View Single Post
  #15   Spotlight this post!  
Unread 13-03-2007, 20:55
rachal's Avatar
rachal rachal is offline
multipurpose tool
AKA: Rachel
FRC #1072 (Harker Robotics)
Team Role: Alumni
 
Join Date: Jan 2006
Rookie Year: 2005
Location: San Jose, CA -> Cambridge, MA
Posts: 172
rachal is just really nicerachal is just really nicerachal is just really nicerachal is just really nicerachal is just really nice
Send a message via AIM to rachal
Re: "New" 2nd Week Scheduling Algorithm

Quote:
Originally Posted by Wayne C. View Post
My gripe with the first week schedule was not playing with rookies. Rather it was playing the same teams 5 times over.
Here are my team's (1072) qualifying matches for LA (there were 52 teams present):
Allies
597, 1722
599, 1669
702, 1759
399, 2174
597, 1836
702, 2174
254, 1702
606, 1759
691, 1855

Opponents
599, 1138, 1759
687, 812, 1644
4, 1722, 1197
330, 1438, 2272
294, 1266, 2178
599, 1266, 1669
399, 980, 2029
330, 867, 1669
330, 968, 2178

We had nine matches and interacted with a total of 31 teams, many of them multiple times. Getting paired with a few teams multiple times, sure that happens. But we were allied with four teams twice, played against four teams twice, and played against 330 three times. I still had fun, but it seemed repetitive after a while.

Quote:
Originally Posted by Ben Mitchell View Post
Personally I don't think that team pairings should be made with a particular algorithm - pure randomization works for me.
Same here. Complete randomization will let teams demonstrate their capabilities much more accurately (at least relative to each other), since lucky and unlucky alliances will balance out on average.
__________________
2005 Sacramento Winners with 245 and 766
2005 SVR Semi-Finalists with 840 and 668
2005 SVR Judges Award