clock menu more-arrow no yes

Filed under:

Stats Corner: Bias in the NCAA Seeding?

Which teams were over-seeded or under-seeded

Baylor v Gonzaga

With the Final Four coming up, and the Mountain West having another bad tournament year which we are not going to talk about (we don’t talk about Broncos, no, no, no), Stat’s Corner is going to cover the teams who were over-seeded and under-seeded in the tournament. I have written about this is previous years, see here, and I usually get some comment about “And in other news, water is wet”. While this may seem like an exercise in futility, if the problem is not pointed out repeatedly, as I did here, it will never get fixed.

There are several, supposedly, metrics which the committee uses to determine seeding, but according the NCAA website “The 2021-22 men’s basketball season marks the fourth season of the NCAA Evaluation Tool (NET) rankings, which replaced the RPI prior to the 2018-19 season as the primary sorting tool for evaluating teams.” I have written several articles on the NET ranking, explaining what it is and comparing it to the RPI, but a brief summary is there are four quadrants, emphasis is placed in Q1 wins and Q3/Q4 losses with the quadrants being

Quadrant 1: Home 1-30, Neutral 1-50, Away 1-75

Quadrant 2: Home 31-75, Neutral 51-100, Away 76-135

Quadrant 3: Home 76-160, Neutral 101-200, Away 135-240

Quadrant 4: Home 161-353, Neutral 201-353, Away 241-353

There are 68 teams in the tournament, and with automatic bids, the top 68 teams will not get in. Also, this is a look at where teams should have been seeded, not on the result in the tournament. St. Peter’s made a great run, but as the 124th ranked team, the argument could be made they should not have been there. The committee should or cannot make a tournament based on what they think will happened, but on the results of the teams during the regular season. Due to this, we are only looking at the top 48 seeds in the tournament and their corresponding NET rankings, when I first wrote this I included KenPom and Sagarin rankings, but after finding the quote on the NCAA website, I decided to stick with just NET. The full seeding and NET rankings can be found at the end of the article

Those who were seeded higher

There were seven teams who were seeded ten or more places higher than their NET rankings, and surprise they were all from “Power” conferences. In fact, the top 10 over seeded schools are all from Power conferences. The biggest discrepancy was Rutgers who had a NET ranking of 77, but was seeded 44th, which is a difference of 33. Miami had a difference of 24, being seeded 38th and a NET ranking of 62. Creighton was 22, seeded 33 and ranked 55. The two most influential in terms of the tournament were Providence getting a 15 seeded (4th seed) despite a NET ranking of 33 (which would be a 9th seed) and Wisconsin getting being seeded 9th (3rd seed) in spite of a NET ranking of 24 (which should have been a 6th seed). The last two double digits boosts went to USC NET 35 to seeded 25th and TCU NET 44 to seeded 34.

Those who were seeded lower

On the reverse side there were four schools who seeded ten or more places lower than they should have, and if the paragraph above did not surprise you I am not sure you will be surprised to see that three out of the four were Mid-Majors. The one exception was Virginia Tech who had a NET ranking of 27 (worthy of a 7th seeded) but was seeded 46th (a 12th seed). The three Mid-Majors who were cheated are Loyola Chicago who’s NET 23 would have been a 6th seed, instead of being seeded 39th with a 10th seed, San Francisco with a NET 22 instead was handed a 37th(see Loyola for the seeding it is the same). The biggest influence on the tournament in this category was Houston who had a NET of 3 which would have netted them (sorry, had to use that pun once) a 1 seed, but instead was seeded 18th and given a 5th seed.

Overview

As with most things in the NCAA the Power Schools get the advantage and the Mid-Majors/Weaker Conferences have the deck stack against them as evidenced by 100% of teams who were seeded double digits higher than they should have been coming from Power Conference and 75% of teams who were seeded double digits lower that they should have been coming from Mid-Majors. Part of the issue is the Tournament is a huge money maker for the NCAA and they want to have the biggest draw to have the most viewers. If they were to be completely honest, the 68-team tournament is a terrible way to choose the champion. If you wanted a way to have the best team win the title, it should be 16 teams at most and then the NIT could expand to 128 teams if you wanted more teams to have a post-season. However, since 68 teams generates a billion dollars a year, things are not going to change. That being said (or written), it does not justify seeding Power teams too high so that they have an easier path and will last longer in the tournament and seeding Mid Majors too low so that they will be happy to make it and exit quickly so that the big draws will be on TV. One could argue that Houston did not deserve a #1 seed with their NET 3 ranking, and I would agree, but if you do not like the system you put in place, either change it or quit telling people it is the “primary sorting tool” when it is being ignored to help certain conferences.

