Can anyone help me? I was looking for something on my hard drive that I can’t seem to locate. Probably because I don’t have a copy any longer.
In GM Walter Browne’s old WBCA he had a RR system of pairings that was interesting. It involved the players more than the TD. Each player was give a pairing number. That number was odd or even. Players had to know their paring number. Players did not have to wait for pairings/next round to find their opponents and what color they played. The TD had all the players together for round one. They picked an opponent. When their game was done they reported the results and then found another opponent, any new opponent. When opponent’s met they indicated if their pairing number was odd or even. That info told the players what color to essay. I seem to recall that of if one player was odd numbered and the other was even numbered that indicated who got white and who got black (and that technique is what I can’t locate). If both players had odd pairing numbers, or they both had even pairing numbers, then that also indicated which player got white and which one got to play black.
It was a nice system for moving things along without having to wait for a round to end before you started your next game. All that seems to matter is that you played everyone, reported your results, and used the odd pairing number-even pairing number system of color allocation.
Anyone remember what the color allocation rules were for that kind of system?
I don’t, but I’ve done enough 8-player round robins to know by heart that in the USCF tables the rule is:
Odd vs Even (except #8) the better seed gets white
Odd vs Odd or Even vs Even (except #8) the weaker seed gets white
1-4 get white vs 8
5-7 get black vs 8
So
1 white vs 2, 4, 6, 8
2 white vs 3, 5, 7, 8
3 white vs 1, 4, 6, 8
4 white vs 2, 5, 7, 8
5 white vs 1, 3, 6
6 white vs 2, 4, 7
7 white vs 1, 3, 5
8 white vs 5, 6, 7
For forty or more years, the Pittsburgh Chess Club ran Saturday night 5 minute round robins with a similar scheme to the one that Jeff outlined. One of the members, Fred Sorenson, created a special chart and the rules to use for these tournaments. Players were seeded by number on the chart in order of their entry, not by rating, IIRC. The next time I drop by there, I will see if I can find a copy. It allowed the players who finished games early to play without sitting around. The strong players often jockeyed around to play weaker foes early while they warmed up. Then they would play their rivals later in the final rounds. It worked very efficiently and let Fred, who ran these tournaments, go home early. If I find the chart, I think we might try it out at our club to see if the players like the method.
The algorithm is simple – odd vs even, smaller number has white. Both odd or both even, larger number has white. For example, 3 has white vs 8, while 7 has white vs 5.
“Number” means pairing number, not rating. Pairing numbers should be assigned randomly, not by rating; otherwise the top player would always have white against the second player.
The No Exit did it this way. So did the club on Lunt Avenue on Friday nights. No need to wait for “official” round times. Late-arriving players are handled easily, too – just add their names to the list.
Sometimes it would be desirable for the TD to step in and request two players who were standing around without opponents, and who had not yet played each other, to pair themselves against each other then and there. Occasionally there are players who try to avoid playing each other until later in the evening.
I have also played another scheme for WBCA (and unrated) RR blitz tournaments. This eliminates the problems of forgetting your pairing number, what odd and even mean, and whether seed “1” is higher or lower than seed “2”.
Set up 5 boards (or whatever number you need) in a row, alternating colors. Seat one player on the end as anchor. This player will sit at the same place for the entire tournament. At the end of each round, the other 9 players move one seat clockwise. The anchor switches colors for every round, but the other four boards stay the same color orientation they were originally. [Note: The TD may have to help with the initial seating arrangement.]
Aside from eliminating the need for 3rd grade math, this scheme works well if you have one mobility impaired player. snicker
This model shows how the players rotate around the “long table” and alternate colors. My idea had players meeting opponents in ascending order. On the link, click on “Next Rd” to see the rotation. round robin model.
The systems described herein work swimmingly well unless and until someone withdraws before the midway point and color switches are necessary. The TD has to stop the event, figure out where each remaining player’s color distribution stands (discounting games against the withdrawn player, which are no longer scored), and hope the games remaining allow for an appropriate combination of switches.
For this reason, strict adherence to the Crenshaw-Berger tables, which make provisions for a withdrawals, is a clear best practice in my book. Accelerated play is stil possible, albeit only in the first n-3 lines of the table, where n is the number of rounds.
I also have a set of color reversal tables: build2.htm.
This is my page for creating round robin tables and has links to a “pull numbers out of a hat” page and other things I’ve found about about RR tournaments: rrpair.htm. A more complete listing is in a prior thread.
Tried to promote this at a US Open Business Meeting in Dallas a few years ago, but didn’t make much progress.
For a weekly (typically Friday night) round robin, usually at a chess club or coffeehouse, I prefer the free-for-all method I described above. As you finish each game, you look for another available player. The conversation typically goes, “Would you like to play our game now?” “Sure. I’m number 7. What’s your number?” “I’m number 3. So you have the white pieces.”
Such events can easily handle 20 (or more) players, and can run start to finish with virtually no TD intervention. The TD can easily play in the tournament, too. Except for handling an occasional dispute, the TD’s only duties are signing people up as they arrive, and handing out prizes at the end.
Drop-outs don’t matter. The player’s remaining opponents simply get forfeit wins. In an informal event like this, it’s not worth the trouble to deal with color reversals. It’s no big deal if somebody gets two more blacks than whites.
Late arrivals are easily handled, too. Just add their names to the list and keep playing. (There should be a limit on how late a player could be, so you don’t end up with the entire tournament finishing except for several games all involving this same player.)
Some of those tables don’t quite follow the rule. Games involving the highest-numbered player sometimes are shown with the “wrong” color.
It’s worth pointing out that the phrases “larger number” and “smaller number” (or even “larger pairing number” and “smaller pairing number”) are preferable to “higher player” and “lower player”. The latter could mean higher- vs lower-rated, which is likely to be just the opposite from larger- vs smaller-numbered.
I would agree with you only in the case of an important big-money blitz round-robin held in conjunction with an important big-money regular-rated Swiss. Here you would need a little more formality. Otherwise, it’s just not worth the extra trouble, especially when it comes to withdrawals or late arrivals.
No problem when it works, but with too many late entries or players who have to miss a few scheduled rounds, a log jam is possible when several opponents want a piece of the same player.
The tables assume a formal schedule. If the “Ghost” (highest-numbered player) used the same rule for colors as everyone else, that player would only alternate colors on every 2nd game (perhaps starting in the 2nd round). For an informal procedure, use the next larger sized table if you use one at all.