This happened at a Game Day event I played today, wondering if it's kosher. It was 16 players, 4 rounds, round 3 of 4. There was a draw in round 2, no draws in round 1 as far as I'm aware. This caused a double-pair-up/down situation, where 2 different 1-1 players were paired against 2 different 0-2 players. What information do I need to investigate this bug?

Did the 1-1 players play each other in either of the first two rounds? If so its totally possible.

It's a lower-case “b” bug, not behavior that causes major problems, but one that probably shouldn't've happened under ideal scenarios. At an event, I probably wouldn't worry about when this happens, unless it has a strong possibility of affecting significant prizing. Additionally, WotC tends to say something along the lines of “the correct pairings are what WER pairs”, i.e. TOs shouldn't be going around and modifying what WER (made by WotC, thus what WotC) decides are the pairings.

<explanation>

Swiss pairings in software work by taking players and sorting them in some order, then pairing them up in that sorting order. (most rounds in WotC software start with players in standings order) If players in this initial pairing have already played each other, then the software needs to be able to fix it, both if there are alternatives at the same point total, as well as if there aren't.

The problem with the “aren't” scenario is that this is not a problem easily processed by computers, so it's difficult to ensure that pairings are processed in an efficient manner: proving that there isn't a valid pairing without pairing down means testing every possibility that could. This means that software will tend to bend the rules relatively quickly, so that it can get a valid pairing without waiting forever on the algorithm.

</explanation>

The end result of all this can sometimes mean what you saw at Game Day: the algorithm reached a point that it balanced expediency over proving it couldn't find a valid pairing, and thus made two pair-downs.

Originally posted by Patrick Nelson:

Yes, they did. However, this being round 3 of a 4-round Swiss tournament, there was probably (I didn't do a full analysis) a different configuration that would have produced a more-valid pairing. I chalked it up at the time to “draws make things weird”, but I was wondering if there was a way it could be investigated.