Final(ish) Regular Season Ratings & Bid Voting

First things first, I want to make it clear, once again, that I work for Michigan.  Relatedly, the ratings are the result of a mathematical algorithm and are not a reflection of my personal opinions.  If anybody has questions about particular placements, contact me on facebook and I might be able to provide an explanation.

Second, these ratings do not include the Dartmouth Round Robin.  Those results are not yet posted on Tabroom, and it would be a serious hassle for me to try to manually code them.  If I were to guess what effect their inclusion would have, I believe there's a good chance Emory SK could jump into the same rating ballpark as Michigan/Harvard and Wake Forest AS would drop to 7th.

EDIT: The ratings do now include Dartmouth.  The effect was basically as predicted.

Third, I know there's a lot of stress about bid voting.  I want to make it clear that these ratings are not intended to replicate the decisions of bid voters.  I would not encourage anybody to take the top 16 here and assume that those teams are necessarily first rounds (nor would I expect that any bid voter would do this).  The ratings provide one way of processing and understanding who the best teams in the country are.  However, because of the nature of the beast, the ratings are also capable of some misses when something is limited/different/peculiar about the data for a team: maybe a team hasn't attended as many tournaments, maybe their schedule has included more regionals at the expense of national level tournaments, maybe they've had multiple partners, etc.

Fourth, caveats aside, the ratings are information.  They are a way of processing results that produces a good picture of who is likely to beat whom.  To that extent, I think that they can a useful tool for bid voters.  I encourage people to maybe deemphasize the "ranking" number and instead look more at the "rating" number and find places where ratings cluster together.  Small ratings differences might indicate that there's not a meaningful difference between teams.  For example, a 15 point rating difference means that the higher ranked team is only about a 52% favorite against the lower ranked team (if you want to play around with these numbers, the "Prediction Calculator" page found on the menu bar above lets you calculate win probabilities).  Small differences can be a product of noise or some random variation that happens naturally.

Fifth, participation awards for everybody!  Everybody's a winner!

Sixth, some teams are more winners than others!  Congrats to Emory KS, Wake Forest AS, Rutgers NM, and Liberty BC on some great performances over the last few weeks.