Round-robin groups (3x12), tiebreaking games, playoff brackets (6x6), still more tiebreaking games, and finals produced this ordering at the top: 1. Michigan A (undefeated) 2. Virginia 3. Chicago 4. Princeton A (undergraduate champion) 5. Berkeley A 6. Illinois 7. Michigan B (undergraduate finalist) 8. Florida 9. Harvard 10. Florida Atlantic 11. Emory 12. Yale A Chicago, Princeton A, and Berkeley A had an exciting tiebreaker playoff (as explained below) to rank the 3rd through 5th place teams. Michigan B, Florida, and Harvard also played off, not only to rank 7th through 9th but also to determine who would face Princeton A in this year's undergraduate final. (I believe this is the first time we used a final match to determine the undergraduate champion, as opposed to final standings. We guaranteed a final if the top two undergraduate teams finished within six spots of each other.) I'm curious, both personally and as an NAQT member, what people thought of this year's format, since it differed from the format used in previous ICT's. (Past years' format described at the end of this post.) Feedback is encouraged: Even if you'd prefer not to post, you can e- mail naqt_at_... to support or oppose {round robin with playoff brackets}, {power-matching and ladder play}, or any other aspect of how the tournament runs -- or anything else about NAQT. The 36 teams were split into three 12-team brackets for full round- robin play in rounds 1-6 (Friday night) and 7-11 (Saturday morning). After round 11 (and immediately before lunch), any teams in the same bracket with the same overall record faced each other in a tiebreaking game (or sequence of 13-tossup half-games) to resolve each bracket into a 1-12 ordering of teams. (For half-games to break three way ties, order of play was determined by points per tossup, with the top team facing the winner of the other two teams.) The top two finishers from each bracket formed a playoff bracket, as did the teams in spots 3-4, 5-6, and so on. Within each playoff bracket, the higher finishers from each prelim bracket (1st, 3rd, 5th, etc.) entered with a one-game advantage over the lower finishers. In rounds 12-15, teams played the four playoff bracket opponents whom they hadn't yet already faced. Those games plus the one-game advantage (or disadvantage) yielded playoff records over five games. Following round 15 came still more tiebreaking games (or half- games). In addition to the ties for third and seventh, all teams that finished in ties (by playoff bracket record) were entitled to play off those ties; many teams availed themselves of this. Knock on wood that I have the scoresheets for all the teams who did. (The tiebreaks after round 11 involved slightly more than half the field. Post-playoff tiebreaks could have involved about that many teams, though some sets of teams opted not to break their ties. Still, regularly scheduled games plus tiebreaks yielded 17 packs worth of play for a lot of teams.) In past years, the Div I format has been roughly the following (someone can correct or complete this as needed): 1. Friday night (six rounds), pre-assigned matchups wherein each team's set of opponents would be roughly balanced by both geography and expected strength 2. Saturday morning (five rounds), power-matching where pairings for round N+1 are determined by results through N-1 (so that the act of making pairings doesn't delay the event) 3. Saturday afternoon (four rounds), ladder play. (Teams #1 and #2 face each other, #3 and #4, #5 and #6, etc. If the team with the worse ladder card wins, teams switch cards. Next round, #2 faces #3, #4 vs. #5, etc.) My brief take on the format comparison is that round robins are much easier to run (even the pre-playoff tiebreaks were straightforward) but don't produce as reliable a full 1-to-N ranking.
This archive was generated by hypermail 2.4.0: Sat 12 Feb 2022 12:30:45 AM EST EST