NCAA Team by Seed, a positive difference means the team was seeded higher than their NET ranking a negative difference means the team was seeded lower than their NET ranking

Seed NET Difference

Gonzaga (26-4) 1 1 0

Arizona (31-3) 2 2 0

Kansas (28-6) 3 6 3

Baylor (26-6) 4 4 0

Auburn (27-5) 5 11 6

Kentucky (26-7) 6 5 -1

Villanova (26-7) 7 8 1

Duke (28-6) 8 12 4

Wisconsin (24-7) 9 24 15

Tennessee (26-7) 10 7 -3

Purdue (27-7) 11 13 2

Texas Tech (25-9) 12 9 -3

UCLA (25-7) 13 10 -3

Illinois (22-9) 14 15 1

Providence (25-5) 15 33 18

Arkansas (25-8) 16 20 4

UConn (23-9) 17 17 0

Houston (29-5) 18 3 -15

Saint Mary’s (25-7) 19 19 0

Iowa (26-9) 20 14 -6

Alabama (19-13) 21 30 9

LSU (22-11) 22 18 -4

Texas (21-11) 23 16 -7

Colorado State (25-5) 24 28 4

USC (26-7) 25 35 10

Murray State (30-2) 26 21 -5

Michigan State (22-12) 27 36 9

Ohio State (19-11) 28 26 -2

Boise State (27-7) 29 29 0

North Carolina (24-9) 30 31 1

San Diego State (23-8) 31 25 -6

Seton Hall (21-10) 32 37 5

Creighton (22-11) 33 55 22

TCU (20-12) 34 44 10

Marquette (19-12) 35 42 7

Memphis (21-10) 36 33 -3

San Francisco (24-9) 37 22 -15

Miami (23-10) 38 62 24

Loyola Chicago (25-7) 39 23 -16

Davidson (27-6) 40 41 1

Iowa State (20-12) 41 49 8

Michigan (17-14) 42 36 -6

Wyoming (25-8) 43 50 7

Rutgers (18-13) 44 77 33

Indiana (20-13) 45 38 -7

Virginia Tech (23-12) 46 27 -19

Notre Dame (22-10) 47 53 6

UAB (27-7) 48 46 -2

NCAA Tournament Teams by seeding Higher or Lower than NET ranking. Positive differences are seeded higher than ranking, negative differences are seeded lower than ranking Seed

Seed NET Difference

Rutgers (18-13) 44 77 33

Miami (23-10) 38 62 24

Creighton (22-11) 33 55 22

Providence (25-5) 15 33 18

Wisconsin (24-7) 9 24 15

USC (26-7) 25 35 10

TCU (20-12) 34 44 10

Alabama (19-13) 21 30 9

Michigan State (22-12) 27 36 9

Iowa State (20-12) 41 49 8

Marquette (19-12) 35 42 7

Wyoming (25-8) 43 50 7

Auburn (27-5) 5 11 6

Notre Dame (22-10) 47 53 6

Seton Hall (21-10) 32 37 5

Duke (28-6) 8 12 4

Arkansas (25-8) 16 20 4

Colorado State (25-5) 24 28 4

Kansas (28-6) 3 6 3

Purdue (27-7) 11 13 2

Villanova (26-7) 7 8 1

Illinois (22-9) 14 15 1

North Carolina (24-9) 30 31 1

Davidson (27-6) 40 41 1

Gonzaga (26-4) 1 1 0

Arizona (31-3) 2 2 0

Baylor (26-6) 4 4 0

UConn (23-9) 17 17 0

Saint Mary’s (25-7) 19 19 0

Boise State (27-7) 29 29 0

Kentucky (26-7) 6 5 -1

Ohio State (19-11) 28 26 -2

UAB (27-7) 48 46 -2

Tennessee (26-7) 10 7 -3

Texas Tech (25-9) 12 9 -3

UCLA (25-7) 13 10 -3

Memphis (21-10) 36 33 -3

LSU (22-11) 22 18 -4

Murray State (30-2) 26 21 -5

Iowa (26-9) 20 14 -6

San Diego State (23-8)31 25 -6

Michigan (17-14) 42 36 -6

Texas (21-11) 23 16 -7

Indiana (20-13) 45 38 -7

Houston (29-5) 18 3 -15

San Francisco (24-9) 37 22 -15

Loyola Chicago (25-7) 39 23 -16

Virginia Tech (23-12) 46 27 -19