Here is a good site to bookmark: https://www.masseyratings.com/cb/compare.htm
It will show you the computer rankings for 31 different models, including NET, Kenpom (Pomeroy), Sagarin, Massey and others (you can download via CSV if desired).
Here are the current composite rankings as of Feb 14th for the Top 16 (Top 4 seeds):
So Iowa's composite ranking across all 31 computer models - an average or mean - puts them at 9th, or a 3 seed.
However, if you look at MEDIAN instead of MEAN, Iowa would rank 6th, or a 2 seed.
What this means is that there is a wide disparity of rankings for Iowa, which is evidenced by the Iowa having the largest Standard Deviation of 9.05.
Here's what the more noteworthy models have for Iowa:
Note that the average for these 4 alone is 5.5, putting Iowa as a 2 seed.
Here are the 5 lowest ratings for Iowa among the 31 models:
Forgive me if I challenge a few of these models just on name alone - JELLY JUKE, and JOBY NITTY GRITTY. Not sure on those![Smile :) :)](data:image/gif;base64,R0lGODlhAQABAIAAAAAAAP///yH5BAEAAAAALAAAAAABAAEAAAIBRAA7)
I have heard of the other 3 at least. But you can see that Iowa's mean or average is skewed mostly by the Jelly Juke model, which has them 29th. Perhaps our color scheme of black and gold has some negative relationship to a poor tasting jelly bean....who knows.
In the end, I would hope that the committee might start with some type of weighted composite model, where maybe there is 50% weight on NET rating, and then 10% weight on 5 other models. You could use the result to create initial seeding, with the rule that a team can't move more than 1 seed placement up or down from the composite rating based on subjective judgments of the panel.
Based on that approach, I would think Iowa would be a 2 seed, at worst a 3, but I'd really like to know the panel's reasoning if they bump Iowa down a spot. Ideally the panel is transparent about all of this so that if a team is moved, the Panel is willing to explain it's reasoning (injuries, et al).
Iowa was a 4 seed in the initial Top 16 seeds put out last week, and was a 4 seed in last Friday's Lunardi Bracket.
I'm not sure I understand either of those, quite frankly, since all the objective measures have Iowa as a 2 seed, essentially, and we actually have injuries in our favor if one considers circumstantial factors. These models already take into account all data based factors such as SOS, scoring margin (varies by model), and who you lost to. /confused
It will show you the computer rankings for 31 different models, including NET, Kenpom (Pomeroy), Sagarin, Massey and others (you can download via CSV if desired).
Here are the current composite rankings as of Feb 14th for the Top 16 (Top 4 seeds):
Team | Conf | WL | Rank | Mean | Trimmed | Median | StDev |
Gonzaga | WCC | 20-0 | 1 | 1.06 | 1.03 | 1 | 0.25 |
Baylor | B12 | 17-0 | 2 | 2 | 1.97 | 2 | 0.45 |
Michigan | B10 | 14-1 | 3 | 4.18 | 3.86 | 3 | 2.65 |
Houston | AAC | 17-2 | 4 | 6.6 | 6.48 | 7 | 2.51 |
Ohio St | B10 | 17-4 | 5 | 7.24 | 6.97 | 7 | 3.55 |
Illinois | B10 | 14-5 | 6 | 7.82 | 6.79 | 6 | 7.6 |
Alabama | SEC | 17-5 | 7 | 8.16 | 7.48 | 7 | 4.84 |
Virginia | ACC | 15-3 | 8 | 8.45 | 8.24 | 8 | 3.11 |
Iowa | B10 | 15-6 | 9 | 10.23 | 9.24 | 6 | 9.05 |
USC | P12 | 17-3 | 10 | 13.61 | 13.41 | 13 | 4.47 |
Villanova | BE | 13-3 | 11 | 13.61 | 12.83 | 11 | 8.06 |
West Virginia | B12 | 14-6 | 12 | 15.98 | 15.24 | 14 | 6.27 |
Creighton | BE | 16-5 | 13 | 17.45 | 17.03 | 15 | 6.89 |
Oklahoma | B12 | 13-5 | 14 | 17.77 | 17.14 | 18 | 7.08 |
Florida St | ACC | 11-3 | 15 | 17.9 | 17.9 | 17 | 6.13 |
Tennessee | SEC | 14-5 | 16 | 19.11 | 18.86 | 18 | 6.88 |
So Iowa's composite ranking across all 31 computer models - an average or mean - puts them at 9th, or a 3 seed.
