Sunday, February 11, 2018


Tom Keegan: RPI remains hidden impediment to NCAA tournament fairness

Kansas players crowd around head coach as they are presented with the Midwest Regional trophy as they celebrate the Jayhawks' 80-67 win over North Carolina to advance to the Final Four on Sunday,  March 25, 2012 at the Edward Jones Dome in St. Louis.

Kansas players crowd around head coach as they are presented with the Midwest Regional trophy as they celebrate the Jayhawks' 80-67 win over North Carolina to advance to the Final Four on Sunday, March 25, 2012 at the Edward Jones Dome in St. Louis.


Unless and until the NCAA, its affiliated television networks and all media covering college basketball denounce and distance themselves from the dreaded Rating Percentage Index, or RPI, rankings, March Madness seeding won’t approach anything that resembles fair.

The RPI places far too much emphasis on a team’s schedule and too little on how teams perform in games. And every time you hear that a school had such and such record against top 50 schools, the top 50 referenced is the deeply flawed RPI rankings.

When the actual NCAA tournament bracket is released, fill out your sheet and advance teams based on how they rank in the RPI. Then do the same for the Associated Press poll, and Sagarin Predictor. What you will find is that the AP poll wins and RPI finishes last.

Yet, for some reason, when humans assemble the brackets, they fall back on computers, falsely assuming machines to have more intelligence than the humans who vote.

It’s such a surrender mentality. If only Rod Serling were here to warn us about how that sort of thinking puts us on the fast track to extinction.

Anyway, as a means of creating Selection Sunday buzz, CBS held its mock selection show Sunday, revealing its projected No. 1, 2, 3 and 4 seeds.

Consider the curious case for Michigan State (24-3, fresh off a victory over Purdue, riding an eight-game winning streak) appearing as a No. 3 seed. The justification: The Spartans have just two top 50 victories.

That shocked me, since I give no credibility whatsoever to RPI and, therefore, never look at it. In the far superior Sagarin rankings, Michigan State is 9-5 vs. top 50 schools.

I point that out to show the huge disparity in computer rankings. But the truth is, common sense trumps all machine-generated attempts at fairness.

Michigan State (24-3) has not lost to a team that doesn’t appear in the top 20 in the current AP poll.

The Spartans blasted North Carolina by 18 points, torched a decent Notre Dame team by the same margin and abused a so-so Maryland squad by 30 points the first time, and then six points on the road.

At worst, Michigan State, ranked No. 4 in the AP poll, should be a 2 seed.

And then, there is the curious case of Kansas as a No. 2 seed and Texas Tech as a No. 3.

The Red Raiders hold a one-game lead on Kansas in the Big 12 and beat the Jayhawks by a dozen points in Allen Fieldhouse.

Humans can figure out that to this point, Tech has had a better season than KU. But too many humans prefer to trust computers.

(Once humans decide virtual reality is superior to reality in all areas, public and private, it will mark the beginning of the end of the species, not to mention bars.)

Texas Tech has an average scoring margin of plus-6 in the Big 12, compared to plus-0.9 for Kansas. But Kansas played a tougher, nonconference schedule and, therefore, has a higher RPI (seven, five spots ahead of Tech) and has fared better on the road than the Red Raiders, who have not lost at home. The committee is putting a premium on road victories this season.

As long as the RPI subtly remains the crutch used by those with Selection Sunday influence, Kansas never will be seeded worse than it deserves. There will be years where KU has the misfortune of being dropped in a loaded bracket, but an unfair seeding never will be a problem because it always plays a strong schedule.

In some years, Kansas gets a higher seed than it deserves, which leads to the misconception that the Jayhawks underachieve in the tournament, because they have been to just two Final Fours in 14 finished seasons under Bill Self. Never mind that KU has made it to seven Elite Eights, or 50 percent of the time. Some will always feel that if a school is seeded No. 1, it by definition underachieves if it doesn't advance to the Final Four.


Edward Daub 3 years, 11 months ago

January , February , Izzo.

MSU is projected as a #3 Seed (Mock CBS) because they have only 2 victories over Top 50 Teams.

However, the Sagarin Rankings show MSU as 9-5 versus Top 50 Teams?

Golly, how is this possible, mathematically speaking?

Well, DUH , it must be a different definition of Top 50 Teams.

If KU is truly a #2 Seed, you can also blame the "Dilution of Talent" caused directly by the ill-advised "1 and 1" Done Rule.

Bob Zielinski 3 years, 11 months ago

Keegan is miscounting the Sagarin wins. The teams played and beaten in the Top 25 games are also included in the Top 50 record. So, Michigan State is actually 6-3 against Top 50 teams right now.

Marius Rowlanski 3 years, 11 months ago

Keegan isn't miscounting anything. It's the NCAA's decision to use the RPI when it's time to seed the tournament.

I would much prefer using Sagarin ratings instead.

Bryce Landon 3 years, 11 months ago

I don't want Kansas seeded #2. They look like they could lose to anyone at any time, and they very well could join the infamous and exclusive "2-seeds that lost to 15-seeds" club - a club that includes Iowa State and Missouri, and even fellow bluebloods Duke, Arizona, Georgetown, and Michigan State.

Robert Brock 3 years, 11 months ago

KU looks like a 4 or even a 5 seed. Out of gas...

Brad Watson 3 years, 11 months ago

Just an opinion....but I think every team in the Big 12 is good enough to be in the tournament and I think all could win at least one game.....this league is really really good this year...very good basketball being played and it should help us in March.

Bob Zielinski 3 years, 11 months ago

Keegan may not like RPI and prefer Sagarin but Sagarin also shows KU with the most Top 25 wins and the most Top 50 wins of all teams in Division I. Only Oklahoma has as many Top 25 wins. Obviously this is not a vintage KU team but to say KU does not deserve a 2 seed and the Sagarin is a better measure of a team's quality kind of ignores the reality of what the preferred measure (Sagarin) is showing us.

Aaron Paisley 3 years, 11 months ago

KU's current body of work does not justify them being a top 2 seed. I'll list off 8 teams right now better than Kansas that deserve a top 2 seed. Virginia, Michigan St., Villanova, Purdue (those are the 4 teams that deserve a 1 seed right now). Texas Tech, Xavier, Cincinnati, and Auburn (those are the teams that deserve a 2 seed right now). KU is at best a 3 seed right now and playing their way to a 4 seed.

Tony Bandle 3 years, 11 months ago

Sagarin/smagarin...the NCAA is all about what it is always all about..$$$$$$$$$!!

The teams that get in are the teams that put fannies in the seats and bucks in the NCAA pockets. You can use stats to make anything look good and when it comes to seeding and selection, the NCAA is the Dark Lord of the Realm!!

Marius Rowlanski 3 years, 11 months ago

RPI = garbage in – garbage out. Basically allowing the NCAA to maximize revenue.

Allin Herring 3 years, 11 months ago

I am not sure how Brad can say the big 12 is really good this year? They can not even WIN the big 12/SEC challenge with Kentucky having a down year.

Steve Zimmerman 3 years, 11 months ago

Ditto here. which season do those reporters ever say big12 a so-so league?

Tracey Graham 3 years, 11 months ago

None of these rankings are perfect. And RPI isn't the only thing that the selection committee takes into account when they seed teams.

The problem with the AP rankings, is how do you seed the teams that aren't ranked in the AP Top 25? I know there are a few teams in the "others receiving votes" but that may be another 5-6 teams. Not 39. So how do you determine if (for example) the SWAC champion should be seeded higher than the MEAC champ if the 2 teams have similar W-L records? So the committee needs another tool aside from the AP rankings to seed the teams. Whether RPI should be that other tool can be debated, but the the selection committee can't rely simply on the AP rankings

Commenting has been disabled for this item.