--- In quizbowl_at_yahoogroups.com, dargan_w <no_reply_at_y...> wrote: <<In terms of the NAQT formula and its relative fairness, I will say this -- It sounds an awful lot like what has happened to us in CBI for a team to come in second to a team they're basically equivalent to and not get to go to nationals. I had hoped that this type of situation was what NAQT's formula would remedy.>> That isn't exactly a valid comparison. Going solely by points per 20 TUs, Florida, who won, had 178 points, while Georgia had 148 (350 fewer points, on nine more TUs). Looking at points per 20 TUs across the DivII field, Florida ranked 24th out of all teams, while Georgia was 35th. So if NAQT had simply gone by picking "the best teams," without any sort of guarantees, it's possible (though by no means certain) even Florida would not have received an invite. --STI P.S. Having tried to predict the bids a couple of years ago (and coming fairly close to the actual results), let me just point out two things: 1. Any reasonable system will match at least 80%, maybe even 90% of the teams that do get selected by the real formula. If you tried to predict the bids this year by using some sort of statistical formula, try playing around with it a bit: adding additional variables, re- weighting variables, etc. Unless you add some truly bizarre metric, the list of teams that would receive the invites will remain largely the same. 2. Since NAQT doesn't publicize the order in which at-large bids are awarded (whether they should is debatable), even systems that are far removed from the real formula may give surprisingly high concordance rates.
This archive was generated by hypermail 2.4.0: Sat 12 Feb 2022 12:30:46 AM EST EST