pic: Sum of team numbers versus match number in Curie

[Jeff Waegelin]

This pattern had a great effect on seeding for eliminations, too, at least in the Galileo division. Several of the top teams with lower team numbers had some pretty brutal matches where they beat up on each other, leading many of them to be outside the top 8, even though they were top hurdlers. At the same time, there were several teams in the higher number ranges that made it into the top 8, possibly because of an easier match schedule. I don't begrudge any of those teams - they all performed well with the match schedules they were given. I do think it would have been better, however, if we had gotten a slightly more equitable distribution of teams in each match.

Randomizing the list before running the algorithm would strike me as a fair way to do it. I don't care as much if I end up in a "pool" of teams - it's the only way to assure good match spacing - so long as the teams in that group are randomly assigned, and not dictated by team number.