Round Robin (Crenshaw-Berger) Tables

Looking at the round robin tables from the 5th edition of the Official Rules, I was wondering why the individual boards are placed in the order they were in. If you try to follow a given player, e.g. 1, in a large RR tournament, he would start (in this case) at the right end of any line, then be next to the left end of the next line for the 2nd round. Following just odd numbered rounds or just even numbered will reveal a pattern.

A RR tournament model has one fixed person at a stationary board (the “Ghost” as Richard A. DeVenezia calls him) while everyone else (an odd number of players) rotates one position along a long table where color assignments are permanent for the side of any one board. This table shows how the players rotate and no one ever moves more than one board when going from one round to another.

I don’t know why the table in the Rulebook was formatted that way, but I can tell you from experience that doing it the way you suggest will really annoy some of the players. (“That board is better/worse than the others! Why should he/I always/never get that board!”)

With the exception of The Ghost who stays on the leftmost board in both tables, everyone plays on each board twice, except for playing The Ghost who might not be present if there are an odd number of players.

Another point you are overlooking is spectator interest. Placing a game between the tail-enders, or between two players you can be certain will make a short draw, on the center board most visible to the spectators is a poor idea. (Yes, I know this somewhat contradicts my previous point. The interests or players and organizers are not always identical.) It’s an art, not a science.

I changed the crenshaw page to give the option of displaying pairings just like in the Official Rules, and discovered a few misprints. On page 298, round 1, 7 plays 14. On 299, round 1 , 15 has white against 1.

This link shows how the round robin works for 10 players. Numbering the boards from left to right, they are displayed 1, 3, 5, 4, 2 in Official Rules.

The same algorithm wasn’t used for the offical tables for 4 or 6 players. This is what I have for 4 players:

[code]3 or 4 players

1: 2- 4 3- 1
2: 4- 3 1- 2
3: 1- 4 2- 3
[/code] … and this for six:

5 or 6 players 1: 3- 6 4- 2 5- 1 2: 6- 5 1- 4 2- 3 3: 2- 6 3- 1 4- 5 4: 6- 4 5- 3 1- 2 5: 1- 6 2- 5 3- 4

I have made a few changes in crenshaw.htm, the main one being a “Min. Moves” option. Normally, as someone goes from one table to another in a RR tournament they will move from an odd to even board or vice versa. Since currently the even and odd boards are displayed in a group, a player will flop from one end of the table to the other. This tries to correct the problem. Usually each player will stay at one “board” for two rounds.

The default of the radio buttons is how the tables are displayed in Official Rules.

Edited change: the name of the option has been changed from “No Flop” to “Min Moves”.

The latest change in crenshaw.htm is an option to reverse all colors. This is necessary to allow a single late player into a RR which already has an even number of players. Click here for an example.

*Edited change. Flipping the colors also comes in handy for the 2nd half of a round robin.

I just added a page to generate random pairing numbers for round robins here. It is linked with the crenshaw.htm page.

I have a new page to show a side by side comparison of the three methods of displaying round robin pairings. These tables have ascending schedules like the old Berger tables (B-2, W-3, B-4, etc) instead of the current descending schedules Crenshaw introduced.
The dropout reversal tables have been adapted for this but since those tables are in the copyrighted USCF Rules of Chess, I have only shown them to a few USCF officials.
What I call “board number” order perhaps should be “internal board order” since whatever order the pairings for any given round are given will most likely be the order for board numbers. In the new “Minimum Moves” option, each player appears to stay at one location for two rounds, alternating colors.

The “crenshaw.htm” link shown in earlier posts has been modified to allow a double round robin.

When running a double round robin, there are two people who would have the same color three times in a row. Trying to swap the order of the next to last round in the 1st half with the one before in causes even more people to have 3-peats.

What about playing the 2nd half backwards? The first round of the 2nd half is an immediate rematch of the previous rounds with colors reversed, knocking out any potential 3-peat. Rule 30F says players or teams will have the reverse colors when they play again but it says nothing about the order.

In the following table, ignore the 2nd half round numbers, just go down line by line.

[code] 13 or 14 players
Rd: Pairings:
1: 1-14 8-7 13-2 9-6 12-3 10-5 11-4
2: 14-8 2-1 7-9 3-13 6-10 4-12 5-11
3: 2-14 9-8 1-3 10-7 13-4 11-6 12-5
4: 14-9 3-2 8-10 4-1 7-11 5-13 6-12
5: 3-14 10-9 2-4 11-8 1-5 12-7 13-6
6: 14-10 4-3 9-11 5-2 8-12 6-1 7-13
7: 4-14 11-10 3-5 12-9 2-6 13-8 1-7
8: 14-11 5-4 10-12 6-3 9-13 7-2 8-1
9: 5-14 12-11 4-6 13-10 3-7 1-9 2-8
10: 14-12 6-5 11-13 7-4 10-1 8-3 9-2
11: 6-14 13-12 5-7 1-11 4-8 2-10 3-9
12: 14-13 7-6 12-1 8-5 11-2 9-4 10-3
13: 7-14 1-13 6-8 2-12 5-9 3-11 4-10

