I have used the WinTD pairing software for years, and in general I like it. One problem I have noticed, though, is that it sometimes seems to assign a full point bye to the wrong player. The general rule for assigning a full point bye, should one be necessary, is to give it to the lowest rated player in the lowest score group who has not already had a bye or forfeit win, and who has not committed to a bye later in the tournament. WinTD usually does this. However, in its Help File it states that it may sometimes assign the full point bye to a different player if by so doing teammate conflicts are reduced or color allocations significantly improved. Doing this can make things run a bit more smoothly, but the problem is that as I read them there is nothing in the Rules of Chess that allows you to do it. Section 28L is the one that deals with assigning full point byes, and nowhere does it mention anything about re-assigning the full point bye for these reasons. As I read the rules, FIRST you assign the full point bye, THEN you worry about teammate conflict or color allocation problems. If I am missing or misinterpreting something I would appreciate it if someone could enlighten me.
Rule 28L2 is atypically clear about how the bye is assigned. WinTD violates rule 28L2. I have had discussions with Mr. Doan about this, and his opinion is that WinTD does not violate rule 28L2.
There is some sentiment among members of the rules committee that the manner in which WinTD handles the bye is correct, or at least reasonable. I would note that the FIDE Dutch system of Swiss pairings allows the bye to be given to a player other than the lowest rated in the lowest score group to improve color allocation.
I placed an ADM on this year’s Delegates’ Meeting agenda to change the wording of rule 28L2 to allow what WinTD does. That motion was referred to the rules committee, so it will be taken up again at next year’s meeting. Until then, I would suggest the ugly workaround of assigning the full point bye explicitly before generating pairings for the section to force WinTD to comply with rule 28L2.
Thank you, Ken, for the quick reply. This problem most often rears its ugly head in the even numbered rounds when most players are due an equalizing color. Apparently not getting an alternating color isn’t a significant enough problem to trigger this deviation from Rule 28L2. It’s easy enough to do the work-around. You just have to remember to check each section as it’s paired to make sure the correct person has been assigned the full point bye, if there is one. If not, you delete pairings, assign the correct bye manually, and then re-pair. It would be nice not to have to do this, but it’s not the worst thing in the world.
Using WinTD I sometimes have to manually do 28L5 in four-round tournaments with unrated comprising the bottom score group. I do this by either explicitly assigning the bye or by explicitly giving the unrated full-point byes in rounds 11 and 12 (removing them before doing the prize calculations).
I take the attitude that 28L2 is very clear, and clearly very wrong. I was on the advisory committee for the 5th edition, and pointed out that there were several other situations where the rulebook didn’t include the caveats that everything be subject to the overall Swiss system rules. Apparently this one didn’t get fixed. Note that 28L2 (as written) allows for no exceptions. If your bottom score group has three people at 1/2, and A and B have already played, C still gets the bye and A and B get upfloated even though equal scores is #2 on the priority list, and 28L2 is not listed as one of the exceptions. 28L2 isn’t listed as one of the exceptions to any of the basic Swiss rules. Heck, if dropping C means that you can’t pair the rest of the section without dup’ing a pairing, 28L2, read literally, says tough, drop C. Does anyone seriously believe that’s what is intended—that this is the one decision in the pairing rules that is so vitally important that the basic Swiss rules are to be ignored?
WinTD treats this pretty much like any other downfloat, subject to the additional rule that any player who has or will have an unplayed game not get the bye.
USSE swisssys
+1
If I were directing with WinTD, I would certainly go ahead and let it do its “wrong” thing. And, to head off bullfeathers complaints from players who see the light as poorly as 28L2 does, I might even post a notice to this effect before round 1.
It makes absolutely no sense to be locked into giving the lowest-rated the bye, any more than it would to be locked into dropping the lowest-rated into the next score group. In fact, it’s exactly the same concept. You make transpositions to improve colors, or to improve pairings in other ways.
Bill Smythe
+2