The United States Chess Federation - Keep the Draw

Home Page

Chess Life Online

2011

October

Keep the Draw — Fix the Flaw!

By Tom Braunlich

October 28, 2011

Like the farmers in Grant Wood's masterwork American Gothic, the chess world can often appear resistant to change.

In recent months innovative proposals for dealing with draws in a new way by GM Rustam Kasimdzhanov on chessbase, GM Sergey Shipov on chessbase, and IM Greg Shahade here on CLO have been controversial and met mostly with widespread skepticism by the chess community.

From the commentary I've seen it seems most people did not understand the true purpose of the proposals, mistakenly thinking they were just another scheme to "prevent draws." Statistician Jeff Sonas, for example, responded with a detailed historical analysis of draws in top-level chess, found that not much has changed, and asked, "What exactly is the problem?"

But many criticisms miss the main point. The K/S/S proposals are not an attempt to eliminate draws. They are, rather, a new proposition - to alter the format of important chess tournaments to deal with draws in a much more spectator-friendly way. They recognize that the tournament formats we have come to use in chess are fundamentally flawed in the sense that they create an esoteric barrier between the sport and the general public. Other sports have been willing to make major changes to fix their flaws in order to maximize their potential for popularity. Why not chess?

This is the true purpose of the K/S/S proposals. They don't view draws as something bad. Beautiful hard-fought full-length draws would still be appreciated and celebrated, as they should be. There is no attempt to outlaw draws (even boring short ones) or pressure the players to play a minimum number of moves, award more points for wins, or any of the myriad other artificial anti-draw schemes.

Instead, the new proposals simply say that all results must be decisive ... that a draw is fine but doesn't count toward the required decisive outcome, thus requiring further play. I hope to show here - with specific details and comparisons to other sports - that this simple change in attitude would have critically important positive ramifications for chess events and their potential popularity.

"Clearly, we need to change the philosophy behind our competitions. Chess needs reforms, directed towards improving its reception by the wider public. And the first logical step is to ensure that every pairing results, every evening, in a definite winner, just as happens in other sports.” (GM Shipov)

"Matchplay"

The clearest way to think of the new proposals is to adopt a simple change in our common terminology. Currently, the two players who are paired together take part in a "game" which can be either won or drawn. The new approach says each tournament pairing would be considered a "match" and it can only have one result: somebody wins. If the initial game of the match is an exciting draw, okay, but draws don't count for the result of the match. There must be further play (according to the simple specific rules below) until there is a decisive result and a winner for the match.

I propose this type of tournament be given a name to distinguish it from regular tournament competition. My suggestion is to call it "Matchplay," to emphasize that a match is being played and not necessarily a single game. It has several important advantages for tournament organizers and spectators discussed below. There are also some criticisms that I will also try to address.

Kasimdzhanov compared chess to tennis in his original proposal. It is true that "chess is not tennis," as several pundits unhelpfully stated, as if that was enough to dismiss the new reform ideas. But this analogy can be useful to us when we think about the ways that tennis has continually worked to fix its flaws and improve its appeal over the years from a slow dignified sport to a fast-paced favorite. For example, like chess, tennis used to be mired in traditional scoring systems that gave it a stodgy reputation for long matches, sometimes with "marathon" sets (i.e., 29-27), all in an awkward attempt to deal with tied set scores. Finally in the 1950s a movement began to revamp the scoring system to be more spectator-friendly, and after a few experiments, the current 7-point tiebreaker was instituted in the early 1970s, coinciding with the sudden surge in the sport's popularity, quality, and media coverage. In other words, they had the courage to change their scoring system to better deal with the problem of "draws" (tied sets) in an exciting way (the tiebreaker) that both sped up play and had more spectator appeal.

