Rethinking the One-Day Chess Tournament

[Note: I am bumping this post to the top of the page for a couple of days. New content will appear periodically below. Item originally published at 12:31 am EDT, 10/24/2015.  - Todd Durham]

Over the last year or two, I have been privileged to be involved in several interesting discussions with Wayne Strickland, a board member with the Central Florida Chess Club. The CFCC is the premier chess organizing entity in the Orlando area, and they do great work.

Click here to learn about the CFCC!

Our discussions have centered around ways to increase the player turnout at the CFCC Tornadoes, where attendance is generally poor, and consistent only in its inconsistency.  Even worse, some of the games lack competitive value- in the last one I played in, I (with my 1700 something rating) played John Ludwig at 2350+ for one game, a pair against a 1900 player and a 1400 player and one against a new player with a 499 rating.

Learning value of a game against John aside,  I went home with a 2.0 score, thinking I had wasted half the day.  To add insult to injury, I won the U1800 prize, and received a check for $37 in the mail days later.  I opened the envelop, looked at the check, and marveled at the absurdity of it all.

Wayne has had some great ideas, including drawings for free entries in the larger weekend tournaments, and also the implementation of a Club Grand Prix (an idea that exactly patterns what I proposed to the Florida Chess Association when I was on the board, but could not prevent it from being modified in a way that lost players.  Wayne succeeded in keeping the original proposal intact).

These ideas have had some flashes of success, but I believe there are three ideas that current organizers do not fully appreciate.  They are:

1)  For the vast majority of players, rating improvement is the number one motivator, and cash prizes are only a secondary consideration.

2)  Other things equal, a player would prefer to play a competitive game against an opponent of similar strength, where the outcome of the game is in doubt prior to the start of the game.

3)  A balance to the time control and the format, so that games are not too fast (my informal survey indicates that G/60 or slower is preferred), and a timely schedule that does not run well into the night. 

Of these goals, #3 is the most nebulous, so there will never be a consensus, but #'s 1 and 2 are, I believe, opinions shared by a great many players.

OK, that's enough whining.  What we need are some new ideas, and I would like to start with a pair of ideas that come straight from the US Chess Federation's Official Rules of Chess, 6th Edition, edited by Tim Just (Random House is the publisher).

The first idea is about pairings, and I will quote the rule book, specifically Rule Variation 29L1, 1 vs 2 Pairings:

"This pairing system is exactly the same as the Swiss pairing system except that the players in the upper half do not play the players in the lower half in the first or any other round.  Instead in round 1 after the players have been ranked, each odd ranked player is paired with the even ranked player following them on the ordered ranking list : i.e., 1 vs. 2, 3 vs. 4, 5 vs 6, etc.

In all other rounds the players are ranked in the rating order within their respective score groups and paired in groups of two starting with the two top-rated players in the top score group.  Odd players should be paired to the player in the next lower score group who closest matches with them in rating, using normal color priorities.  Color allocation, transposition, avoiding players meeting twice, byes, late entries, and withdrawals are applied exactly the same way as in a Swiss event.

TD TIP: Because it handles withdrawals and late entries as easily as the Swiss system, this 1 vs. 2 pairing system is easier, for some directors, to administer than 29 L, Using round robin table in small Swiss, in events with the number of players almost equal to the number of rounds; however, the 1 vs. 2 pairing system can accommodate many players in any size tournament.  This system has seen some popularity at the club and local level.  It is a hybrid of the traditional club "ladder" system (where one player challenges another player for his spot on the ladder) and the Swiss system."

I believe that the above rule would go a great way towards meeting player expectation #2 above.  It combines the competitive nature of round robins (mostly quads for us) where players will play other players close to them in rating, but it adds the "rise and descent" effect of Swiss pairings, where players will move up or down based on tournament performance.  Essentially, we still get a bit of "Swiss movement", but without the radical "multi-hundred" rating point swings of the normal Swiss.

With pairings like this, my two games against a 2350+ and a <500-rated player would have been replaced with two nominally competitive games- and every other player would experience the same effect.

Of course, I hear you asking "But Paul, with this kind of pairing setup, how do we allocate prize money?  With consecutive pairings, anyone could rattle off a perfect score!" 

This leads to my second idea- allow me to refer you to Rule 33E. Prizes based on points 

"Some organizers base prizes on points scored rather than place.  Such events often award prizes to all plus scores, a popular feature for players who doubt their ability to win the top-place prizes.

For example, in a five-round Swiss with an entry fee of $30, it could be announced that 5 points will win $100, 4.5 points will win $50, 4 points will win $30, 3.5 points will win $20, and 3 points will win $10.  With a fee of $60, prizes might be 5 points will win $300, 4.5 points will win $150, 4 points will win $90, 3.5 points will win $60, 3 points will win $30. These levels provide relative safety for the organizer, since even with a poor turnout prizes will often be less than entry fees.

Prizes based on points have proven more popular with Experts and below than Masters, so organizers should consider not using them in top sections.  It is desirable to have such events in sections of no more than 200-point classes each or to offer supplementary prizes for lower classes unlikely to make plus scores.  A separate section for each class is ideal if the expected turnout is sufficient (blogger's note: I consider this paragraph to be inapplicable if rule 29L1 is used, since anyone could have a plus score with consecutive pairings, but I included it anyway because I did not want to edit the citation in any way in my presentation).

The based-on-points method has a unique advantage over prizes that are guaranteed based on entries.  The announced prizes are never reduced, but their total varies according to the turnout.  This protects the organizer against financial loss without the player disappointment sometimes caused by prize reductions."

With this kind of prize distribution,  each player controls his own destiny.

So where does this leave us?  If we applied the two ideas, this is what we get:

1)  Each player will play an opponent close to his or her playing strength every round, gradually working up or down the rating list based on performance, with no wild pairing swings or "wasted rounds".

2)  Each player will be rewarded based on his or her performance in the tournament beyond a certain score, so any prize money they win will be earned, without waiting to see if a blunder or pre-arranged draw in another game takes them out of the running. (Side note: TD's could also pay out $$ this way much easier, simplifying the last round process for them!)

Will these ideas work?  Are they worth trying?  Please comment below, and let the discussion begin!