View Full Version : Setting up a round robin tournament

Spiny Norman

08-09-2005, 11:10 AM

Our club championship starts in a few weeks. There will be 8 players, 7 rounds, round-robin format (not Swiss). The likely makeup of that player list is:

2 players rated ~2000-2100

2 players rated ~1700

3 players rated ~1350-1450

1 player unrated (estimated strength ~1400)

I have a question or two regarding the setup of the draw in Swiss Perfect:

1) Do we enter in all the players and their local (ACF) rating and let Swiss Perfect manage everything from there; or

2) Do we get each player to "draw a number" from a hat for their position in the list of players, so that things are truly randomised and its the "luck of the draw"?

If we go with #1, does Swiss Perfect ask whether the top-ranked player should have Black or White in their first game, like it does in a Swiss format?

If so, should we toss a coin for that (publicly) so that there can be no allegations (or private feelings) of any possible bias?

3) As players, what would YOU prefer to see in terms of tournament setup. Does it matter? What if you get hit with playing the top-ranked player with Black for the third time in a row in recent tournaments, or something like that?

Garvinator

08-09-2005, 11:59 AM

1) Do we enter in all the players and their local (ACF) rating and let Swiss Perfect manage everything from there; or

2) Do we get each player to "draw a number" from a hat for their position in the list of players, so that things are truly randomised and its the "luck of the draw"?

Either way is fine. Normally most arbiters/organisers will use option 2. I consider one of the big advantages of option 2 is that you will have all the players in the tournament together at one time before the start of the tournament. That is rare indeed and this will then give your club a chance to make any announcements etc before players actually sit down at the board to play.

If we go with #1, does Swiss Perfect ask whether the top-ranked player should have Black or White in their first game, like it does in a Swiss format?

Sp will give you two options, but they arent do with colour. It will give you a choice of how you want the players to be paired. Either by standard or rutsch. You can choose either, or you can use the berger pairing tables, located here: http://members.trump.net.au/taschess/berger.htm. Berger pairing tables are most commonly used.

Note though that if you are going to use the board numbers that sp gives you, then one player will always be playing on the same board. Might be an idea to change that around a little.

3) As players, what would YOU prefer to see in terms of tournament setup. Does it matter? What if you get hit with playing the top-ranked player with Black for the third time in a row in recent tournaments, or something like that?[/QUOTE] As a player, I do prefer the second method, but again, that is so all the announcements can be made before sitting down to play.

Trent Parker

08-09-2005, 11:59 AM

I reckon Drawing numbers out of a hat would be fairer. If i remember correctly the first five numbers get a 4 white 3 black split with the Berger tables. Do you want all of your top players receiving more whites ;)

Kerry Stead

08-09-2005, 12:07 PM

Our club championship starts in a few weeks. There will be 8 players, 7 rounds, round-robin format (not Swiss). The likely makeup of that player list is:

2 players rated ~2000-2100

2 players rated ~1700

3 players rated ~1350-1450

1 player unrated (estimated strength ~1400)

I have a question or two regarding the setup of the draw in Swiss Perfect:

1) Do we enter in all the players and their local (ACF) rating and let Swiss Perfect manage everything from there; or

2) Do we get each player to "draw a number" from a hat for their position in the list of players, so that things are truly randomised and its the "luck of the draw"?

If we go with #1, does Swiss Perfect ask whether the top-ranked player should have Black or White in their first game, like it does in a Swiss format?

If so, should we toss a coin for that (publicly) so that there can be no allegations (or private feelings) of any possible bias?

3) As players, what would YOU prefer to see in terms of tournament setup. Does it matter? What if you get hit with playing the top-ranked player with Black for the third time in a row in recent tournaments, or something like that?

Without a doubt, option 2.

Swiss Perfect doesn't actually randomise anything for a round robin draw - it just uses the numbers (effctively the order of name entry) as pairing numbers and puts them into a standard Berger table. If you did it by rating order, then you will get 2 rounds that are contests, and all other rounds will be mismatches ....

So draw lots, use these numbers as the order of entering players names into Swiss Perfect ... and THEN let Swiss Perfect do its thing.

Ian Rout

08-09-2005, 01:24 PM

You use option 2 (random draw). Option 1 has two weaknesses

1. (minor) Players with pairing number 1 and 2 meet in the second round in the Berger tables (I think that's the name) used by SP.

2. (major) Four players get an extra White, because there are an odd number of rounds. These are the players with pairing numbers 1 to 4 (or the top half for other numbers of participants), so Option 1 gives these automatically to the higher rated players rather than everyone having an equal chance of getting them.

Kevin Bonham

08-09-2005, 01:34 PM

If we go with #1, does Swiss Perfect ask whether the top-ranked player should have Black or White in their first game, like it does in a Swiss format?

No. It will just award the top player in the list white for round 1 and some players will get +1 or -1 colour splits on the basis of ranking rather than merit. It's worse than that because not only does the top seed get the +1 colour balance, they automatically get it in the form of white against the second seed in round 1. It is definitely better to randomise.

