| |
BGonline.org Forums
potential swiss rules modification
Posted By: Bill In Response To: potential swiss rules modification (Chuck Bower)
Date: Tuesday, 31 March 2015, at 5:45 p.m.
Chuck,
I see a particular problem with this idea. It may have some value too, so please don't take this as a dismissal. This is a quick reaction, not a thorough analysis.
In general, when doing the draw for a particular round, we think in terms of cohorts. Literally, we take everyone with the best record and do a random draw to pair them. If there is an odd number, we randomly select a person to "play down" against someone with the next-best record. Then we randomize the next cohort and select one person to "play up." (It doesn't matter whether that makes the next cohort even or odd. We just repeat the process from best record to worst.) In the record keeping, we track who has played up or down and use that as a secondary filter. We try to make sure no one in a cohort gets to play down (or has to play up) a second time unless everyone else in that cohort has already done so once.
Using your modification, the players that receive byes end up as small in-between cohorts. In half of the cases, they will play up and in the other half, they will play down.
Let's take the example of a tournament with 41 players. After the first round, you will have 20 [1-0], 1 [0-0] and 20 [0-1]. The player that received the first round bye will be paired up against an [0-1] player - essentially getting an easier draw because of the bye. Of course, if you take a tournament with 43 players, the cohort sizes come out differently: 21 [1-0], 1 [0-0] and 21 [0-1]. This case ends up being the same as our current method, where the bye player is always placed in the higher cohort.
Is this quirk okay? Or a problem?
| |
BGonline.org Forums is maintained by Stick with WebBBS 5.12.