The Rulebook states “Pairing numbers are assigned by lot.” Take 4 slips of paper and number them 1 thru 4, then fold them or lay them face down and let the players draw for their number.
Generally the WinTD assigned numbers are essentially a lot drawing that is done totally inside the computer. If you want to have slips of paper then put in dummy names (maybe a through h for an 8-player round robin), pair the round robin, find the player who alternates w-b throughout (player number 4 in an 8-player round robin), use that as a basis for determining how the computer assigned 1 through 8, change the letters to those numbers, and then you are ready for the slips of paper drawing.
The last time I did this, SwissSys asked for the pairing numbers of the players. It also gave an option to sort randomly, but that was not required. Note that quad rules require you to do it by rating. Also, the RR tables are set for a reason. I wouldn’t mess with them. As I said above, if you are going to tamper, you’ll need to announce the changes in all advance publicity, and also, there is no reason to use SwissSys to do a round robin.
As an added disincentive, note that SwissSys doesn’t give the option to send RR tables to USCF. You’ll have to convert to a swiss before it will let you make the .dbf files. If you use the Tournament Upload on TD/Affiliate directly, you can make the tournament look like a RR, and it probably takes less time.
Alex Relyea
As I read the rulebook (30G, p. 172) it says that the top 4 players by rating are the first quad, etc. However, then rule 30A takes effect, and the players are assigned pairing numbers within their quad by lot, so that the top rated player in any quad is not guaranteed to be pairing #1 (and get White twice.)
If I had a distribution of players that didn’t break well using the top-down assignment method, I’d probably talk to the players and see if they’d be willing to split it up differently. I’ve done this in events and I don’t think the players have ever said no to such a request.
5 masters in a quad would be a pairings disaster. With masters you can’t very well use cross-round pairings (the games would take too long) so you’d end up giving full-point byes to 3 of the 5 players at some point during the tournament.
If you can get unanimous consent among the 5, you could do something creative. For example, if the announced time control was G/60, you could instead make it G/90 and have two double rounds (4 games per player total). In each double round, each player would play two opponents simultaneously. Pairings in the first double-round could be:
1 - 2 - 3 - 4 - 5 - 1
where each player plays white against the player listed on his left, and black against the player listed on his right. In the second double-round you could have:
1 - 3 - 5 - 2 - 4 - 1
The table arrangement for this is interesting. Arrange five 6-foot by 30-inch tables like the blades on a five-blade ceiling fan, with the table corners touching, so that the[i] short /i table sides form a small regular pentagon in the center. Seat each player in one of the V’s between the fan blades, having him play white against the player on his left, black against the player on his right. Face all clocks toward the center.
The arbiter stands in the small central pentagon at all times, where he can view all the boards and all the clocks, and is not permitted to leave until all the games are over. THIS LAST SENTENCE IS A JOKE – no phone calls please.
Try it, you’ll like it.
Bill Smythe
Sadly, our room is too small and our tables are too wide; but we do something nearly as creative by playing Olympic Quads; where we deliberately pair the first set of two quads with a very strong, strong, average and weak players; then the second set pairs the top two winners from each quad and bottom two player in the second round of quads. Lots of fun with the intention of mixing the players up. Played at our Tuesday night club at G/10 and sometimes, like once a month, rated.
Running quads with SwissSys is very simple. Search the help for “quad tournaments.”
The basic idea is this: Create a tournament and a dummy section where you will register all the players for your quads. Then, in the tournament setup dialog, you can choose “quad setup” under the “tools” menu. That tool will automatically copy players from the dummy section into quads by rating order. You will not be asked to assign random pairing numbers, since SwissSys understands these are quads. You will be asked if you want to randomly toss for color in the last round (the correct answer is “yes”).
