1: A(2318) 1 point, due White in round 2
2: B(2316) 1 point, due Black
3: C(2289) 1 point, due neither (forfeit win in round 1)
4: D(2131) 1 point, due Black
5: E(2003) 1 point, due White
6: F(1501) 0.5 points, due neither (half-point bye in round 1)
All other players had 0 points.
Several players expected round 2 pairings to be: A-D, C-B, E-F, believing that the lowest player in a score group should always be the one who gets dropped unless colors would be improved by dropping the next-lowest, which is not the case here.
Instead SwissSys produced: A-C, E-B, F-D, apparently because it keeps the “natural” pairing on board 1, which it seems to consider a higher priority than dropping the lowest player.
Noob TD answer here, correct me if I’m wrong: The first step is to divide the score group in half, and the lower half is larger (AB/CDE). That gives you the AC pairing. BD is a color conflict, so you get BE. That leaves D to drop to the next score group down.
Notice that to select D as the odd player and E as B’s opponent requires a transposition of D and E, a rating difference of 128. That is within the 200-point limit for transpositions made to improve color equalization. However, if one starts with the natural pairings (A vs. C and B vs. D) and applies a transposition of C and D, the rating difference is just two points (the difference in the ratings of A and B). Then, the pairings become A-D, C-B, E-F.
SwissSys’s pairings are not incorrect, but it does appear preferable to treat E as the odd player from the 1.0 score group (since the remaining players in the 1.0 score group can be paired with no color conflict).
I know from personal experience that Thad Suits is very good about answering such questions. I would encourage you to e-mail him the .S2A file from that section and ask him about the pairings.
Some TDs, and many would-be TDs (including your “several players”) are far too reluctant to make transpositions involving players from different score groups.
If you had presented your example without ratings –
– telling us only that the five players in the 1-point group are listed in ratings order – then there would be no inherent advantage or disadvantage in either of the suggested transpositions. Switching 3 with 4 (resulting in 1-4, 2-3, 5-6) would be neither better nor worse than switching 4 with 5 (resulting in 1-3, 2-5, 4-6).
But once you throw the ratings back in, it becomes obvious that the first transposition is better. It amounts to a 2-point switch (the lesser difference of 2318-2316 and 2289-2131) versus a 128-point switch (2131-2003).
So your “several players” were right, but for the wrong reason. Change the top player’s rating from 2318 to 2490, and suddenly the second transposition becomes better.
That’s called top-down pairings – start with the top pairing, and don’t worry about what comes later until you get there. Look-ahead pairings are better – examine the entire score group (or in this case, the top two score groups), making whichever transposition does the least total violence to rating differences.
Thanks for the replies. I exchanged a few emails with the programmer. He indicated that when there is a choice of switches, it tends to push the switches down to the bottom of the score group, because he used to receive complaints from TDs when it switched at the top. But he did sound willing to fix this if the consensus would prefer it. Judging from the replies, that seems to be the case.
Speaking only for myself, a possible reason for this may be that the rulebook doesn’t specifically mention whether transpositions involving the odd player to be dropped from a score group should be considered equally as any other transposition with the same rating difference. Maybe the next edition could include an example to illustrate this?
Indeed, the rulebook could be a little more explicit. But, in the discussion of the odd player (29D), it does hint at this in a couple of places:
[29D1b] " … In deciding whether to make a switch of either the odd player or the opponent, … switches to correct colors should stay within the appropriate limits (29E5)" [80-point rule and 200-point rule] …
– and, about half a page later:
" … It is acceptable to pair the [dropped] player against a somewhat lower-rated player to equalize or alternate colors, but only within the rules for transposition as explained in 29E5 … "