Author: Dave Gomboc
Date: 20:00:55 06/08/99
Go up one level in this thread
On June 08, 1999 at 18:21:08, Robert Hyatt wrote: >On June 08, 1999 at 13:51:13, Dave Gomboc wrote: > >>On June 08, 1999 at 12:44:04, José de Jesús García Ruvalcaba wrote: >> >>>On June 08, 1999 at 09:36:12, Dave Gomboc wrote: >>> >>>>On June 08, 1999 at 08:13:55, Robert Hyatt wrote: >>>> >>>>>On June 08, 1999 at 03:00:22, Bruce Moreland wrote: >>>>> >>>>>> >>>>>>On June 08, 1999 at 01:36:33, Dave Gomboc wrote: >>>>>> >>>>>>>The web page http://www.uni-paderborn.de/~wccc99/ reports that standard >>>>>>>(non-accelerated) pairings will be used, but from my conversation with some of >>>>>>>the ICCA executive, they do intend to use accelerated pairings (mainly because >>>>>>>it will force more games to be played between the strongest opponents.) >>>>>>> >>>>>>>So I am writing this so that everyone is not shocked when they get there. :-) >>>>>> >>>>>>I predict a big fight. >>>>>> >>>>>>bruce >>>>> >>>>> >>>>>I don't know who the TD is, but you can tell him for me that this is _stupid_. >>>>>All he has to do is ask _any_ legitimate TD and he'll discover that accelerated >>>>>pairings are _wrong_ when you have more than log2(players) rounds. And since >>>>>2^7 (7 rounds) is > number of players, this is useless... >>>>> >>>>>Some people never learn. Or they refuse to learn. :) >>>> >>>>I specifically gave this argument, Bob. But David Levy replied that they wanted >>>>to maximize the number of games between strong opponents, and that using >>>>accelerated pairings would do that. I went and researched this a bit, and >>>>here's what I found in the Chess Federation of Canada handbook (at >>>>"http://www.chess.ca/section6"): >>>> >>>>Accelerated Methods of Pairing Early Rounds (Variations) >>>> >>>>633. In a tournament where the players-to-rounds ratio exceeds the optimum >>>>(16:4, 32:5, 64:6, etc.), the chances of producing a clear winner are decreased. >>>> >>>>Accelerated pairings increase the frequency of meeting of the higher ranked >>>>players and are therefore also useful in longer tournaments where the winner is >>>>unlikely to have a perfect score. >>>> >>>>The effect of these variations decreases when the players are of about the same >>>>playing strength (as in a tournament divided into sections by playing strength). >>>> >>>>[after this, specific VARIATIONS 633.1 and 633.2 are discussed] >>>> >>>>They are using accelerated pairings to "increase the frequency of meetings of >>>>the higher ranked players" because that is "useful in longer tournaments where >>>>the winner is unlikely to have a perfect score." I thought about it, decided >>>>that this description fit the WCCC very well, and consequently was persuaded by >>>>David's argument. >>>> >>>>Dave >>> >>> What I most dislike is that nobody knows which are the strongest entries, so I >>>do not see how the accelerated pairings will help to match them more frequently. >>> I think one of the premises for accelerated pairings to work is to have a good >>>ranking of the players, like an established ratings list. But I remember >>>somebody said that in these tournaments the entries are ranked according to the >>>TD's guesses. I do not think that is a good ranking. >>>José. >> >>There's plenty of background material to rank the players on, including the >>result of previous tournaments and, for some entrants, the SSDF list. It's not >>as good as it would be in a human tournament, but it is acceptable. >> >>One time I played in the Quebec open: they use a different rating system in >>Quebec (FQE) than in the rest of Canada. So they took my 1927 CFC rating and >>subtracted 100 points, and paired me as an 1827. The young 2100+ FQE player who >>I beat left without resigning in person: I went to look for him, and found him >>crying in another room. How could he have known that I was a little bit better >>than a weak 'A' player? I was already underrated at 1927, and there apparently >>wasn't really a 100 point difference between the two rating systems. He went on >>to became Quebec's junior champion the next year. >> >>Upsets happen, but since a reasonable ranking can be made before the event, it >>is okay to use accelerated pairings. >> >>Dave > > >You only have to go back to 1997's event to see why this isn't a great idea. >By the time the final round gets there, all the top programs have _already_ >played. And the tournament is essentially over. you just move the final >(important) rounds up one level by doing this... and with < 32 teams, there >is _no_ reason. check the math on what happens if you have 16 weak programs, >and 16 strong programs... with 5, 6 and 7 rounds... with and without >accelerating the pairings... If the goal is to find the appropriate winner, then 1997 was excellent. Junior played many strong programs, and no one was questioning its ability by the end of the 11th round. It was very clear that Ban and Bushinsky had a very good program. We only have 7 rounds this time, so I think there will be more suspense this time around. 11 rounds for ~32 players might not have left any suspense even in a regular swiss, Bob. I know that it's not the standard practice for accelerated pairings to be used where I play chess unless the 16:4, 32:5, 64:6 entrants-to-rounds ratio is oversubscribed, but this is probably something that David Levy picked up in his internation chess playing experience (he is an IM after all, as I know you are aware of). But it isn't just an off-the-cuff decision, it is being made deliberately with the understanding of the factors involved. To me, the tournament will be either the same or slightly better because of it, so why not do it? I don't think it will lead to a disaster, though maybe some commercial entrants will dislike it because they could be knocked out of contention for first place in round one instead of round two (big deal). Dave
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.