In a sense, chess is in a similar predicament. The awkwardness of draws has limited our sport's competitive structures, since we have always gone with tradition and let draws dictate our formats, rather than designing them first by what would be most exciting to the public. We need a similarly innovative solution to that of tennis — not to "outlaw" draws in some way, but to simply change our scoring structure so that a drawn result is no longer sufficient. This is best done by considering each pairing to be a match that must have a decisive conclusion.

The new proposals advocated by K/S/S do exactly this, in slightly different ways, and thereby could open up vital new opportunities for the organizer and spectator the same way it did for tennis, paving the way for increased chess popularity with the general public.

Matchplay Rules

First, before we consider the pros and cons, how exactly would this kind of "Matchplay" work? Here is my suggestion for the specific rules, which is a slight variation on the K/S/S proposals which I think has some advantages:

1.      The two players paired together will play a "match" which will be won by the first player to win a game.

2.      If the initial game is a draw, the clock is stopped and the players take a short break while the arbiter resets for the next game using each player's remaining base time (as detailed in #3 below). The players switch colors and play another game using these new times.

3.      Time Control - The players start the match with a pre-agreed "base time" and a "delay time" amount, plus a "minimum time" and a "secondary delay." (For example, the base time might be 100 minutes per player plus the delay time 30 seconds per move, with a minimum of 5 minutes and secondary delay time of 5 seconds per move. This could be written shorthand as "100m/30s - 5m/5s") If the initial game is a win for either player, the match is over. If it is a draw, the clock is stopped. It shows how much time each player has remaining of the original base time. This will be the player's new base time for the next game, (except that if it is below the minimum it will be reset to the minimum). The delay time will be reset to the secondary amount for all the remaining games. Each time there is a draw this process is followed until there is a winner.

An example helps to make it clear - suppose GMs Kasimdzhanov and Shipov are paired for this kind of Matchplay with the time control detailed above. Let's say the first game ends in a 40-move draw with Kasimdzhanov (white) having 34 minutes left, and Shipov having 48 left. These minutes become their remaining base time for the next game. They take a short break while the arbiters reset the board with reversed colors with Shipov (now white) playing G/48 and Kasimdzhanov playing G/34. Both will have a delay increment of 5 seconds now (the predetermined amount for the secondary delay increment). Suppose the 2^nd game also ends in another 40-move draw with Kasimdzhanov having 9 minutes left and Shipov having less than 1 minute. The third game is reset with colors reversed again, and this time because Shipov went under the minimum of 5 minutes his time will be reset to the minimum of 5/5 for the third game, while Kasimdzhanov will have 9/5. At this point one of them wins brilliantly after 40 moves (I won't tell you who!) and the match has its decisive result in the third game.

Notes:

It's also possible to simply use pre-set time controls for the 2^nd game and subsequent games, as Kasimdzhanov proposed, instead of basing it on each player's remaining base time. But many (including Shipov) have complained this could lead to a much longer overall playing session, especially if the first game is a very long one, and too much fatigue for the players as a result. The above rules should minimize the amount of extra time players might need compared to the initial overall scheduled time. Note that if the break time between games were 5 minutes the example match would have only lasted 14 minutes longer than a single long full game with the same initial time control that had the same total number of moves (120).
Shipov advocated going immediately to blitz game playoffs to minimize fatigue, but this seems very odd to me if the first game was a short draw. Blitz specialists would be happy with early draws, even as white, counting on having good odds of winning the blitz. My suggestion, which shares similarities with Greg Shahade's idea, discourages this tactic by preserving the initial "pace" set for the first game of the match. It only goes overtime or reaches a blitz pace when they've used up their base time.
The clocks must use a "delay" format in "Matchplay." If an "add-on" time increment is used which allows the players to accumulate base time it won't work because they can make additional pointless moves in a drawn position in order to run up their base time for the next game. Also, "delay" allows the organizer to more easily estimate the maximum amount of time used per match for planning purposes.

What are the Advantages of "Matchplay"?

