I was thinking about doing a continuous pairing format for a K-3 u500 section which I expect will have about 10-15 players in it.
Most of the games get over in 10-20 minutes, but there can be some that take a lot longer - so why not just pair kids as soon as a reasonable pairing is available instead of the kids waiting around?
Awards are based on Score/Games Played
I see little downside to doing it this way (well a busy TD and the chess pairing software won’t be of much help - but easy enough with a queue sheet, pairing sheet, and cross table - I need to be careful about choosing tie breaks, that is easy to calculate by hand)
I like the fact that an odd number of players is not that bad - kids have potential to play a lot more games (more accurate rating?) - I don’t have to round them up or try to guess when the next round might start - more combos of results possible so should be less ties.
Anyone ever do anything like this?
Pitfalls?
Suggestions?
I’m not so sure this is a good idea. What is a “reasonable pairing”? By random chance, you may get different strength opponents and more importantly, the top players may not face each other if their games are finishing a different times. Another factor might be that some players will use the scholar’s mate repeatedly and play a lot more games than others. If the top players do face each other and draw, would it be fair that the player who goes 5.5-0.5 lose out to a player that goes 6.5-0.5 simply because one’s opponents lost more quickly?
While handling the odd player may work better, a cross round pairing can accomplish the same thing.
How much time do you have available? At my son’s middle school I did a non-rated event with 9 rounds in about 4 hours (with a half-hour lunch break included in that time). If your numbers are small enough then you might be able to do a round robin. If you have a large room such as the normal sized cafeteria we had for 26 players then you might have a game or two of Uno set up at one end with chess at the other. That would give the kids something to do while waiting for the round to finish.
Taking an extreme example, you may have two kids capable of playing a good game and eleven who are somewhat random. If the two strong kids play in the first round and draw then their game may take all of the available time, during which time another kid wins ten times with scholar’s mate.
I know you meant it humorously, but I’ll answer as if you were being serious.
You’d only deduct the five minutes if you were using a delay clock. At the K-3 level with new players, a delay clock will often actually extend the length of games due to the players playing clock instead of chess and having both players making totally random moves in the delay time rather than possibly having one player making moves with a plan.
You may not have delay clocks available. In the middle school tournament where I did nine rounds in 3.5 hours of play (116 games - one player left early resulting in the final round being the only one with a bye awarded), only one game ever had a clock, and that was an analog clock that was put on after 20 minutes into the round in a one-sided position.
This looks like a variation on the ASAP tournments that I have seen run. But instead of starting the round earlier, you appear to want to pair players immediately when done so that they can play more games. As long as this just a “rating tournament” and not one for any prize or title, then fine. But note that better players tend to take more time, think longer over their moves, and play higher quality games. Isn’t this a value you want to instill in your young players? Having to wait around for a new round should provide a teachable moment to make them think about how they can play better.
This type of tournament is ripe for abuse. On another thread was a discussion of a TD who used unusual pairings in order to increase his own rating. Pairing players relatively randomly gets more games played but is an odd way to find a winner of an event or produce valid ratings. Could a TD just let a bunch of kids play casual games without rules or clocks and then send these in for rating? Sure, but it is an abuse of process. There should be more than a tip of the hat to pairing rules for an event to be USCF rateable.
Actually, it is not immediately obvious to me that playing random opponents is an odd way of producing valid ratings. I might think quite the opposite, in fact.
As I said in the other thread, the ratings algorithm doesn’t care why two players were paired, as long as it doesn’t lead to abusive situations. Assuming the pairings are reasonably competitive, I don’t see an issue with rating it.
The advantage a Swiss has is that as it progresses, players should be playing opponents closer to their own strength.
The biggest problem I can see with a continuous pairing system is that the ‘fast’ players might essentially wind up in one pairing pool and the ‘slow’ players in another, and that’s more of an administrative/social element. Other than that, it’s not that different from what happen in unrated play in many clubs, two players who are available at roughly the same time sit down and play each other.
Thanks for all the comments.
My idea for reasonable pairing would mean to pair same/close score groups, so there could be a wait for the “better” pairing.
