In SwissSys there are a couple of settings that can be important to round robin sections (ie Quads).
They are “Coin toss for color in the last round” and “Random lot numbers for round robin”. You can find these on the “Rules for Pairing” screen.
Tom Langland
The USCF default “Random lot numbers for round robin” was on!
Thanks everyone for your comments and help
I stated some options. I said, nor indicated that I thought the random draw system or the rating top-down system was fair. In my 7+ years of posting (hard as it may be to believe) this is the first time someone has put words in my mouth. I start to understand all those complaints from other posters.
There is always leeway for the TD/organizer provided there is proper notification of the variation from the usual rules. I can site at least one instance where there was a problem due to a question of proper notification, if you want me to do so. Once again options (or alternatives) are not necessarily an opinion on something.
Larry S. Cohen
My apologies. I misremembered “fair” for “proper”.
Alex Relyea
Why should we care more about what masters think about running quads than about what C players think?
My experience has been that if you imagine the worst possible set of circumstances, the real world will (eventually) come up with something even worse.
I’m curious. If you want to run a 3 round tournament with 4 players, is there any particular advantage to running it as a round robin instead of as a swiss? A 3 round swiss with 4 players results in every player playing every other player exactly once, just as in a round robin.
If running it as a Swiss, then 1 (by rating) plays 3 and 2 plays 4 in the first round. That probably means 1 plays 2 and 3 plays 4 in the second round, so 1 plays 4 and 2 plays 3 in the third round.
If running it as a RR by rating rather than by random lot, then 1 and 2 play in the final round, so the two highest rated players meet in the final round, rather than having the highest and lowest rated players meet in the final round.
Is having the top two players meet in the final round important?
Also, the higher rated players might not win in round 1, and that could affect the pairing for the remaining rounds.
One of the benefits to having pairings pre-determined (even if assigned by lot rather than by rating) is that players may have the option to start their games early. That can’t happen if you’re going to use the results of the first round to determine second and third round pairings.
So, in the long run it comes down to this: What’s important to you as an organizer and to your players?
Mike gave some of the advantages of a Round Robin. Another is that if it is a multi-week event (probably one game per week) then a round robin gives more scheduling flexibility and deferral options in case somebody cannot make one particular week.
If two more players enter after round one is paired then it is much easier to get them into a 3-round Swiss than change it over to a 5-round Round Robin.
That would be true with more than 4 players, but for 4 players isn’t a round robin as constrained in that regard as a swiss? If you have one game still being played from round k, neither of the players who have finished their round k game can start their round k+1 game, because their round k+1 opponents are the people still playing the round k game.
It seems to always make the lowest section the swiss if you have a number of players that isn’t divisible by 4. Is there a setting so it makes the swiss whichever group makes all the sections the most even in terms of rating?
Take some club tournament from the MSA, pull out the first 22 players and demonstrate how your idea would work in practice to make them into quads and one six player section. “Most even” is a vague, non-operational concept. With 22 players, there are five possible locations for inserting the hex; come up with a measure of evenness that can be computed to determine which is the “best.” Then we can discuss whether doing it that way even makes sense.
One thing to remember is that you would expect roughly 1/4 of the time, you will have a multiple of 4, 1/2 of the time you’ll have either 5 or 7 extra and 1/4 of the time you’ll have 6 extra. It’s only the last case where shifting the location of the Swiss from the bottom is likely to work reasonably—with an odd number of extras, the bottom section is the best place for byes because you’re more likely to be able to cross-round. Change Mike Nolan’s scenario to six (rather than five) masters, and I (and probably a lot of other TD’s) would probably suggest that the “Swiss” be made of the top six. If you wanted to carve the six out of someplace in the middle, you would need a really good reason for it.
Have you considered having 3 rounds of a Hex?
Then 3 hexes plus one quad would solve the 22 player turnout.
Regardless of whether the division is quads, hexes, octets or whatever else, it doesn’t solve the “most even” problem in all cases. Since the format is advertised in advance, it could only be switched if every player in attendance agreed to do so (and this assumes we’re talking about a small club/local event that doesn’t award USCF GP points).
My guess is that at least one of the most negatively affected players (for example, if proposing on site to switch from quad to hex, player #5 in the rating-sorted overall player list) will object.
In my experience, when there are clusters of players with big ratings gaps between them, if the TD gathers the players together before pairing the first round and shows them the turnout and discusses options, the players will almost always accept a suggestion for non-traditional groupings.
Few players enjoy being the only high rated player in with a bunch of much lower rated players or the sole low rated player in with a bunch of high rated players. The whole point to quads is to try to put players in competitive situations, eg, pair them against similarly rated opponents.
I’m not sure that hexing the top six and running quads below would require consent from anyone other than the top six since everyone else is getting (what one assumes) are competitive quads as advertised. Assuming you have prizes, you also need to have figured out how you might adjust the prize fund for a group of six rather than four.
22 was my “for instance” not an actual tournament. And how does that “solve” the problem? Instead of 4 quads and 1 hex, you have 3 hexes and 1 quad and you still have to decide where to put the quad.
I think it is incumbent upon the organizer to know his market, know his players, and do what is ‘right’ when the turnout doesn’t match expectations.
The reason to run quads as opposed to a Swiss is to try to get players in groups of similar strength. If you can’t do that, why are you running quads?
It’s player #5 who is most likely to complain in all cases. (This has happened to me once before.) And players 7/8 in the second hex, for example, may be looking for higher-rated competition and grumble about being “dumped” into a lower quad (though, since they are in a quad as advertised, they don’t have as much to complain about).
My only point was that discussing the options and securing unanimous agreement to change advertised conditions are both mandatory.
I agree with both sentences. The problem is that, as long as you use any even number of players for grouping, the condition in your first sentence can’t be avoided if the distribution provides odd numbers.
If a TD is running these events regularly, and sees that the normal “fault lines” among the players indicates a different format would be superior for his purposes, the TD should make those changes, IMO.
What is the advantage of running “3 rounds of a Hex” over a three round Swiss?
Alex Relyea