§         Draws maintain their integrity and importance. Draws are essential for chess and there is no attempt to outlaw or punish them. But since draws don't provide the "match" with a decisive result, the players must continue. Since colors switch for the next game, whoever currently has white has yet another reason to play hard even if the position is equal.

§         Decisive results are more exciting to spectators and fans. This is an unproven assertion, but I think there is ample evidence. A draw has a feeling of incompleteness, and the spectators often do not feel like they have received full value for the time and money spent watching the contest. This is especially true of the short lifeless draws such as those seen at the recent Candidates Tournament, which leave observers feeling cheated and which are very bad for the public image of chess. Decisive results under Matchplay should create more excitement and greater interest among the general audience for chess - it wouldn't alone guarantee success, but it would "fix the flaw" and remove a major obstruction.

§         The key advantage is that Matchplay opens up important new vistas for organizers to try to create events that are more attractive to a casual chess audience and which therefore have at least the potential to grow the sport's popularity. Why? Because now the organizer is freed from the need to deal with the awkward nature of draws or lengthy playoffs in his tournament structure. This both simplifies tournaments (making them more easily understandable to a novice) and allows for more exciting finales.

All of the traditional forms of chess tournaments can still be used with Matchplay rules - round robins and Swiss systems, etc. Kasimdzhanov points out how round robin events could be made more interesting in this kind of format, because the decisive results creates more tension, more chances for comebacks, and more prospects for sensational results such as a player like Carlsen running the field 13-0 in a 14-player round robin.

However, Matchplay would allow tournament forms that international sports indicates are the most attractive to fans - namely, Elimination tournaments (sometimes call "knockout" events), and "Preliminary/Playoff" formats.

To maximize popularity with the general public, a tournament needs to tell an exciting story, which means it needs to come down to a final climax. All modern sports strive to create the growing tension and exhilaration that is generated when the event features players getting dramatically eliminated until there are only a few left and then finally the last two. The spectacle of the "fight to the death" that such formats create is obviously a lot more thrilling than what we get from typical chess tournaments, such as a round robin in which during the final round the leader needs only to draw against the player in 10^th place. (And as Greg Shahade has pointed out this situation is ripe for the temptation of cheating.) Sometimes in such events the winner is decided even before the last round and the final games are completely superfluous. This is a terrible problem for chess promotion, as has been argued forcefully by GM Maurice Ashley, among others.

Comparisons with Other Sports

Other sports almost always structure their events to maximize last-round tension, and do it in similar ways. Let's take a moment to go through some informative examples so that we can see what we can learn when compared to chess:

One classic form of tension-creating tournament is the Wimbledon-style elimination format used in tennis. It generates maximum excitement. However, since this gives no chance for a comeback after one bad result most sports use a variation on it that involves a "preliminary" schedule of several games aimed at qualifying some of the players/teams for the "playoff" elimination matches. The world cup of football (soccer) for example has preliminary rounds that lead to about half the teams going on to elimination playoffs. This allows a team a chance to overcome one "bad day" in the early rounds and come back to make it to the playoffs and even to the championship title. (Note also that tie games are allowed during the preliminaries but are not allowed during the playoffs. Those matches must continue until the result is decisive, much like in the Matchplay rules for chess advocated here.)

Many other sports follow this example. In some cases the preliminary rounds consist of a "season" of very many games to determine with relatively high accuracy the best teams who qualify for the playoffs, as in American football, baseball, basketball, hockey, and most international football (soccer) leagues, etc.

Even those sports that don't lend themselves directly to the playoff format strive to emulate something similar in their tournaments. The World Series of Poker main event, for example, eliminates the starting group of thousands of players as they run out of chips in order to eventually reach a "final table" of nine surviving players. In this case the individuals play as a group (rather than in head-to-head competition) and eliminate each other one by one until only a single champion remains. But the effect is the same. The tension is always ratcheted upward.

