College Women's USAU Rankings

2024-25 Season

Data updated through April 28 at 11:45pm EDT

FAQ
Division I // Division III
Rank    Change Team                                                 Record Rating Change Region Conference Div   SoS PDC %
34 1 Cornell ME 1 21-3 1640.76 328 Metro East Eastern Metro East DI D-I 1366.83 273.94 0.2
64 6 Connecticut 17-9 1319.26 249 Metro East Eastern Metro East DI D-I 1194.48 124.79 0.1
79 3 Columbia 13-12 1218.99 289 Metro East Eastern Metro East DI D-I 1172.48 46.52 0.04
93 34 NYU 16-8 1111.44 1 Metro East Eastern Metro East DI D-I 975.98 135.46 0.14
104 15 Yale 11-13 1040.63 141 Metro East Eastern Metro East DI D-I 1100.56 -59.92 -0.05
139 Rutgers 13-15 792.55 227 Metro East Eastern Metro East DI D-I 899.34 -106.79 -0.12
185 31 SUNY-Stony Brook 4-15 518.7 78 Metro East Eastern Metro East DI D-I 830.6 -311.9 -0.38
243 4 Columbia-B 1-9 40.29 34 Metro East Eastern Metro East DI Dev 220.64 -180.35 -0.82

FAQ

The results on this page ("USAU") are the results of an implementation of the USA Ultimate Top 20 algorithm, which is used to allocate post season bids to both colleg and club ultimate teams. The data was obtained by scraping USAU's score reporting website. Learn more about the algorithm here. TL;DR, here is the rating function. Every game a team plays gets a rating equal to the opponents rating +/- the score value. With all these data points, we iterate team ratings until convergence. There is also a rule for discounting blowout games (see next FAQ)
For reference, here is handy table with frequent game scrores and the resulting game value:
"...if a team is rated more than 600 points higher than its opponent, and wins with a score that is more than twice the losing score plus one, the game is ignored for ratings purposes. However, this is only done if the winning team has at least N other results that are not being ignored, where N=5."

Translation: if a team plays a game where even earning the max point win would hurt them, they can have the game ignored provided they win by enough and have suffficient unignored results.