I did run a few basic simulations with a 13 player section and it surprised me that waits were short and even had no bad luck getting colors to even out with only one time getting a BWB v BWB in what would be their last game. It is possible to force some bad situations, but those would more likely with a smaller number of players.
I’m doing this K-3 U500 section so it could be JTP eligible - the rest of of the tournament is quads at G/30 (G/25 d5!) and we are pretty tight with getting out on time if we get a game in each round going the full hour.
I think we all go crazy trying to get the kids to slow down their play speed.
There’s the old story of a floor TD noticing a kid constantly looking at other things that were happening in the tournament room not paying too much attention to his own game. Being concerned, the TD asked him if there was a problem and he replied that his coach told him if wanted to get better at chess he should not make his moves too quickly.
The suggested system sounds similar to that used by Dr. Milton Hanauer to run his junior high school tournaments in NYC in the 1960s. I directed the high school section for him for a few years which had one game each Saturday morning, and witnessed some odd situations in the junior high, which he paired and which played something like 10 games over 5 mornings.
Hanauer didn’t believe in using clocks, and what would happen was that if a game took long enough, the players would play only one game that morning, but if it ended quickly, he would try to give that player a second game or sometimes even a third game. He didn’t pay a lot of attention to score, and didn’t use ratings at all, and frequently would make a pairing like a 3-0 or 4-0 vs. a 1-1 or 2-2 because someone needed a game. Sometimes a player could get up to 6-0 or 7-0 before Hanauer decided that he needed to face strong opponents, and he was more careful with his pairings in the last few rounds.
A fairly common situation was that one of the top players would win quickly, and Hanauer would say, “Wait here, I’ll try to find you another opponent.” That player would then sit near Hanauer waiting for a pairing, and some other players figured out that if their game was the next to end and Hanauer noticed that, they might obtain an unwelcome opponent. So some players in this situation tried to sneak out when Hanauer wasn’t looking, or kept their game going longer than necessary until the feared opponent had been paired and was no longer a threat. On at least one occasion, a kid tried to leave while a strong player was waiting to be paired, Hanauer said “Wait, you have to play another game,” and the player said, “Can’t do it, I have a dentist appointment.” And that gambit was successful.
Bill’s post brings up another possible weakness in the continuous pairing concept, someone who is in contention to earning a prize and winning his current game may not want to finish his current game and play another round because that might keep him from winning a prize. That might encourage dawdling, which is different from playing slower in order to improve one’s play.
I think that if the kids in a K-3 u500 section knew how to manipulate sitting on a won game to their advantage then they most likely would not be be in the K-3 u500 section. They are generally oblivious to who gets prizes beyond the more games you win the more likely you will get a trophy - I have had one kid who was rated around 400 that went 3-0 in a quad ask me if he will get a trophy and all the time I get kids that go 1-2 in a quad ask me if they will get a trophy.
In this case if they withdraw they forfeit any prize consideration. Since this is normally a quad tournament I post “if you can’t commit to playing all games then please do not enter.” I will be stronger in my wording.
What about this idea. You have two sections. One is the normal tournament that is paired ASAP when the last game is finished. The other is a waiting section. In this section players are paired randomly as they become available after their normal tournament game is finished. When the normal tournament is ready for a new round, all in progress extra games are canceled.
Mike
How would you like to be in the middle of a game you are winning and have the TD come up to you and say “Sorry, this game is being canceled so you can start a new game with a different opponent”?
That sounds like a way to get kids AND parents upset at you.
All it takes is one kid to be told he’s ahead in the standings (by a parent, usually.)
And I think there’s a big difference between taking time to think longer and thus moving more slowly and just dawdling between moves. I’ve seen the latter even among players rated under 200.
I decided to go ahead and use the continuous pairings and things worked out perfectly!
Everybody got 3 blacks and 3 whites and wound up playing the same opponents.
I feel they must have set a records for most times not noticing a king was in check and for stalemating their opponents when they were up a tremendous amount of material (one time the kid even quoted to me that they should always check to avoid stalemate)