairfield in the MAAC semifinals, the consensus was they had little chance at a bid, even with such a weak bubble. But last night, their name was called, leaving other bubble teams like Drexel shaking their heads. And when Drexel's resume is compared side by side with Iona's, it's nearly impossible to make an argument for Iona.
Drexel
vs RPI 1 through 50: 1-1 (Both against Virginia Commonwealth)
vs RPI 51 through 100: 5-2
vs RPI 101 through 200: 6-3
vs RPI 201+: 15-0
Iona
vs RPI 1 through 50: 0-2 (losses to Purdue and Marshall)
vs RPI 51 through 100: 5-1
vs RPI 101 through 200: 8-2
vs RPI 200+: 12-2
So Drexel has more top 50 wins, more top 100 wins and fewer bad losses. Drexel also played in a more difficult conference according to the conference RPI rankings: the CAA is ranked 15th and the MAAC is ranked 18th. Drexel's resume so far is clearly better, but then you consider the circumstances of their losses:
Four of their six losses came in a three-week span from Nov. 18 through Dec. 3, a stretch where starting guard Chris Fouch was injured for three of the games and limited in the other two.
As far as I know, none of Iona's losses can be explained by injuries. One point often brought up in Iona's favor was that they won an impressive 15 games away from home on the season. That's certainly a point in their favor, but only one of those wins came against a top 100 team (at RPI #94 Denver). Drexel, meanwhile, won 14 games away from home and won at RPI #85 Cleveland State. So record away from home doesn't explain this difference either.
You probably figured this out already from the title, but there is one large point in Iona's favor here: non-conference strength of schedule (SOS). Iona played the 43rd toughest non-conference schedule while Drexel played just the 223rd toughest non-conference schedule. The committee has shown in the past that it likes to reward teams for scheduling tough opponents, and clearly that pattern continued here, because there's no other logical explanation for Iona getting in over Drexel. Selection committee chair Jeff Hathaway even emphasized Iona's challenging non-conference schedule when questioned on CBS last night.
There are several issues with the NCAA using this criteria. First off, while top programs may have full control over whom they can schedule in non-conference play, mid-majors often struggle to find good teams willing to play them. Drexel, despite being located in Philadelphia, is not included in the prestigious Big 5, and thus not given the opportunity to play annual non-conference games against perennial powers Temple and Villanova. It's not like Drexel is ducking the Big 5; in fact it's the other way around.
A second issue is that the committee's goal is simple: select the 37 best at-large teams. Thus, they should consider the entire body of work. Why does it matter if a team played a weak non-conference schedule and then a strong conference schedule as opposed to a strong non-conference schedule and a weak conference schedule? The overall SOS is the only thing that should matter. If a coach knows he's facing a brutally tough conference slate starting in January, there's nothing wrong with him playing a few weak teams early in the year to build his team's confidence and give his reserves valuable game experience, as long as that team proves it can beat good teams during conference play.
But even if the committee only considers overall SOS, there's still a major flaw in the system: the way SOS is calculated. Iona's overall SOS was 143rd, while Drexel's was 213th. This disparity makes little sense when looking at the resumes side by side: both teams played two top 50 opponents, and Drexel played nine top 50 opponents to Iona's eight. And each team played eighteen top 200 opponents.
The reason for the disparity can be seen when analyzing the dregs of each schedule. Drexel played eleven games against teams rated 250th or worse in the RPI, while Iona played just six games against such teams (and lost one of them, to Hofstra). The problem is that in the SOS formula, there is a huge difference between playing the 200th best team in the country instead of the 345th best. Even though both schools faced an equal number of competent opponents, and Drexel faced more top 100 opponents, Iona has a large overall edge because the terrible, easily beatable teams Drexel played were slightly worse than the terrible, easily beatable teams Iona played.
In reality, an NCAA tournament caliber team will be a heavy favorite over any team ranked 200th or worse in the country, and shouldn't get any extra credit for beating team #200 instead of team #345. Even the committee agrees with this, as they give teams equal credit (that is, no credit) for beating opponents rated worse than 100th in the country. Yet their SOS formula thinks that for an NCAA caliber team, the difference between playing the #1 team vs the #145 team is the same as the difference between playing the #200 team and the #345 team. It defies common sense, but it's the system we currently have.
If the NCAA is going to insist on using non-conference SOS to select the field, they should at least use a version of SOS that makes sense, and assign equal weight to all games against #200 and below. Otherwise, we end up with nonsense like Iona over Drexel.