Even those sports that are completely different than chess in terms of their style and configuration - such as gymnastics, figure skating, Olympic races, etc. - still utilize elimination concepts and last-round tension in their tournament structures to enhance excitement. For example, golf is a unique sport in which the players don't directly battle each other, but rather compete to get the best score against the golf course as an individual. Head-to-head match play with an elimination format is possible in golf, but they have discovered it doesn't work too well for TV since several minutes elapse between each player's shots and the TV coverage has a poor pace. So they prefer to have lots of players in the final round so that they can switch back and forth between them and fill time with action. Even so, note that they have structured their rules over the years to create escalating tension and to bring the play down to the final shot - players have to get through elimination tournaments just to make it to the main event, then they must survive a "cut" that eliminates half the players after two days. Then the remaining players start the final round in reverse score order, so that the leaders tend to be among the last few players to finish, and it often comes down to the final putt. If there is a tie at the end they don't accept that as a result and go on to a playoff!

Yes, I know you might be saying that chess is different from all these sports. The main difference is - these sports are far more successful!

Tournament Formats Affect Popularity

We should learn from their example and realize that tournament formats affect popularity and that breaking out of our old traditions might have a positive effect. Chess is the only major international sport I can think of that does not typically use climactic tournament structures of some kind. If we do use them, as in the World Cup, they are clunky two-game matches with separate playoff days, often filled with short draws, and thus don't have the excitement level they should.

Using "Matchplay" chess, a pure Wimbledon-style elimination tournament is possible and would be fun, especially for a closed invitational event, but for open events I'm sure most players would appreciate having several preliminary rounds before the do-or-die playoffs begin so that they would have a chance to come back from a poor start. So, for example, an American Open using Matchplay might have the following format: 8 rounds of standard Swiss System play (in which draws count normally, not Matchplay) that qualify 16 finalists for elimination playoffs. The seeding into the playoff brackets is done according to the players' final ranking at the end of the Swiss. (i.e., if the player with the fourth highest rating ended the preliminaries alone with the highest score, then he/she will be seeded number one in the playoff brackets). The playoffs then proceed as an elimination event with Matchplay, escalating the tension from 16 players down to 8 to 4 to 2 and finally to a champion. Everyone would get at least eight games in the prelims, and the finalists would also have up to four matches.

This same format would work for a closed national championship in the U.S. or other countries. (For instance, twenty invitees play a number of regular preliminary Swiss-system games to qualify, say, eight players for the playoffs using Matchplay rules.)

"Matchplay" rules could also be used for traditional chess "matches" such as the world championship with some positive effects. Instead of a match of 12 or 16 games in which draws count - some of which are frustratingly short - using Matchplay rules would produce a winner every round of the match. Spectators won't need to risk coming on a day of a short draw anymore, because every round there will be blood. Isn't this better than a traditional match that is tied after the pre-set number of games and the entire result of the match comes down to only one set of rapid/blitz playoffs? Under Matchplay rules the blitz play, if it happens at all, is spread throughout the match and leaves open the possibility of a comeback from one blitz mistake. Matchplay rules never lead to breaking an ongoing tie with infamous "Armageddon" final game.

You might think that changing these traditional events in such a way is heresy. One must consider whether making chess competitions more exciting and attractive to the general public is worth changing a comfortable but ineffective status quo. Would chess not gain much more than it loses? Other sports have made similarly controversial improvements of their flaws in the past, and improved their appeal. Here are some more specific criticisms with my reply:

