|
|
|
Topic: Annual Tournament -- looking for input
| |
|
|
Author
| Message |
|---|
|
|
I'm thinking what changes I'm going to make for the next annual tournament and would like to run an idea I had by you.
What I really want to know is if you find any holes in my proposal, please do not post other completely unrelated ideas, those have been discussed a lot in the past and I can always go back to read them when I want to make bigger changes.
So the problem with last year's annual tournament is that rounds take a very long time, and the next round can't begin until the last game of the current round is completed.
I want to keep the Swiss System modality for this year's tournament, so there will be a single game per round, same as with the current tournament.
The first part of my solution to accelerate the rounds is to implement faster time controls, and several good ideas have been suggested in these forums so we don't need to talk about that part.
But even with smarter times per move, there will always be a player taking a vacation that can delay a round considerabily. So here is my solution:
Rounds will have a "completion time" associated with them. When the current round reaches that time the tournament will advance to the next round. For the purposes of pairing the system will assume that all the games that are unfinished ended in a draw.
Note that those games from the old round will continue in parallel with the games from the new round, and players will get the score according to the real outcome, it is only for the pairing of the next round that the scoring for a draw will be used.
The last round will not have a completion time, that round will run until all games from that round and previous rounds are completed.
The benefit is that only the last round will have an undetermined length, all the previous rounds will last for a fixed time. The disadvantage, of course, is that pairings will not be as good.
Do you see any problems with this approach?
Miguel
|
|
|
No problem at all.
Good approach.
The minor problem about parings being not so good is not too much relevant (when you consider, for instance, that even in an ideal world, pairings will never be good).
The completion time should be well estimated. For example, if games will be 3 days/ply, an average game of 50 moves (100 plies) has the maximum of 300 days. But people use to play in 2 days, so 200 days will be the average.
With exception of vacations, I believe that we should limite the round to the maximum, 300 days.
(I am not sure why 300 days looks so long... Did I mistake anything?)
Cheers,
Beco.
|
|
|
It sounds ok - but it is perhaps a little bit unfortunate that it is possible to influence the pairings with some slow play in a winning position or some holidays at the right moment.
A slightly better pairing could make a difference.
But if a lot of players do that it will level itself out i guess.
|
|
|
Having a completion date for all rounds except the final round is very good to keep things moving. This method is used by several correspondence organizations.
However, when rounds are concluded in this manner the unfinished games are adjudicated by the correspondence organization. When we used snail mail in the past to make moves, each player had to submit his moves along with analysis to the Tournament Director for adjudication. Doing this is far better than assuming unfinished games are draws. Some games will be very lopsided, and so, easily adjudicated a win for one side.
I know that using this method places a bigger burden on you, Miguel, so I would strongly suggest you delegate this task to others willing and having the chess knowledge to make these determinations. You could have a panel of perhaps 3 strong players willing to do the adjudications, and it would be best that they NOT be playing in the event, so as not to favor one player over another in close games. Too, with strong players doing the adjudications, the need for participants to submit their own analysis would be unnecessary.
After all unfinished games have been adjudicated you would be able to make the best possible pairing for the next round.
mustang 
|
|
|
I should desagree with human interference. I am not comfortable with this in many ways, like mistakes or human-hours of work.
BUT...
I could agree with an automatic evaluation of the position by a free strong chess engine, that would simple give the point TEMPORARILY to the better position, to pairing the next round...
And after that, the game just continue, and the point is given to who really won.
Again, maybe it is better go give 1/2-1/2 to unclear position than 1-0 to pair some player, and after that correct the actual score to 0-1 (or vice-versa).
Cheers,
Beco.
|
|
|
In case of a draw due to time lapse, there will be a temporary score for both players (considering it is a draw) so the pairing will be a little off but I guess that's ok. Also, when do you start eliminating players while the round progresses? if there is an elimination prior the final round then the draw mechanism might not be a good idea because a player in a winning position gets deprived of his actual score that can mean himself advancing to the next round.
|
| 1 2 3 4 Next |
|
|
|