 13 or 14 players reversed colors

Rd: Pairings:
13: 14-7 13-1 8-6 12-2 9-5 11-3 10-4
12: 13-14 6-7 1-12 5-8 2-11 4-9 3-10
11: 14-6 12-13 7-5 11-1 8-4 10-2 9-3
10: 12-14 5-6 13-11 4-7 1-10 3-8 2-9
9: 14-5 11-12 6-4 10-13 7-3 9-1 8-2
8: 11-14 4-5 12-10 3-6 13-9 2-7 1-8
7: 14-4 10-11 5-3 9-12 6-2 8-13 7-1
6: 10-14 3-4 11-9 2-5 12-8 1-6 13-7
5: 14-3 9-10 4-2 8-11 5-1 7-12 6-13
4: 9-14 2-3 10-8 1-4 11-7 13-5 12-6
3: 14-2 8-9 3-1 7-10 4-13 6-11 5-12
2: 8-14 1-2 9-7 13-3 10-6 12-4 11-5
1: 14-1 7-8 2-13 6-9 3-12 5-10 4-11


              15 or 16 players

Rd: Pairings:
1: 1-16 9-8 15-2 10-7 14-3 11-6 13-4 12-5
2: 16-9 2-1 8-10 3-15 7-11 4-14 6-12 5-13
3: 2-16 10-9 1-3 11-8 15-4 12-7 14-5 13-6
4: 16-10 3-2 9-11 4-1 8-12 5-15 7-13 6-14
5: 3-16 11-10 2-4 12-9 1-5 13-8 15-6 14-7
6: 16-11 4-3 10-12 5-2 9-13 6-1 8-14 7-15
7: 4-16 12-11 3-5 13-10 2-6 14-9 1-7 15-8
8: 16-12 5-4 11-13 6-3 10-14 7-2 9-15 8-1
9: 5-16 13-12 4-6 14-11 3-7 15-10 2-8 1-9
10: 16-13 6-5 12-14 7-4 11-15 8-3 10-1 9-2
11: 6-16 14-13 5-7 15-12 4-8 1-11 3-9 2-10
12: 16-14 7-6 13-15 8-5 12-1 9-4 11-2 10-3
13: 7-16 15-14 6-8 1-13 5-9 2-12 4-10 3-11
14: 16-15 8-7 14-1 9-6 13-2 10-5 12-3 11-4
15: 8-16 1-15 7-9 2-14 6-10 3-13 5-11 4-12

   15 or 16 players reversed colors

Rd: Pairings:
15: 16-8 15-1 9-7 14-2 10-6 13-3 11-5 12-4
14: 15-16 7-8 1-14 6-9 2-13 5-10 3-12 4-11
13: 16-7 14-15 8-6 13-1 9-5 12-2 10-4 11-3
12: 14-16 6-7 15-13 5-8 1-12 4-9 2-11 3-10
11: 16-6 13-14 7-5 12-15 8-4 11-1 9-3 10-2
10: 13-16 5-6 14-12 4-7 15-11 3-8 1-10 2-9
9: 16-5 12-13 6-4 11-14 7-3 10-15 8-2 9-1
8: 12-16 4-5 13-11 3-6 14-10 2-7 15-9 1-8
7: 16-4 11-12 5-3 10-13 6-2 9-14 7-1 8-15
6: 11-16 3-4 12-10 2-5 13-9 1-6 14-8 15-7
5: 16-3 10-11 4-2 9-12 5-1 8-13 6-15 7-14
4: 10-16 2-3 11-9 1-4 12-8 15-5 13-7 14-6
3: 16-2 9-10 3-1 8-11 4-15 7-12 5-14 6-13
2: 9-16 1-2 10-8 15-3 11-7 14-4 12-6 13-5
1: 16-1 8-9 2-15 7-10 3-14 6-11 4-13 5-12

[/code]

How much of an issue is this? It’s not a Swiss, so everybody would get the same number of whites and blacks.

I have played quite a few blitz tournaments under this model. It is an easy way to keep track of alternating colors. Being disabled, I usually had the honor of being assigned the “ghost” spot.

Another common method used at blitz tournaments is: odd-odd or even-even → higher seeded gets white and odd-even → higher seeded gets black. The players only need to remember their pairing number.

Michael Aigner

How about a “pivot” or an “anchor”?
Surely there is more substance to you than a ghost has! :smiley:

We would do this at club a lot for a RR speed tournament.

I corrected the link in my inner quote to show how players rotate–the original link creates the actual tables.
The color rule in the Berger table appears to be vice versa, but either one works. The same color rule shouldn’t apply to the Ghost since he would routinely have the same color twice in a row. In a double round robin, the color rule would be reversed for everyone in the 2nd half unless every pairing was a two game match.
Pairing numbers should be assigned randomly, seeded seems to imply the pairing numbers are based on ratings like a swiss, otherwise the color assignments between specific opponents would remain constant if the same people played every week.

Of course, what’s best is what works for you and your club and it looks like y’all have found it already.

When I first started working on this over 40 years ago, I called him the “Bye Guy” since he existed to give the guy with the bye “something to do”. I think DeVenezia called him the Ghost since this player doesn’t always exist. His term certainly works much better than mine. :slight_smile: Any word works if your audience knows what you are talking about.

There is a new page which creates a narrative to let your players create their own schedule in a round robin. This is what it does for 7 players:

For a sample with an even number, this is the narrative for 10:

I have had to move my website since I can no longer access my site on bellsouth. Click on the globe —>
for my new one. My pairing generator is now rrpair.htm and I am reducing the support for the Crenshaw tables.

The address of my round robin generator is now rrpair.htm and it now generates the default Crenshaw-Berger tables. It has links to other aids like how to splice in a latecomer and models to show how a round robin works and how to derive a schedule if there is no rulebook handy.