Pairing Systems

Running Chess Tournaments
1

Moderator Mode: Off

I just Googled “chess pairing systems” and found the following:

  1. The Swiss System began in 1895 in Zurich, hence the name Swiss System. It is used in tournaments that are too large for a round robin format and better than an elimination tournament when all the players can play at the same time.

  2. As above the only 2 other pairing systems mentioned are the Round Robin and elimination format.

So, my question is are any other pairing systems used in Chess tournaments?

Also, how do they operate?

2

For a specialty event pitting team against team, there is the Scheveningen system. Every player on each team plays every other player on the opposing team. I saw this used to help players to get a FIDE rating, too. That tournament pitted 6 FIDE rated players versus 6 others, most of whom either had no FIDE rating or not enough games for a publishable FIDE rating. Not sure if FIDE still allows this anymore, but as long as norms are not at stake, this seems like a nice format for publicity and to grow a base of FIDE rated players.

In the Pittsburgh Chess League, we use a combined Swiss and RR format for the last division. We have a 5 round prelininary Swiss to determine seeding for the Quad Finals. If the last group has six teams they play a three round Swiss final. This maximizes the number of teams that can win finals trophies.

3

Keizer System: This was targeted for multiple week events at chess clubs - so that if players didn’t show up there was a pairing around it. It is intended for club tournaments, not weekend events.

  • Each game is paired between nearly equal players
  • Players may play each other more than once
  • Not everyone will play everyone else (participants may exceed rounds, OR games are more interesting by having equal participants replay or both)
  • Everyone plays in every round unless a round has an odd number
    -SCORING varies depending upon opponent strength: Each player is assigned a value, with the highest value to the top ranked player, the second highest value to the second ranked player,etc. “Roughly” the value of the top ranked player is approximately three times the value of the lowest ranked player, and each next player has a value one less than the previous one. Example: A tournament of 35 players would set the value for the top most player at 50, and the bottom ranked player would be 16. SCORING is that you get the value of your game score x the player value for your points in a round. So in this example if you beat the top player you get 1 x 50, draw you get .5 x 50, lose you get 0 x 50.

Color balance is given to the player with the greater score. If two players have the same color balance then the player who has played the most successive games with the same color will alternate. If this criterion ties too then the history is checked.

Absences don’t necessarily score 0 points, since an absent player would drop on the ranking and would meet a much weaker player in the next round, which is not the goal here, equal-ish games are. So absences are usually rewarded with 1/3 of the player’s own value.

This part really requires work: The values of players are recalculated after each round, and then the score is recalculated back to the tournament beginning. That is, after the first round, the players are now ranked by their scores, the 50-16 Keizer scores are re-assigned, and the tournament score is now recalculated. One can imagine doing this in Excel.

Holland System (was used extensively in chess for years until the Swiss became popular. US Opens (and the predecessor events) often used this. The Peoria Chess website mentions this with respect to the 1945 U.S. Open held at the Pere Marquette Hotel. Round robin preliminary groups with seeding into round robin finals (and “class” finals)

Scheveningen was already mentioned.

There are Swiss variants -

Dutch removes the repairing of two players restriction

McMahon System actually awards extra points at the start of the tournament (for example, all players above 2000 receive 1 extra point, all players 1800-2000 receive 1/2, then pairings are made from that point forward. Points awarded count in the final standings. (The McMahon Score.) It is named for Lee McMahon of the NY Go Club.

Monrad sets the pairing numbers at random and then pairs as a Swiss.

Konrad is a Swiss with X rounds - but only the x-n scoring rounds (sorted by score) count for the player’s score - that is, maybe there are 8 rounds, but only the top 6 scoring rounds go into the player’s tournament standing. This might be good for a weekly club event over 2 months, for example, where people may miss some rounds.

Pittsburgh System: Swiss preliminary with round robin finals - although its probably been used before, this was used in the 1946 U.S. Open in Pittsburgh. 1947 was the first year that it was a straight Swiss.

The elimination system mentioned above can also use various methods for seeding.

4

The rulebook mentions the 1 vs 2 system…as it implies 1 vs 2, 3 vs 4, etc. Players are placed in ratings order in their score groups and essay contests against the next player on the list starting from the top on down. The rules for color assignment and not playing someone twice still apply.

We used a “pick a pairing” system at the club a few times. I don’t remember the details but after a round or two the players really didn’t care who they played so picking became a challenge.

5

Yes, I was going to come back this morning and mention that Wayne Clark and you had developed a 1 v 2 system that had some surface similarities and goals similar to the Keizer system.

Sevilla does Keizer pairings - it is available at jbfsoftware.com/joomla15/ind … &Itemid=51

My understanding is that it is free, and also does some types of Swiss pairings.

6

On occasion at our club we’ve run rated games, not a tournament per se, over a period of weeks with only one or two rounds per week. No prizes, just the chance to get in some longer-than-usual time control rated games. Knowing that a lot of people won’t be there every week, we’ve used the 1 vs 2 system, but without regard to score groups.

