The final round pairings for a recent tournament were pointed out to me and I was asked what I’d do. I neither came up with the same decision that two other TDs who were also asked to look this over did, nor the one that the Chief TD used. Also, those two decisions do not match.
Please pair the top three score groups:
A 1153 4.0 WBWB Has already played F, G.
B 1927 3.5 BWBW Has already played C, D.
C 1510 3.5 WBWB Has already played B.
D 1062 3.0 WBWB Has already played B.
E 801 3.0 WBWB
F 1041 3.0 BWBW Has already played A.
G 718 3.0 BWBW Has already played A.
H 622 3.0 WBWB
That seems rather simple. A plays B, C plays D, F plays H, E plays G. There’s no reason to change the upfloating 3.0 to fix color since it simply shifts the color problem down to the 3.0’s.
The natural pairings say A-B, C-D, F-G, E-H. But this creates color issues on the middle two pairings. Swapping D for F not only fixes the color issues for those two boards, it is also the smallest possible swap (in terms of ratings differential) anywhere in the pairings. So, IMO, I would pair A-B, C-F, D-G, E-H.
That’s what you get when you write an answer while on the way out the door. Upon further review, I agree with Jeff and Boyd. The natural pairings have two bad colors. One’s unavoidable. The only fix to the other that’s under the 80 point limit is to upfloat F to play C.
One color problem in the 3.0 group is unavoidable. The color problem in the 3.5 drop-down to the 3.0 group is avoidable with a 21 point transposition. A-B, C-F, D-G, E-H (higher ranked gets due color).
I wanted to give 24 hours after my post for anyone interested to work out the problem.
We were only given A-F, but of course the Chief TD had the whole section at his fingertips. A well-respected ANTD and a Local TD came up with B-A, C-D, and E-F.
I can’t see how the last pairing can be supported. I could only see the first pairing being supported if there were something weird about the lower, unstated score groups that required dropping G and H to have a legal pairing (it happens, and has happened to me).
How did those coming up with the first pairing justify giving B white and A black?
EDIT: The Chief TD appears to have taken a situation requiring one color mismatch and generated a pairing with three color mismatches and substantial departures from the natural pairing. In the absence of substantial, unstated justification (e.g., A and B having played each other before a re-entry, team competition without the plus-two rule, or maybe A and B being related–and if there were prizes involved, that’s just tough, IMO), I am constrained to state concerns . . . [edited] .
EDIT AGAIN: Substantial, previously unstated justification appears to have existed; see next page of thread. Statement of concerns withdrawn.
A 1153 4.0 WBWB Has already played F, G.
B 1927 3.5 BWBW Has already played C, D.
C 1510 3.5 WBWB Has already played B.
D 1062 3.0 WBWB Has already played B.
E 801 3.0 WBWB
F 1041 3.0 BWBW Has already played A.
G 718 3.0 BWBW Has already played A.
H 622 3.0 WBWB
There are five due white and three due black, so two due white will have to play each other.
A human TD going “top down” (rather than “looking ahead” enough) might, therefore, start off with
A-B
C-D
reasoning that the C-D pairing pairs two due white against each other, as is necessary somewhere. No reason to transpose just to move the bad color down to a lower score group, if it doesn’t reduce the total number of bad colors, right?
But now the fertilizer hits the fan. The remaining raw (“natural” – I hate that word) pairings
F-G
E-H
result in bad colors in a total of three pairings, not just one. Transposing G with H
H-F
E-G
is more than an 80-point swap, and transposing E with G (an “interchange”)
E-F
H-G
is also more than 80 points.
So back to the drawing board on the top two boards. If we upfloat F instead of D (less than 80 points), we have
A-B
C-F
(not for the purpose of pushing the bad color down to a lower group, but rather in the hope that it will improve colors overall) so now the bottom pairings become
I’m guessing that B-A is a typo for the first pairing and it is really A-B. Since they were only given A-F to look at (with G and H essentially hidden from them), the C-D color conflict is the only one over three boards and thus it would make sense to leave it alone.
If A and B are prohibited from playing each other for some valid reason then the second set of pairings, including the three color conflicts, makes sense. A plays C (the only other 3.5) even though there is a color conflict. B plays the top 3-1 that hasn’t already played B and it would be a 240-point transposition to avoid the color conflict. The remaining four players are paired normally with one more color conflict in their two games.
OK – first, I am going to pair as I think they should be, and then I’ll read everyone else’s responses.
First cut. Players not getting due color have asterisk:
A-B
C-D*
G*-F
E-H
So the question is, can we swap D with G or C with F to give the maximum number of players their due colors?
In both cases, the rating difference is too large (>80) to justify a switch. Also, this is the last round, so mucking up the pairings of future rounds by players’ not having their due colors this round is not a consideration.
I would leave them as they are above.
Now, I will post this an then see what you more experienced people have to say!
EDIT: Missed that H also should have White.
Still, honestly, I’m not sure that I’d bust my hump too much to resolve colors in the last round, as long as it’s not balance that is a problem. We cannot swap G with H or F with E, so leave it.
OK – I didn’t consider the possibility of swapping D with F, because that IN ITSELF did not fix anything, but does when you then switch the Board 3 colors of from G-F to F-G.
Eliciting further information from the OP, I learned the following.
First (minor point that doesn’t make any difference in the pairings), WinTD was used.
Second (major point), C had a rating more than 80 points higher than B in a different system from the one used to initially list the ratings.
That implies that WinTD was set to use the greater of the two ratings when pairing. In that case, a strict adherence to the 80-point limit for transpositions would have given the chief TD’s pairings. After all is said and done it looks quite likely that the chief TD may have actually done pairings exactly by the book.
This thread is one reason why I dislike the idea of people using the forums to call out named TDs, since even a very knowledgeable person may overlook mentioning what turns out to be critical information (and they don’t come much more knowledgeable than Alex).
This qualifies as substantial, previously unstated justification, and I have withdrawn my calling out of the unnamed TD on the previous page. I would not have called out a named TD at all–there are venues for that; agreed this is not one.
