Author: Richard Pijl
Date: 13:41:05 10/24/04
Go up one level in this thread
On October 24, 2004 at 16:23:38, Uri Blass wrote: >On October 24, 2004 at 15:48:17, Gian-Carlo Pascutto wrote: > >>On October 24, 2004 at 14:40:19, Uri Blass wrote: >> >>>On October 24, 2004 at 12:57:06, Michel Langeveld wrote: >>> >>>>On October 24, 2004 at 12:22:21, Volker Richey wrote: >>>> >>>>>Hi, >>>>> >>>>>Swiss Perfect calculates an other Berger value? >>>>> >>>>>where is the error? >>>>> >>>>>Volker >>>> >>>>This method was elected by the players: >>>> >>>>"In this tournament a special tie-breaker is in effect to compensate for the >>>>effect of Swiss-system paring in a tournament with a small number of >>>>participants. It applies to the players in the lead with an equal number of >>>>points(P). Only results against mutual opponents and the game among themselves >>>>count for SB. Buchholz (sum of opponent) points were discarded." >>>> >>>>Congrats for Vincent for this Victory! >>> >>>I do not understand the method. >>>What is the meaning of mutual opponents? >>> >>>I understand that games against Isichess was not counted when Isichess drew with >>>the king when the king drew with Diep so I do not think that it is correct to >>>use that method in the future because games against Isichess are not an obvious >>>easy win based on the results. >> >>Isichess did not play all 3 first place finishers, it's completely irrelevant >>how strong Isichess is. >> >>This method removes the problem that if you are unlucky (having to play against >>a weak opponent), this may cause you to lose the tournament because of a lower >>SOP/SB score. >> >>Actually, if this method hadn't been used, the tournament director would have >>basically appointed the winner, since the last round pairing had to be done >>manually. Nobody would have accepted that. >> >>-- >>GCP > >1)The tournament director could pick the winner only in case that he could be >100% sure about the results. > >I see no reason to be 100% sure(I remember that xinix beated Ruffian in previous >tournament so even weak programs can surprise in a good day). > >The only claim that you can say is that the tournament director could influence >the results and it is not fair. > >2)I see no reason that the pairing in the last round had to be done manually. Be assured that it was not the wish of the tournament director, but because the software couldn't handle this round. There were several rules that could not be resolved by the program. >3)Even in case that it had to be done manually it was possible to calculate the >possible pairing and choose a random pairing. Remember it was the 11th round of a 14-player tournament. Finding a pairing that did not give one side an unfair advantage was not easy. Richard. > >I understand that the ranking was not decided by which player played stronger >opponents and I think that it is better to decide by which players played >stronger opponents. Who plays stronger opponents is not an accomplishment of the program because that is largely determined by the swiss pairing program. Remember, we're talking about a tournament with almost as many rounds as players. >I can understand throwing very weak players out of that calculation and >assuming that they scored the same because if A and B are very weak then it is >not important if you beat A or beat B but I cannot agree to not using the >question which player played better opponents to decide unless the players >except very weak players are the same and in this case isichess is not a very >weak player(it drew in WCCC with Junior) so the condition that the players >except very weak player are the same does not hold. I'm sure that anyone that played a very weak opponent would have liked to play a stronger one instead if that would improve their chance of winning the tiebreak Richard.
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.