However, if you look at MEDIAN instead of MEAN, Iowa would rank 6th, or a 2 seed.
What this means is that there is a wide disparity of rankings for Iowa, which is evidenced by the Iowa having the largest Standard Deviation of 9.05.
Here's what the more noteworthy models have for Iowa:
Team | MASSEY | NET | POM (KENPOM | SAGARIN |
Gonzaga | 1 | 1 | 1 | 1 |
Baylor | 2 | 2 | 2 | 2 |
Michigan | 3 | 3 | 3 | 3 |
Houston | 7 | 5 | 6 | 7 |
Ohio St | 4 | 7 | 7 | 8 |
Illinois | 5 | 4 | 5 | 5 |
Alabama | 8 | 9 | 8 | 9 |
Virginia | 10 | 6 | 9 | 6 |
Iowa | 6 | 8 | 4 | 4 |
USC | 13 | 15 | 12 | 20 |
Villanova | 9 | 13 | 11 | 10 |
West Virginia | 17 | 16 | 17 | 13 |
Creighton | 18 | 22 | 15 | 11 |
Oklahoma | 12 | 17 | 21 | 19 |
Florida St | 22 | 25 | 20 | 16 |
Tennessee | 19 | 12 | 16 | 15 |
Note that the average for these 4 alone is 5.5, putting Iowa as a 2 seed.
Here are the 5 lowest ratings for Iowa among the 31 models:
Team | BIHL | JELLY JUKE | JOBY NITTY GRITTY | KIRKPATRICK | PUGH |
Gonzaga | 1 | 1 | 1 | 1 | 1 |
Baylor | 2 | 2 | 2 | 2 | 2 |
Michigan | 3 | 3 | 3 | 3 | 3 |
Houston | 5 | 4 | 10 | 5 | 5 |
Ohio St | 4 | 5 | 4 | 4 | 4 |
Illinois | 8 | 14 | 5 | 7 | 6 |
Alabama | 7 | 7 | 6 | 12 | 8 |
Virginia | 6 | 8 | 8 | 10 | 7 |
Iowa | 14 | 29 | 15 | 16 | 18 |
USC | 9 | 6 | 17 | 15 | 12 |
Villanova | 11 | 15 | 7 | 8 | 15 |
West Virginia | 12 | 10 | 11 | 9 | 14 |
Creighton | 23 | 33 | 25 | 19 | 24 |
Oklahoma | 10 | 11 | 9 | 22 | 9 |
Florida St | 25 | 25 | 18 | 13 | 11 |
Tennessee | 24 | 27 | 16 | 24 | 25 |
Forgive me if I challenge a few of these models just on name alone - JELLY JUKE, and JOBY NITTY GRITTY. Not sure on those
I have heard of the other 3 at least. But you can see that Iowa's mean or average is skewed mostly by the Jelly Juke model, which has them 29th. Perhaps our color scheme of black and gold has some negative relationship to a poor tasting jelly bean....who knows.
In the end, I would hope that the committee might start with some type of weighted composite model, where maybe there is 50% weight on NET rating, and then 10% weight on 5 other models. You could use the result to create initial seeding, with the rule that a team can't move more than 1 seed placement up or down from the composite rating based on subjective judgments of the panel.
Based on that approach, I would think Iowa would be a 2 seed, at worst a 3, but I'd really like to know the panel's reasoning if they bump Iowa down a spot. Ideally the panel is transparent about all of this so that if a team is moved, the Panel is willing to explain it's reasoning (injuries, et al).
Iowa was a 4 seed in the initial Top 16 seeds put out last week, and was a 4 seed in last Friday's Lunardi Bracket.
I'm not sure I understand either of those, quite frankly, since all the objective measures have Iowa as a 2 seed, essentially, and we actually have injuries in our favor if one considers circumstantial factors. These models already take into account all data based factors such as SOS, scoring margin (varies by model), and who you lost to. /confused
Last edited: