Delta and Echo have drawn, and Charlie has beaten Foxtrot.
My question is how to pair the fourth round. Board 1 is easy.
Abel - Baker
My intention would be to pair the rest of this group as
Charlie - Delta
Foxtrot - Echo
This was a five round tournament, if that matters. In any event, the computer program paired this group as
Abel - Baker
Charlie - Echo
Delta - Foxtrot
Delta and Foxtrot were (understandably IMO) very upset with these pairings, but neither said anything to the TD before the round had started and there was time to change the pairings, and I’m not sure if the TD would have changed the pairings anyway.
My question is what happened here? Understand that these are only approximate ratings, and that I don’t know the exact version of the software, but I don’t get it. I like to have good colors as much as the next guy, but this seems to me so far over the line that I am curious as to what happened. Can someone please help me understand?
It looks to me like the pairing was incorrect. The lowest rated odd player in a group should play the highest rated player in the next lower group unless there’s a big reason not to. Maybe the program was concerned about color but that shouldn’t have taken precidence. What pairing software was used?
I agree with the assessment that the pairing was incorrect. Read Tim Just’s TD Corner this month and WildTommy’s analysis in a different thread for a related question.
The crux of these questions seem to be A.Can color balance take priority over score group? and B.How are computers defining “score group” when odd players are dropping from one group to the next?
The answer to question A should be “No” according to my 5th Edition. Playing within your score group (27A2) is given higher priority than colors (27A4&5). In Alex’s example, color seems to have been given priority over playing within a heterogenous score group composed of lowest odd 3.0/highest 2.5 not already played (27A1). Even if the ratings were close enough for a transposition/interchange (which they are not), this is irrelevant to justify the switch because transposition and interchange are defined within a score group.
Question B can only be answered by rigorous testing of a program or by the software author himself. Perhaps this and other discussions of pairing variances should include the software name and version. If we are going to rely upon these programs to do a lot of the work (and thinking) for us, they ought to be scrutinized.
What some pairing programs do is balance the colors against the “correct” pairings so that future rounds are easier to correct. I’ve only used WinTD and I know that it will do that, unless the rating difference of the players involved is greater than 200 points. However, you can change that setting as well as set other preferences that avoid interchanges of points, etc. that probably would have avoided this situation.
Having said that, even if a computer is being used to pair the tournament, the TD should still “check” those pairings and ensure that obvious situations like this are at least looked at, whether or not they choose to correct them.
I agree that the computer pairing is incorrect. 29D1 permits switching the opponent of the odd player (Charlie) to a lower ranking player in the next score group to improve colors. However, the switch must be within the 200-point rating difference for equalization, which in this case it clearly is not.
Variation 29E5h provides for equalizing colors without regard for rating difference; this is apparently the FIDE rule. Both SwissSys and WinTD have options to implement this, and this results in the computer pairing which Alex described.
For the record, I tried pairing the fourth round with the current version of SwissSys (6.084) with USCF defaults, and it came up with the “correct” pairings (A-B, C-D, F-E).
I agree with others that the pairing seems incorrect, but I disagree with some of the reasoning in a couple of the replies.
First, a caveat – we may not know the whole story. An important fact may have been conveniently omitted (such as, X has already played Y) or the ratings may not be exactly as stated.
But, disregarding those possibilities for the moment –
There is no issue of color taking priority over score in this example. Both the suggested pairing and the actual pairing involve:
3.0 vs 3.0
3.0 vs 2.5
2.5 vs 2.0
The pairing across score groups is FORCED by the fact that there is an odd number in the top group. The only question is, WHICH 3.0 should face WHICH 2.5, etc. Transpositions can be made – dropping a 3.0 other than the lowest, and/or bringing up a 2.5 other than the highest – for the usual reasons (color, etc).
The only reason the actual pairing appears illegal is that the transposition exceeded the 200-point limit for color equalization.
When dropping an odd player to play someone of a different score group, the USCF advocates pairing the lowest-rated odd player against the highest-rated player in the next score group with caveats:
-the remaining members of both affected score groups can be paired with each other.
-the odd player has not played all the members of the next lower group.
-color consequences are acceptable.
In 29D1b, it recommends making “switches” on either end of this drop pairing when the above caveats break down. It goes on to say that there is no limit to the rating difference in order to keep the score groups intact, but it also says “switches to correct colors should stay within the appropriate limits” apparently referring to the 80-point rule for alternation and the 200-point rule for equalization in evaluating transpositions and interchanges. I guess if the rating difference between Delta (2250) and Echo (1700) had been no more than 200 points, then such a “switch” would seem less improper.
What if Alex had said that this was the fourth and final round of this tournament? And what if the “correct” pairings Charlie-Delta, Foxtrot-Echo had been posted? Master Delta could complain that he got 3 blacks out of 4 games at the same time pointing out that his pre-tournament rival, Expert Foxtrot got 3 whites out of 4 games.
I finally was able to check the complete tournament crosstable against my own pairings that I got when I ran SwissSys on it, and I think that I discovered the problem. I think that the computer was set to FIDE defaults. They say that both equalization and alternation colors are 3000 points. Does FIDE really care so much about color? I’d hate to have this pairing happen in a FIDE rated event.
As far as I know, no one uses FIDE pairing rules for swisses in the U.S. This has nothing to do with FIDE-rating an event. Those rules are used primarily in European swisses.