3) As players, what would YOU prefer to see in terms of tournament setup. Does it matter? What if you get hit with playing the top-ranked player with Black for the third time in a row in recent tournaments, or something like that?

That has sometimes come up as an issue in our club's round-robin system where players simply write their names on a piece of paper and then who has what colour vs who is determined by order of entry. We were going to make an escape clause that a player who had had the same colour vs the same opponent 3 times in a row could request that that colour be reversed next time but people lost interest in the idea as there were bigger fish to fry.

antichrist

08-09-2005, 02:12 PM

For choice of colours try my trick of each round one player on each board puts 2 pieces behind his back and the other players picks a hand - worked like a charm for a few rounds until the heavyweights squirmed.

I dare you!

Spiny Norman

08-09-2005, 06:40 PM

Thanks everyone for the very helpful info ... I'm glad now that I asked the questions!

Spiny Norman

16-09-2005, 10:04 AM

We drew numbers from a 'hat' last night. Process seemed to be well received by all participants. I will now enter the players into Swiss Perfect in that order and publish the draw on our website. Will post again once that's done in case anyone is curious to see the result.

Spiny Norman

16-09-2005, 06:26 PM

Here's how the names came out of the hat:

Seed Name Rating

1 FROST, Stephen 1360

2 SMITH, Mike 1378

3 FROST, Jared 0

4 WALLER, Tony 1703

5 STONES, Douglas 1641

6 WONG, Ngiam Yee 2106

7 GUY, Wayne 1963

8 ABRAHAMS, Charles 1443

Round 1 Pairings are "standard Berger" as suggested:

1v8 2v7 3v6 4v5

Board Rating Name Result Name Rating

1 1360 FROST, Stephen (0) ABRAHAMS, Charles (0) 1443

2 1378 SMITH, Mike (0) GUY, Wayne (0) 1963

3 0 FROST, Jared (0) WONG, Ngiam Yee (0) 2106

4 1703 WALLER, Tony (0) STONES, Douglas (0) 1641

Spiny Norman

16-09-2005, 06:59 PM

Round 1 Pairings are "standard Berger" as suggested

Interesting though that the SP pairings for later rounds came out different to the TasChess Berger pairing tables referenced above. Should I be concerned about that, or just let SP do its thing? Also, the pairings SP produced for round one were:

1v8 2v7 3v6 4v5

but standard Berger is supposedly

1v8 7v2 6v3 5v4

What's the story with that?

Bill Gletsos

16-09-2005, 07:08 PM

Interesting though that the SP pairings for later rounds came out different to the TasChess Berger pairing tables referenced above. Should I be concerned about that, or just let SP do its thing? Also, the pairings SP produced for round one were:

1v8 2v7 3v6 4v5

but standard Berger is supposedly

1v8 7v2 6v3 5v4

What's the story with that?Those on the TasChess site look nothing like Standard Berger.

Standard Berger for round 1 is 1v8 2v7 3v6 4v5

Bill Gletsos

16-09-2005, 07:12 PM

You can work out standard Berger tables for yourself fairly easily.

The following shows how to do it for 8 players.

In the first round the smallest number plays the largest, the second smallest the second largest etc.

If there are an odd number of players make you make the group even by adding last number as a bye.

The first numbered player in each line is white.

Round 1 is therefore:

1 V 8

2 V 7

3 V 6

4 V 5

Now in successive rounds player 8 swaps colours (therefore they are black in odd rounds and white in even). Player 8's opponent is always the bottom right hand player from the last pairing of the previous round. Therefore in round 2 player 8 plays player 5. The remaining numbers are allocated in a anticlockwise direction starting with the number following player 8's opponent (note that player 1 follows player 7).

round 2 is thus

8 V 5

6 V 4

7 V 3

1 V 2

Round 3

2 V 8

3 V 1

4 V 7

5 V 6

Round 4

8 V 6

7 V 5

1 V 4

2 v 3

Round 5

3 V 8

4 V 2

5 V 1

6 V 7

Round 6

8 V 7

1 V 6

2 V 5

3 V 4

Round 7

4 V 8

5 V 3

6 V 2

7 v 1

Note that the 4 lowest numbers (1-4) play one extra white than the numbers (5-8). This is normal. The only time they are all equal is if player 8 is the bye.

Spiny Norman

17-09-2005, 10:15 AM

Thanks Bill for clearing that up. I guess I can go ahead and publish the draw now! Cheers, Steve.

four four two

18-09-2005, 01:48 AM

Frosty is Ngiam yee Wong a recent arrival from interstate or overseas? A foreign student perhaps? ;)

Spiny Norman

18-09-2005, 08:36 AM

Frosty is Ngiam yee Wong a recent arrival from interstate or overseas? A foreign student perhaps? ;)

Not at all. "N.Y." as he likes to be called, is a former Singapore citizen (and, I believe, a former Singapore national team member). He has lived/worked in Australia for many years and is now retired. Having been out of chess for a long time, he is now getting back into it again "for fun". Very knowledgeable fellow. At 2100-odd ACF, I suspect he is a little under-rated.

