I've been having a pleasant Monday discussion with some friends on Facebook about where the Committee appears to have properly seeded teams (Iowa State) versus where the seeds appear to be way off (Wisconsin, Minnesota).
Of course, my mind ran with the idea and I ended up with this thought: let's see how three major ratings indices evaluate the work of the Committee. I'm going to look at (1) RPI, (2) Ken Pomeroy's ratings, and (3) Jeff Sagarin's ratings to see where the Committee lined up well with the metrics and where it appears to have missed. I'll group the results by how close (or not) the Committee got the seeds compared to what the metrics say.
For the purposes of this article, I'm not going to examine the 13-16 seeds as they would be the bottom 16 teams in the Tournament regardless of who qualifies ahead of them. I will include automatic-bid Nevada and Middle Tennessee State as both schools made the 12 line for consistency's sake.
The beauty of 2017 is that the selection Committee releases a full list of seeds from 1-68 so we don't have to speculate on exactly how close a particular 7 and 8 seed are: the Committee has already told us.
In the end, I examined the placement at every team that showed up in the top-60 of at least one of the three rankings considered. Here are the findings with the team's actual seed listed first and its expected seed listed second (except for those seeds predicted correctly).
CORRECTLY SEEDED (9)
Teams: Villanova (1), North Carolina (1), Gonzaga (1), Louisville (2), Purdue (4), Iowa State (5), Marquette (10), Nevada (12), Middle Tennessee State (12)
MISSED BY ONE (18)
Teams: Kansas (1 - 2), Kentucky (2 - 1), Duke (2 - 3), Oregon (3 - 4), Baylor (3 - 4), Florida (4 - 3), West Virginia (4 - 3), Notre Dame (5 - 6), Cincinnati (6 - 5), Creighton (6 - 7), Michigan (7 - 6), Wisconsin (8 - 7), Miami (8 - 9), Arkansas (8 - 9), Vanderbilt (9 - 10), VCU (10 - 11), Rhode Island (11 - 10), USC (12 - #1 NIT)
MISSED BY TWO (14)
Teams: Arizona (2 - 4), Florida State (3 - 5), UCLA (3 - 5), Butler (4 - 6), Virginia (5 - 3), SMU (6 - 4), St. Mary's (7 - 5), South Carolina (7 - 9), Dayton (7 - 9), Michigan State (9 - 11), Oklahoma State (10 - 8), Xavier (11 - 9), Wake Forest (11 - 9), Kansas State (12 - 10)
MISSED BY THREE (3)
Teams: Minnesota (5 - 8), Northwestern (8 - 11), Virginia Tech (9 - 12)
MISSED BY FOUR (3)
Teams: Maryland (6 - 10), Seton Hall (9 - #1 NIT), Providence (11 - #3 NIT)
MISSED BY FIVE (1)
Teams: Wichita State (10 - 5)
When I look at this list, I'm struck by just how many clubs fall into the last four groups, suggesting a rather significant seeding miss by the committee. In particular, it appears as though the committee has significantly overseeded the Rose Bowl conferences, the Pac 12 and the Big Ten. Seeing four of the Big Ten's contingent be overseeded by an average of three seeds apiece is alarming, suggesting that the conference's teams may not be nearly as strong as the committee has insinuated.
There are a couple of nuggets that I'd like to point out having compiled this research, as follows:
1. RPI loves Arizona (2), but KenPom (20) and Sagarin (18) both see the 'Cats as a 5. Yikes.
2. The Committee has a lot of love for Butler, ranking them as the top #4 seed. RPI (15) saw the Bulldogs as a solid 4, but both KenPom and Sagarin (26 each) saw them as a solid #7.
3. RPI (18) and the Committee (#5 seed) gave Virginia a raw deal. The Hoos should have been on the 2/3 border according to KenPom (7) and Sagarin (9).
4. What in the world happened with Maryland? The Committee gave the Terps a 5 seed, but RPI (34), KenPom (45), and Sagarin (42) all saw them as much more of a bubble team.
5. RPI killed Wisconsin, suggesting a 9 seed for the Badgers. The Committee gave them an 8, but KenPom (23) had them as a 6 and Sagarin (17) gave them the top 5 seed.
6. Wichita State is tough to stomach. RPI wasn't a fan (32), though even RPI gave the Shockers an 8 seed, but KenPom (8) had them as a 2 seed (!) while Sagarin (11) gave them a 3. Sitting on the 10 line is brutal and borderline inexplicable.
7. Not sure what the Committee saw in Providence. All three systems had the Friars between 56th and 58th, yet the Committee gave them the #42 overall seed instead of sticking them in the middle of the NIT field.
8. The rawest deal yesterday of any team in America? Texas Tech. KenPom and Sagarin both had the Red Raiders at #46 -- equivalent to a trip to the First Four as a #12 seed in the NCAA Tournament -- but RPI had them at 123rd. Their result? No postseason at all (not even the NIT).
9. Both Illinois State and Clemson should have made the Tournament as First Four participants, though their status as two of the top-eight seeds in the NIT suggests that they were more or less properly seeded.
10. A team to watch: UNC-Wilmington. The metrics suggest that, even if conference championships weren't used to determine automatic bids to the Tournament, UNCW would have ended up on the 13 line.
11. Tough misses for Georgia, Cal, and Houston. All three squads rated no worse than 58th in any metric yet none came in higher than 49th in any service. They lived on the wrong side of the bubble. Like Illinois State and Clemson, all three are top-eight clubs in the NIT.
No comments:
Post a Comment