#222 Harvard (6-8)

avg: 709.9  •  sd: 60.03  •  top 16/20: 0%

Click on a column to sort  • 
# Opponent Result Game Rating Status Date Event
356 Harvard-B** Win 13-2 628.93 Ignored Mar 9th MIT Invite
340 MIT-B Win 13-1 759.35 Mar 9th MIT Invite
218 MIT Loss 6-7 614.3 Mar 9th MIT Invite
305 Northeastern-C Win 8-5 812.44 Mar 9th MIT Invite
110 Berry Loss 9-15 692.07 Mar 15th Tally Classic XIX
139 Florida State Loss 11-15 702.52 Mar 15th Tally Classic XIX
216 Minnesota-Duluth Win 12-10 979.9 Mar 15th Tally Classic XIX
120 Connecticut Loss 5-12 579.02 Mar 29th East Coast Invite 2025
174 Delaware Loss 7-11 471.11 Mar 29th East Coast Invite 2025
179 Pennsylvania Loss 7-10 522.23 Mar 29th East Coast Invite 2025
192 Princeton Win 8-7 979.4 Mar 29th East Coast Invite 2025
157 Johns Hopkins Loss 7-14 422.95 Mar 30th East Coast Invite 2025
192 Princeton Loss 8-11 488.8 Mar 30th East Coast Invite 2025
236 NYU Win 15-4 1249.53 Mar 30th East Coast Invite 2025
**Blowout Eligible

FAQ

What is the uncertainty of the mean?
The uncertainty of the mean is equal to the standard deviation of the set of game ratings, divided by the square root of the number of games. We treated a team’s ranking as a normally distributed random variable, with the USAU ranking as the mean and the uncertainty of the ranking as the standard deviation
What's the algorithm again?
  1. Calculate uncertainy for USAU ranking averge
  2. Model ranking as a normal distribution around USAU averge with standard deviation equal to uncertainty
  3. Simulate seasons by drawing a rank for each team from their distribution. Note the teams in the top 16 (club) or top 20 (college)
  4. Sum the fractions for each region for how often each of it's teams appeared in the top 16 (club) or top 20 (college)
  5. Subtract one from each fraction for "autobids"
  6. Award remainings bids to the regions with the highest remaining fraction, subtracting one from the fraction each time a bid is awarded
Is there a longer explanation of all this?
There is an article on Ulitworld written by Scott Dunham and I that gives a little more context (though it probably was the thing that linked you here)