BTW, I feel a little sorry for my lad Jared ... he has to play N.Y. in R1 and then Wayne Guy in R2. :doh: Not easy being 12 years old, unrated, and playing against seasoned players ...

Rincewind

18-09-2005, 10:18 AM

BTW, I feel a little sorry for my lad Jared ... he has to play N.Y. in R1 and then Wayne Guy in R2. :doh: Not easy being 12 years old, unrated, and playing against seasoned players ...

You know what Nietzsche said.

Spiny Norman

18-09-2005, 06:46 PM

You know what Nietzsche said.

Something about being killed? ;)

Rincewind

18-09-2005, 06:51 PM

Something about being killed? ;)

No, something about not being killed.

Spiny Norman

18-09-2005, 07:11 PM

No, something about not being killed.

I knew there was a trick to it ... :)

Sp will give you two options, but they arent do with colour. It will give you a choice of how you want the players to be paired. Either by standard or rutsch. You can choose either, or you can use the berger pairing tables, located here: http://members.trump.net.au/taschess/berger.htm. Berger pairing tables are most commonly used.

Note though that if you are going to use the board numbers that sp gives you, then one player will always be playing on the same board. Might be an idea to change that around a little.

3) As players, what would YOU prefer to see in terms of tournament setup. Does it matter? What if you get hit with playing the top-ranked player with Black for the third time in a row in recent tournaments, or something like that? As a player, I do prefer the second method, but again, that is so all the announcements can be made before sitting down to play.[/QUOTE]

Apart from this thread I haven't come across any other discussion about pairings that even mention Rutsch. If players are sorted by rating then Rutsch is far better. Standard has 1 v 8 then 1 v 2, 3 v 1 and 1 plays easier seeds each round. Whereas Rutsch has 1 v 8 then 7 v 1, 1 v 6 and so on playing higher rated opponents each time ending with the final round match between 1 and 2. Thus with Rutsch you are much more likely to get an interesting finish to your tournament with the top seeds playing off.

TomekP

26-03-2015, 04:59 AM

If players are sorted by rating then Rutsch is far better.

Rutsch is another round-robin, therefore player are not sorted by rating players but by the random.

Pepechuy

29-03-2015, 04:29 AM

You can get creative and try something different than the hat.

In one tournament, players hit a high striker with the hammer, and ordered accordingly.

In another one, players chose gold bars.

Or ask the players several days in advance to predict some currency exchange rate. The closest one chooses his pairing number, and so on.

In any case, random is better than sorted by rating.

Sergio Pagano

30-03-2015, 05:10 PM

In any case, random is better than sorted by rating.

I strongly disagree with you. Random is unfair, it could be possible that 2 players having the same rating (= expectation to win the tournament) will have different number of whites, and that having more whites having white in the direct match. By the way, I don't know how the Rutsch works, can someone post here more details or links? Thanks in advance! I'm interested because I created tables based on rating and I would like to check if they are differences between Rutsch and mine tables. Thanks in advance!

Garvinator

30-03-2015, 05:54 PM

Remember for a round robin event, lots are meant to be drawn at an agreed time to determine the starting position of each player. So in sp, the 1 v 8, 2 v 7 factor does not mean much because the arbiter will have drawn lots to determine who is placed in position 1, 2, 3 etc. So it will not be highest rated player in seed 1.

Sergio Pagano

01-04-2015, 12:23 AM

I don't know how the Rutsch works, can someone post here more details or links?

Solved, it is the carousel, I was not aware of the synonym. BTW it is different from my tables, here you have an example of them for 8 players ordered by descending rating and having 1 white more for the stronger players:

1 1 – 8 2 – 7 3 – 6 4 - 5

2 7 – 1 8 – 2 5 – 3 6 - 4

3 1 – 6 2 – 5 3 – 8 4 - 7

4 5 – 1 6 – 2 3 – 7 4 - 8

5 1 – 4 2 – 3 8 – 5 7 – 6

6 3 – 1 2 – 4 5 – 7 8 – 6

7 1 – 2 4 – 3 6 – 5 7 - 8

Pepechuy

08-04-2015, 06:59 AM

Please note that I wrote "Random is better than sorted by rating", not "Random is best".

I was under the assumption that Berger tables would be used.

I can see that some sort of pairing that takes the rating into account can be attractive (you provide an interesting conceptual example), but still prefer random.

Anyway, professional players do not seem to mind a random draw. For amateurs, I really do not believe that the difference between white and black is big enough to worry about it.

I strongly disagree with you. Random is unfair, it could be possible that 2 players having the same rating (= expectation to win the tournament) will have different number of whites, and that having more whites having white in the direct match. By the way, I don't know how the Rutsch works, can someone post here more details or links? Thanks in advance! I'm interested because I created tables based on rating and I would like to check if they are differences between Rutsch and mine tables. Thanks in advance!

Powered by vBulletin® Version 4.2.5 Copyright © 2021 vBulletin Solutions Inc. All rights reserved.