Author: Robert Hyatt
Date: 20:58:15 11/10/02
Go up one level in this thread
On November 10, 2002 at 22:25:44, Ricardo Gibert wrote: >On November 10, 2002 at 21:53:12, Bob Durrett wrote: > >>On November 10, 2002 at 21:50:20, Robert Hyatt wrote: >> >>>On November 10, 2002 at 11:07:17, Ricardo Gibert wrote: >>> >>>>On November 10, 2002 at 07:40:33, José de Jesús García Ruvalcaba wrote: >>>> >>>>>On November 09, 2002 at 15:55:48, Ricardo Gibert wrote: >>>>> >>>>>> >>>>>>9 rounds are too many for the number of participants playing in a swiss. You end >>>>>>up with weird pairing toward the end of the tournament where a high ranked >>>>>>program gets paired with some program ranked in the middle, because the high >>>>>>ranked program has played all the programs ranked higher already. This makes the >>>>>>last round a bit of an anticlimax. Yuck! >>>>>> >>>>>>It makes more sense to play 5 rounds of 2 games each for a total of 10 games. 5 >>>>>>rounds are perfect for the number of contestants entered and now the pairings >>>>>>will always make some sense particularly because color allocation will not be a >>>>>>problem. You will also get more games played between the high ranked programs >>>>>>which will lend more interest to the contest. >>>>>> >>>>>>To me the choice is clear cut. 5 rounds of 2 games each PLEASE! >>>>> >>>>>Last time there were like 50 participants (I do not remember the actual number), >>>>>so five rounds are not enough. >>>> >>>>5 rounds at 2 games per round makes 10 games. How can 9 games be enough and 10 >>>>games not be enough? Do you really expect a big tie for first place, because >>>>"five rounds are not enough?" >>> >>>I assume you are talking about playing an opponent twice, once black and once >>>white? That means you only have 5 different opponents in an event with >>>50 players. Not enough to get a valid winner in a swiss event, which needs >>>at _least_ log2(entries) rounds to really find the "best" player... >> >>Assuming no draws? >> >>Bob D. > > >Thinking it over I think they're right. With 5 rounds, the 2 top programs might >not get paired with each other. It would need to be 6 rounds. > >With the rule changes being considered, there might be a smaller turn out, but >the idea does not seem to be popular in any case. I think it would be a good idea for a longer event. IE with 50 players, and say 10 rounds of 2 games each (20 games total) would be a pretty revealing event.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.