I don’t understand that at all (in fact I think it’s complete drivel). How are you better with two players with WBB and BWW rather than WBW and BWB? In the first case, the colors in the next round are basically forced to avoid 3 in a row, while the second you at least could give the first player W and the second B. Not optimal, but at least you have more flexibility.
For a small tournament (or section) the difference can be huge.
An extreme example is the 6-player event with 4 or 5 rounds. In that case, if all the colors alternate in both rounds 2 and 3, there will be no pairings at all in round 4.
Almost as extreme is 8 players. In that case, if all colors alternate in rounds 2 and 3, there will be only one set of pairings in round 4 that will equalize colors – and that pairing set will likely pair 2’s against 0’s, or something equally horrid.
And as you go up the scale a bit to 10, 12, or 14 players, things can still get tight. By making the colors work too well in the early rounds, you are essentially splitting the players into two camps – those who started with white, and those who started with black – and you are making only inter-camp pairings, no intra-camp pairings. Eventually, you’ll run out of suitable inter-camp pairings, and you’ll have to make intra-camp pairings, i.e. pairings with both players due the same color.
If the number of players is small enough so that numerous bad colors are eventually inevitable, I would rather have the problems in the odd-numbered rounds (alternation) rather than the even-numbered rounds (equalization).
Bill Smythe
Which is specific to it being six players (thus 3 on each color schedule).
Since we pair to keep score groups together and proper color is down at the bottom of the desiderata, who cares? In a section like that, the colors will almost 100% of the time start to be “wrong” by the time you’re at round 3 and definitely by 4. To recommend that people intentionally mess up the colors hoping (and that’s all it is) that it will have salutary benefits in future rounds is wrong-headed. And, BTW, with 8 players, if you get the colors “right” in the first 4 rounds, you can pair not just 5, but 5, 6 and 7 without a problem, which is completely the opposite of the situation with 6 players.
There are systematic ways to avoid getting boxed in when the number of players is small relative to the number of rounds (such as hybrid RR-Swiss or decelerated pairings). Intentionally messing with the colors isn’t one of them.
What’s always intrigued me about talking with other TDs is the number of differing philosophies as to correct pairing procedure.
My belief has always been that I should try to get the current round as correct as possible, without worrying about future rounds. This seems simpler from both the practical and philosophical points of view. More importantly, I’ve found it easier to explain my pairings to players when pairing this way than when I’ve tried to use ideas such as the “look-ahead” method.
I readily acknowledge the “Small Tournament Effect” Mr. Smythe has discussed on numerous occasions in the Forums. However, I do share (what I interpret as being) Mr. Doan’s aversion to relying heavily on projected results to make current pairings, solely for the benefit of future pairings having improved color allocation. IMHO, one should make the current pairings as well as one can, and let the chips fall where they may for future pairings. YMMV.
Six players, four rounds is a very specific case where it’s easy to get trapped into being unable to pair round four without repeating a pairing, and planning ahead (such as with hybrid pairings for at least the first two rounds) may make sense. Eight players with five rounds, ten players with six you would expect to take care of themselves at least in terms of getting pairings without duplications. The advantage of either hybrid or decelerated pairings (at least for the early rounds) is that they tend to delay some of the matchups of the highest rated players, so you don’t get “pick a victim” pairings in the late rounds. But, instead, you have “pick a victim” pairings early. That’s unavoidable—the bottom-ranked player will per force play more than one top half player no matter how badly he’s doing. The only alternative is to dup a pairing, which, in practice, may be the best way to handle it. (I wouldn’t do that unless forced, without consent of the players involved, but given the alternative to the players of playing way up in points and ratings, I suspect that most of the time that consent wouldn’t be hard to obtain).
With 2-4 players you can always find pairings that avoid rematches as long as the number of rounds is not greater than the potential number of opponents (counting BYE as an opponent). With 5-6 you have to worry when doing the round three pairings to make sure that you have potential round four pairings that do not include rematches (such as if rounds 1-3 have players 1-3 playing only players 4-6). With 7-8 I think you are always safe through round seven (if the first four rounds saw 1-4 play 5-8 then you finish with what is essentially two different quads). With 9-10 you have to worry about the round five pairings (to avoid things like 1-5 playing only 6-10 in the first five rounds). With 11-12 you have to worry about the round nine pairings (to avoid things like 1-6 having played only 7-12 in rounds 1-6 and 1-3 playing only 4-6 in rounds 7-9) and maybe some earlier pairings.
I generally let the pairings go through normally and then at the critical rounds I check to see if I need to make an override.