Criticisms

Length of session - It could result in extremely long playing sessions. Imagine the first game went 115 moves and down to the last seconds on the clock, perhaps over 5 hours of play. Now they have to set up and play again with 20 more minutes on their clocks? That's perhaps another hour of play, with perhaps more to come, and seems exhausting. True. But the Matchplay rules variation improves on Kasimdzhanov's original proposal by minimizing the extra time that might be required when the base time is exhausted. In this case, after a very long initial game, the 2^nd game would go directly to using the minimum (blitz) time control. Yes, if there is a long series of draws the match could go very long. But since the player with white (at least) is always motivated to try to win, a long series of draws is less likely than in current forms of blitz playoff games. No longer would players accept strategic short draws. Very long matches also can happen in tennis without hurting that sport's appeal. Matches at Wimbledon frequently go 5 sets and 5 hours of exhausting physical play, or longer. (The Isner-Mahut match in 2010 lasted over 11 hours!) In any case, Matchplay rules minimize this and I believe would tend to not go much longer than current full time controls. Only experience with the system would tell for sure.

White/Black - The "Matchplay" rules would not compensate for the advantage of playing white, and perhaps are therefore unfair. This is true to some extent. Matchplay does reduce this advantage by switching colors after an initial draw, but the player who gets white in the first game maintains a statistical edge. Yet this inequity exists in all regular chess tournaments and can only be completely eliminated by insisting that each player always have an even number of whites/blacks (i.e., as in a double round robin, or in "playoff matches" which consist of sets of two games, leading to further tied results that still have to be broken anyway). Insistence on perfect parity is yet another factor that restricts chess formats and hinders popularity. It should not be considered the primary consideration. As in regular chess tournaments, the privilege of playing white can be made more fair by awarding it based on who "earned" this edge from the preliminary rounds of the tournament, and there are a variety of ways that could be done.

Only for Elite Events - The Matchplay rules would not be appropriate or relevant for most small tournaments, such as a 5-round weekend Swiss, which must pack two and sometimes three games into one day. That's true, but I don't see this as a negative. Such events typically have no need for formats that enhance the promotional potential of the tournament. Matchplay is mostly for international events, promotional events, exhibition matches, or major championships that are usually played at the pace of one game a day. If you had an important event that is traditionally two games a day, you might adapt them for Matchplay via the Preliminary/Playoff format, with two-a-day preliminary rounds that qualify some number of players for Matchplay playoffs at the end.

Clock Management - The retention of unused base time for the second game might have some effects on strategy in the initial game, which is not a factor in current chess events. Players might feel pressured to move more quickly than they otherwise would in order to maintain their time cushion in case of a draw. My response: Good! Clock management will become just another enhanced strategic factor of the sport. Also, because of this effect, whoever has white is discouraged from making the kind of short draws that Alexander Grischuk, for example, was doing with white at the recent Candidates tournament. Such a strategy would be much more risky under Matchplay since, because base time is preserved, he would be giving his opponent the white pieces plus lots of time to work with them.

Rating the Games is Weird - How would Matchplay be rated? Simple. The first game played at the normal long time control is the only one that is rated for ELO points. The secondary games are not rated.

Comebacks - Elimination or playoff matches leave less chance for a comeback (if a player loses a match) as opposed to a Swiss or round robin. My response to that is, "too bad!" Try again next time. Meanwhile the spectators loved it.

Pace and the Natural Slow Nature of Chess - Matchplay rules won't help the underlying slow pace of chess. Many critics of the K/S/S proposals say that as a slow and thoughtful game, chess can't appeal to the general public and will always have a limited attraction. Therefore, reforms in tournament structures are pointless and we shouldn't bother trying to make formats more exciting. I think there is evidence that this opinion is at least unproven, and quite possibly just wrong. They used to think the same thing about poker until several critical reforms were made to present the similarly slow game, including the hole-card camera, post-tournament editing for TV, etc. Chess has generated widespread media attention during Fischer-Spassky or Kasparov-Deep Blue, but the ground for continuing that momentum was not fertile. If formats were improved schemes for promoting chess on TV would have a fighting chance. Chess tournaments could (in theory at least) use the same strategy of post-tournament editing for TV that worked so well for poker, presenting the games with the kind of bite-size chunks (focusing on the key turning points of the game) that the general public can relate to instead of the four-hour-long games presented live.

Tradition - