clock menu more-arrow no yes mobile

Filed under:

The DPI: First Weekend Carnage

The DPI Has Been Destroyed, Kind Of...

Syndication: Journal Sentinel Ebony Cox / Milwaukee Journal Sentinel / USA TODAY NETWORK

What is left of the S-tier?

S-Tier Week 2

DPI Rank Seed Team DPI Score
DPI Rank Seed Team DPI Score
1 5 Houston 17
2 4 UCLA 18
3 1 Gonzaga 22
4 3 Texas Tech 23
5 2 Villanova 26
6 1 Arizona 30
7 1 Kansas 31

Five of twelve S-Tier teams didn’t make the second weekend. The biggest surprises, to me at least, were Kentucky and Baylor. Baylor had the best comeback that never was and ran into a red-hot North Carolina team.

There was a 50-point difference between North Carolina and Baylor, and something I kept going back to, man, momentum is hard to understand. Teams like Virginia Tech and Iowa were on fire after their respective conference tournaments and flamed out after 40 minutes.

I still believe that the 2022 national champion is on this short-list, but it did take a hit in the first four days.

The B1G Is What We Thought They Were

B1G Frauds

DPI Rank Seed Teams DPI Score
DPI Rank Seed Teams DPI Score
16 5 Iowa 36
21 4 Illinois 47
24 3 Wisconsin 51
42 7 Michigan St. 75
43 7 Ohio St. 77

Not to lie, but I kind of nailed it here. I didn’t have any of these teams past the sweet 16, and I hope you didn’t either. The only Big Ten teams left are Purdue, and somehow, Michigan. Iowa, lost in the first round, the second biggest upset according to DPI.

Bloody Sunday slayed four consecutive Big 10 teams; Illinois, Ohio State, Michigan State, and Wisconsin. Purdue avoided a 5th loss against future SEC school, Texas. (Purdue shot 40 free throws against Texas).

For the second year in the row, an extremely competitive Big Ten conference has underperformed on the biggest stage of them all. Shame.

The DPI Defied

Why Are You Still Here?

DPI Rank Seed Team DPI Score
DPI Rank Seed Team DPI Score
36 11 Michigan 71
39 4 Providence 72
45 11 Iowa State 77
49 10 Miami 96
54 15 Saint Peter's 129

So, these teams, basically have no business being here. My numbers were right, and these teams got them all wrong. Most of these schools, Saint Peter’s and Iowa State excluded, had a toss-up first round, followed by a big second round upset. None of these schools have had a lower DPI than any of the teams they’ve played.

There may be some recency bias in this statement, but here goes. This is the most memorable Iowa State tournament run in my lifetime. I think the added element of no expectations, and us being on borrowed time basically from the jump, has given this first weekend some extra magic. No expectations, just winners. Winners.

The two schools that annoy the hell out of me, that are still here: Michigan and Providence. Michigan should’ve played in Dayton. Providence is literally the worst 4-seed in the history of time.

I’m personally offended that the ‘showtime’... ‘snowtime’ Jackrabbits were only given one game in the big dance. Those dudes throw lobs, behind the back passes, pull-up from anywhere. Damn, let me determine the individual outcome of every game and the tournament would be so much more fun, seriously.

Biggest Upsets


Winner Loser DPI Difference in Score
Winner Loser DPI Difference in Score
Saint Peter's Kentucky 106
Richmond Iowa 70
New Mexico State UCONN 61
North Carolina Baylor 50




I don’t even want to add input here, I just think it’s hilarious that Iowa and Baylor are both on this list. Richmond had no business beating Iowa, and those crazy bastards did it anyways. The hilarity of this table is enough for me to overlook Tony Soprano University beating the team I had in the championship game.


Where I used the DPI the most in my bracket were the coin-flip games. The 7-10 or 8-9 matchups. These games can go either way, and what I learned, if the DPI of two teams is 15 points or less, I wouldn’t trust the DPI at all. Those two teams are equal, and you just got to go with your gut.

What the DPI is best at, so far, is showing where the committee, in my opinion, messed up in seeding. It’s easy to weed out the teams seeded too high or too low, but when it comes to determining individual games... eh. Needs some work.

There are some super confusing losses that the metrics just can’t wrap their heads arounds. But this is March, and these outcomes don’t care about numbers. As David Fizdale once said... Take that for data. (2:42 mark in video)