So, it’s round 2 of a four-round Swiss. After round one, there are three players with 1.0. Here’s a summary for those three players and the top player in the 0.5 group:
Player A (2200) 1.0 W
Player B (2183) 1.0 B
Player C (2147) 1.0 W
Player D (2118) 0.5 B
We expected that player C would be the odd man, producing these pairings:
B vs. A
D vs. C
However, the computer program (SwissSys 9.01) did this:
D vs. A
B vs. C
The question is why. There are no problems at all with the pairings on the lower boards. None of the four players have “do not pair” requests. Player A does have a requested half-point bye for round 4 but I don’t see how that should make a difference. Does anyone have an idea?
I’m not an expert of the problem of 6 players in a 4-round Swiss, but there are problems, I believe, in the 3rd round where no legal pairings can be found. Perhaps the computer is attempting to solve that problem. That’s my only knowledge on this. I don’t know why that happens, but others will, hopefully, opine.
As usual, we’d have to see the complete crosstable (all scores, all ratings, all opponents, all colors) to have an intelligent discussion. Please furnish this.
As far as avoiding the 6-player trap, two points. First, both SwisSys and WinTD have fallen into this trap many times, so I doubt if there are any algorithms in either one to avoid this trap. Second, after one round, you have 3 players at 1.0 and one at 0.5, so how can there possibly be only six players? Wouldn’t there need to be 3 at 0.0 and one more at 0.5, to account for the opponents of the 4 players already listed? That would be eight players, at least.
EDIT: A third point regarding the 6-player trap: The problem occurs in round 4, not round 3. Whether the problem occurs in round 4 depends on the round 3 pairings. You can make any pairings you want in round 2. It’s round 3 where you have to be careful.
What Mr. Smythe says above is generally true. However, IMHO, Mr. Terrie’s original post gives all the information necessary for an analysis of the top score group’s round 2 pairings.
We know there are at least eight players in the event, as the four players in the OP each played a round 1 game (as indicated by their color history).
We know all the ratings, scores and color histories of the involved players.
We know none of the involved players has a requested bye in round 2.
We know none of the involved players have requested not to be paired with one another.
We know player D is the highest-rated player in the 0.5 score group.
Given this information, I cannot produce a scenario where SwissSys would make the pairings it did, unless the “Weikel oddman pairings” box was checked.
I was not the TD, so I don’t have all the colors, only for the first two rounds. Here is the wallchart pairing and color info for the first two rounds:
P.S. I know there must be a way to turn the above into a nicely formatted chart with all the values centered under the headers but I can’t figure out how to do it. Can someone explain the method?
I have encountered odd man pairing errors in Swiss Sys. That being said, I see you used version 9.01. The latest is 9.4. According to the bug report on the Swiss Sys website: “USCF oddman pairings tweaked. (9.3)”. Download the newer version and see if it reoccurs. If it does, then send the files .s1c and .s2a to the programmer. Where to send it is on the “contact” page. I have found that when I document errors, he does correct them.
I’ve created a new topic, Displaying crosstables in the forum, for discussion of how to display crosstables and other tabular information in the forum.