Color Question

Running Chess Tournaments
1

I’m currently running my club championship which is 6 round Swiss over the course of 9 Mondays. There are 26 players in it. We have make up dates in case someone can’t make it. #1 ended out playing #2 in the 3rd round because of games that had to be made up paired as draws. They ended out drawing with each other.

Now I have the following color situation at the top going into round 5.

#1 2277 wbwb 3.5
#2 2271 bwbw 3.5
#3 2231 wbwb 3.5
#4 2179 bwbw 2.5
#5 2151 bwbw 3.0

#1 has played #2 & 5 and is scheduled to play #2 in round 5

2 has played #1 and is scheduled play #5 in round 5.

The Swiss Sys paired for round 5:

White Black
#1 vs #3
#5 vs #2

#1 2277 wbwbw
#2 2271 bwbwb
#3 2231 wbwbb
#4 2179 bwbwb
#5 2151 bwbww

Regardless of who wins between #1 and #3, #3 will play the winner of #5 vs #2. If #2 wins then he would be forced into a 4th black against #3 who will have played 2 blacks in a row.

I would rather reverse colors on board 1, so that in round 6 regardless of whether it’s 2 vs 3 or 2 vs 5 both players would end out with even colors. I figure that since #1 will have played 2,3,5 it would be very likely that he would play #4 who would be due black.

My question: Is this a valid application of 29E6a look ahead method? Am I justified in switching the colors that Swiss Sys assigned?

2

That doesn’t look like something covered in 29E6a.
There is still 29E4b that covers alternating priorities. That rule, however, is limited to players within the same score group (the 3.0 is not in the 3.5 scoregroup), governs cases where all of the affected players have the same due color (1&3 are due white but 2&5 are due black - though the variation was designed for situations where those different type of conflicts would have been handled by simply changing the pairings so that there were no longer both types of conflicts) and reverses the colors on the second of each pair of boards (i.e. 1-3 and 2-5 instead of your desired 3-1 and 5-2).

If 5 wins, then 5 would play 3. Changing from 1-3 5-2 to 3-1 5-2 would mean that 3 would finish with four whites out of six games. Changing to 1-3 2-5 would mean that 5 would finish with four blacks out of six. In both cases, 2 and 3 would have a 50/50 split if 2 wins or draws against 5.

So it looks like your options are:

  1. keep the SwissSys pairings and hope that either 5 beats 2 or that 2 accepts that the Swiss pairings gave 4 blacks in 6 games.

  2. use 29E4b, expand it to include the 3.0-1.0 player in the 3.5-0.5 scoregroup the player is paired into, expand it to include alternations for both white-white and black-black conflicts, reverse the listed order of alternations, and hope that nobody pushes you to show your justification in the rulebook.

  3. use 29E4b, expand it to include the 3.0-1.0 player, expand it to include alternations for both white-white and black-black conflicts, keep the listed order of alternations, and hope that 5 does not beat 2.

  4. ignore 29E4b, and do your change without support in the rulebook.

  5. get 1 and 3 to simply agree to play with 3 having white.

Personally, I’d go with 1, but 5 seems plausible (kind of similar to correcting a game after two players played the wrong color).

3

I’m not proposing change the colors on board two between #2 & #5. If I leave those colors as is and just switch 1 and 3 then I will get #2 playing White against #3 who will be due black and both players will have 3 of each color, or if #5 wins then he would get Black against #3 who would in this case end out with a 4th white. But I still have the problem of #4 ending out with 4 blacks if he plays #1 in round 6.

I could have avoided this whole problem if I had managed to upset #3 last night. My colors would have worked out perfectly. :stuck_out_tongue:

4

I realized that you didn’t propose to change the 5-2 colors, but that is the board that 29E4b says would be changed if you used it. That is why I included it as one of the options (and still kept your original idea as another options).
Seeing as the change could still leave 4 with four blacks out of six, I’d just stay with the SwissSys pairings.

5

I’m pretty sure “look ahead” in 29E6a was never intended to mean “look ahead to possible pairings in the next round” !! :slight_smile:

Rather, it simply means, for example, that when deciding on the top pairing(s) in a score group, it is wise first to check for possible consequences further down in the group.

Suppose, going into round 3, we have a 6-player score group as follows:

  1. wb
  2. bw
  3. wb
  4. wb
  5. bw
  6. wb

The “raw” (untransposed) pairings would be:

  1. (wb) vs 4. (wb) – (1. gets white)
  2. (bw) vs 5. (bw) – (5. gets white)
  3. (wb) vs 6. (wb) – (3. gets white)

The inexperienced TD, probably using pairing cards (and possibly even making the beginner’s mistake of writing down each pairing as it is made), would immediately notice the color problem in the first pairing, and make a switch between the first and second pairings:

  1. (wb) vs 5. (bw) – (1. gets white)
  2. (bw) vs 4. (wb) – (4. gets white)
  3. (wb) vs 6. (wb) – (3. gets white)

This conveniently solves the color problem in the second pairing as well. But now the TD notices that 3. has already played 6. Being too proud (or finding it too inconvenient) to undo his top-down pairing, he now makes an additional transposition, this time between the second and third pairings, and ends up with:

  1. (wb) vs 5. (bw) – (1. gets white)
  2. (bw) vs 6. (wb) – (6. gets white)
  3. (wb) vs 4. (wb) – (3. gets white)

If, instead, the TD had done a little looking ahead, he would have come up with a simpler solution – simply switch between the second and third pairings to begin with:

  1. (wb) vs 4. (wb) – (1. gets white)
  2. (bw) vs 6. (wb) – (6. gets white)
  3. (wb) vs 5. (bw) – (3. gets white)

In general I don’t like such “anticipatory color assignments”. There is simply too much chance that something will go wrong. All it would take would be a single upset, or even a single draw.

In the case of your tournament, you probably have several 2-pointers, some of whom will soon have 2.5 or 3.0 and may end up being paired against the 2.5 and/or 3.0 you already have. Thus, I don’t see how it’s possible to accurately predict anything.

A better idea would to be aware of the “small tournament effect”. If the colors work too well in the early rounds, you will be dividing the players into two camps, those who started with white and those who started with black, and you will have made primarily inter-camp pairings. With a small tournament, suitable inter-camp pairings will become scarce in the later rounds, forcing you to resort to intra-camp pairings, where there will be color problems.

The moral: In a small tournament, don’t make the colors work too well. One way to do this is to transpose only to equalize colors, and not merely to alternate colors. In a pairing program you can change the alternation limit from 80 to 0, while leaving the equalization limit unchanged at 200.

Is 26 players a small tournament? I usually think of a small tournament as consisting of 20 players or fewer. But with six rounds, it’s not too much of a stretch to consider even a 30-player event to be “small”.

So I agree with Jeff – go with the Swis-Sys pairings. Also keep in mind that, in a 6-round event, it is virtually impossible to avoid 4 blacks or 4 whites for at least a few players.

Bill Smythe