Not all RR tables (of which there are many once you’re at 6 or more players) are appropriate for chess since we have relatively strict rules regarding color. (I assume the same might be true of home/away scheduling, though schedules there often seem to be designed to favor two home followed by two away). WinTD doesn’t do the 1 plays 4, 3 then 2, but instead does 1 play 3, 2 and then 4, which is part of a more general algorithm for generating RR tables of any size that avoid color problems (no one gets 3 in a row or ever gets to +2 out of equalization). The two are identical other than “labeling”. There are only two distinct RR tables (i.e. differ in anything other than labels) for 4 players, depending upon whether you reverse all colors in round two or repeat colors for two of the players in 1 and 2 (rather than repeating colors in 2 and 3).
It’s certainly possible to use paper cross tables for the players to enter the data than transcribe it from those. However, if there are more than a few sections, I’m not sure I would recommend that (assuming you need to get it into a computer eventually). When they did the KC SuperNationals blitz using preliminary octos, it took me until 4 a.m. to get the paper information input—enough people are unfamiliar with RR cross tables that some of those were indecipherable. With WinTD you can generate the full RR schedule for all rounds, and print standard board by board pairing sheets. More familiar for the players and the TD.
That’s pretty much the same with WinTD. You create a template section, dump all the players into it and do “Section-Divide Up” (it can be any size, not just quads). It automatically generates sections as needed by appending a number or letter to the base section.
Forget the 4 player Quad, rather run Octogons. 5-8 Player 3-SS events.
As for Quads an alternative to the top down rating approach is to go in reverse. The Highest player in the Quad is player #4, second highest rated is Player #3, etc. to fill out the quad. What this does is gives the 2 lowest rated players White twice and the 2 Highest rated players Black twice.
Of course the random draw is the proper (& FIDE) way to determine player position in any Round Robin event, even one of only 4 players.
Larry S. Cohen
I don’t think the masters would be too thrilled about having byes in a three round event either. There is no really good wy to handle this mix of players. Fortunately, this extreme would never happen in the real world.
Please read the fourth paragraph of 30G. It clearly states that players are supposed to be listed in rating order in the special case of quads. Note that the last round is supposed to be a random draw for colors, so #1 does not necessarily get White twice.
To Mr. Doan above, of course I don’t mean just any RR tables, I mean the ones that are in the USCF rule book.
To Mr. Cohen, you may not think that it is fair, but it is the rule. Please see my previous statements to Mr. Smith.
Alex Relyea
It is not the rule. It’s “preferred”, but that preference is probably ignored more often than not in favor of random assignments.
With the scheme
1-4, 2-3
3-1, 4-2
1-2, 3-4
– only a single coin toss (by the computer or TD) is necessary. Heads, number the players in rating order. Tails, number the players in reverse rating order. This way, the two top-rated players always meet in the final round, no matter what. And it’s 50-50 which of the two gets the extra white.
A further refinement would be to alternate the scheme down the quad sections. If the top quad is in rating order, the second quad can be in reverse order, the third in rating order, etc.
Bill Smythe
We draw by lots as the Rulebook states and do not toss for color for the last round since the random initial setup has already decided that for us. In fact we sometimes play the first 4 rounds of a Hex which gives equal number of blacks and whites to each player, so a toss would be unfair for the last round in this situation.
I tried this but it assigns random pairing numbers in each quad. Is there a way to have the pairing numbers by rating?
What version of SwissSys are you using?
Alex Relyea
The most up-to-date version, version 8.894
Well, you do not have to use pairing software. We use a paper pairing sheet and list the players by lot or any way we choose and, yes, we could list them in rating order; then below on the same sheet we have each of the 3 rounds shown with each player’s name and a place to record the score. When we have it rated, we do that manually, usually as a 3 round 4 player Swiss. USCF asks how many players and how many rounds and builds a XTable for you =just key in each player ID# and post W, L or D and submit to USCF online for rating. Also it could be posted as a round robin.
I tried this with the same version, and it worked fine. Note that you have to choose “Quad” in the last dialog box. Can you list the steps you are going through here?
Alex Relyea