It’s called the Ratings Percentage Index, or RPI for short, and as usual it resulted in a few head-scratching decisions when Selection Sunday came for the NCAA men’s basketball Tournament — most notably the decision to take flawed teams like St. Joseph’s, Villanova, and Kentucky ahead of Arizona State, Ohio State and Virginia Tech.

To see why, we need to understand the big picture. The RPI is a poor measurement of a basketball team’s quality for two reasons. First, it makes no distinction between a two-point loss to a great team (as in Virginia Tech’s ACC Tournament loss to North Carolina this weekend, which likely cost the Hokies a bid) or a blowout defeat (as in Villanova’s 19-point shellacking at the hands of Georgetown in the Big East quarterfinals).

Second, and more importantly, it isn’t even good at what it’s designed to do. The NCAA is loathe to consider victory margin for one very good reason — it encourages teams to run up the score against weak opponents. Since a lot of college basketball games are mismatches talent-wise, this can make for some very poor sportsmanship at the very least.

The way RPI works, though, is to simply take a team’s winning percentage and multiply it by its opponents winning percentage. This effectively means that playing a team like North Carolina will improve your RPI, even if you lose, while playing a team like Oregon State will make it worse, even if you win.

The fact that a team scheduled a game against one of these teams has no bearing on how good they are. But it has everything to do with their RPI, because one knows before the opening tip how it will be impacted — and that’s just wrong.

What’s odd is that the NCAA has a more effective tool at its disposal — USA Today’s Jeff Sagarin uses what he calls his “ELO Chess” model that doesn’t include victory margin, and it’s also supplied to the committee. Why they choose to weigh the RPI more heavily baffles me, as Sagarin’s system is a much more robust way of looking at the type of “we beat them but lost to them,” information the selection committee must distill.

We saw another RPI-related trend in the seedings for the tournament, which is the relative overseeding of teams from “mid-major” conference relative to those from the strongest of conferences, most notably the viciously difficult Pac 10. The latter league oozes talent and may produce as many as half the NBA’s lottery picks this June — eight of its 10 teams were worthy of NCAA consideration and Washington wasn’t far off.

But only six got in, as Arizona State was wrongly shafted, and those that did make it were generally placed too low. Many complained about Arizona making it when they closed with a 3–7 mark in their final 10 games. But it’s inevitable in a conference this strong that somebody will finish the year with such a streak, as college teams tend to play nothing but intra-conference games in the final few weeks.

Meanwhile, Arizona State won half its games in the country’s best league, beat no. 3 seed Xavier by 22 points, and only lost two games outside the league. Yet the Sun Devils suffered because the RPI wrongly understood its near-.500 record and near-.500 opponents’ record to be a sign of middling quality. Actually, it was a sign that everybody in their league was killing each other.

Okay, I’ll get off my soapbox. Let’s give the committee some credit for the things they did correctly, They rightly put the four best teams in the country in the top seeds, and besides the preposterous placement of Vanderbilt as the no. 4 seed in the Midwest they did nothing horribly objectionable on the top five lines. Since those are most of the teams that have a chance of really doing something in this tournament, we should applaud the committee for getting that part mostly right.

Further down the bracket, the puzzles start. Miami somehow earned a no. 7 seed despite finishing behind the snubbed Virginia Tech in the ACC, losing to Tech in the conference tournament and accomplishing nothing overwhelming outside the conference. On the other side, the Big 12 had a similar underseeding problem to the Pac 10, as both Texas A&M and Kansas State were placed far too low in the bracket — each should have been in the top-eight seeds.

For now, the interesting part about the committee’s over-reliance on RPI is that it creates some interesting tournament pairings. By “interesting,” of course, I mean “profitable,” at least for those of you who enter NCAA Tournament pools. Unfortunately, the tournament’s single-elimination format makes random probability a huge factor. Nonetheless, some decent bets can be found.

For starters, you’ve gotta like Marquette’s odds as a no. 6 seed in the South. The Big East’s forgotten power, they have a soft bracket in which they’re paired with arguably the weakest no. 11 seed (Kentucky, who lost leading scorer Patrick Patterson), possibly the weakest no. 3 (Stanford) and clearly the weakest no. 2 (Texas). Having to play the Longhorns in Houston is troubling, but the Golden Eagles are the one low seed that appears to have the best chance of making a deep run.

If they make it to the Elite Eight, they’d be an underdog against no. 1 seed Memphis, but again, they would get the weakest of the no. 1 seeds. Moreover, if Memphis falls nobody else on that side of the bracket looks particularly scary — thought Marquette might not want to face Big East champion Pitt again (the Panthers are lurking at no. 4).

My other sleeper, West Virginia, got placed into an awfully tough neighborhood. After a rough game against unheralded Arizona to open things, they’ll probably have to beat the West’s top three seeds in succession to make a Final Four run, and all three (UCLA, Duke, and Xavier) deserved their seeding.

Besides those two, a couple gimmes loom early on. no. 5 Clemson is a good bet to destroy no. 12 Villanova and no. 4 Vandy en route to the Sweet 16, and no. 10 St. Mary’s should get by an overrated Miami team.

For a Final Four, I’ll take Kansas, UCLA, North Carolina, and my Cinderella pick, Marquette. Hey, I said they got the top seeds right — you wanted me to go with Portland State? It’s just unfortunate that the RPI once again led the selection committee so far astray when it came to seeding the rest of the field.

jhollinger@nysun.com