7

Yes, but that’s still a Swiss (if players with equal scores are still paired against each other). It’s just not a ratings-controlled Swiss (a term invented by Martin Morrison, I think). Or, more accurately, the manner in which the ratings control the pairings is different than in a “standard” Swiss.

Bill Smythe

8

How about the Verber system?

In a large single-section 5-round scholastic Swiss during the Fischer boom, with three rounds played and only two rounds to go, there were still about 35 perfect scores. Chief organizer Richard Verber did not want the event to end with multiple perfect scores, so in round 4 he took over the pairing duties from chief TD Tim Redman. He extended the standard policy of pairing the lowest player (if there are an odd number of players) in a group against the highest player in the next group. In this case, he paired only a few of the very highest-rated perfect scores against each other. The remaining whole-bunch-of lower-rated perfect scores were paired against players with higher ratings but lower scores (2.5, 2.0, and maybe even 1.5).

It worked like a charm. After the fourth round, there were only three perfect scores. In the fifth round, the top two drew each other, while the third lost to a higher-rated opponent in the next score group down. NO perfect scores at the end of the tournament.

I’ve always thought Verber’s idea could be quantified into a more precise set of pairing rules, that could handle hundreds of players in five rounds with little chance of multiple perfect scores.

Bill Smythe

9

And now Dubov System: fide.com/fide/handbook.html … ew=article

10

There are two different types of Holland System - Balanced and Unbalanced.

In a Balanced Holland System all the players are divided into equal sized groups of close to equal strength (it is unlikely to be possible to exactly balance the groups’ strengths.) Each group plays a round robin within itself. The winner, or sometimes the top two, qualify for the Championship finals, also a round robin. Those not making it to the Championship finals are placed in lower level round robins.

In an Unbalanced Holland System all the top players are put into the top group. The next best players are put into the second group, etc. Again, all the groups are equal sized, and again, each group plays a round robin within itself. All plus scores in the top group, the top two scorers in the second group, and the winners of all other groups are put into the Championship final round robin. Those not making it to the Championship finals are again placed in lower level round robins.

The Holland System was once the standard pairing format for U.S. tournaments before being supplanted by the Swiss System.

11

Apparently there were no chess moms in those days.

12

As I recall (this was in 1972) there were few, if any, chess moms hanging around the event. The kids were mostly high-school age.

Richard Verber projected an air of authority that few challenged. During the opening announcements before round 4 (the first round with inventive pairings), he even announced that, despite what everybody might have been thinking, the pairings were correct. That was all it took.

Rigid rule-following is the bane of improvement, and of good new ideas. It’s often a good thing when every so often somebody steps outside the box a bit.

Bill Smythe

13

The year before, in the December 1971 HS tournament, there were around 110 players. It was played in the old Chicago Chess Club, located on S. Wabash, on the second floor - right next to the “L”. I remember watching expert Harold Boas curl his fingers up in his hair and slowly pull as an “L” train screeched by.

The next year, December 1972, the tournament had about 450 players for a 5 round Swiss. I finished 4-1 - taken out in the Verber pairing.

14

“Why yes, Mr. Tournament Director, you may be innovative as long as I believe that it benefits my kid.”

15

110 players was a lot for that site. It must have been really crowded.

I wonder if Harold Boas still has as much hair now as he did then. Probably not.

That year it was held in the old LaSalle Hotel, if I remember correctly. I was on the tournament staff (too old to play in a HS event). I watched as Richard Verber wickedly smiled and gleefully chuckled while loading the round 4 pairing cards into the Peterman Pairing Boards. I think Tim Redman watched, too, covering his face with his chin in his hands as Verber systematically crushed Redman’s anatomically correct straight-out-of-the-rulebook Harkness pairings from rounds 1-3.

Thank you for jogging, refreshing, and confirming my memories of that tournament. Any pairing system capable of producing 0 perfect scores in a 5-round single-section event of 450 players can’t be all bad.

Bill Smythe

16

I had a friend who played the first round in a closet. Literally!

Yes - at LaSalle and Madison. Was it there again the next year when the tournament had 350? (About 2 years later it was torn down.) Or was it at the St. Clair?

17

My first tournament ever was a few years before that, in 1968, at a still-earlier Chicago Chess Club location on Van Buren. At that event, while his staff was trying to get the round 1 pairings together, Tom McCormack was running around announcing, only half-jokingly, that the last three boards would be played in the elevator.

Bill Smythe

18

It’s a great motivator to play well.

19

The joke among teams with poor results at the US Amateur Team East/World Amateur Team is that they will play the next round in the parking lot. And good luck with that - there’s no room in the parking lot either!

20

Well, if it’s the last round, at least they’ll be close to their cars so they can make a quick getaway after they lose.

Bill Smythe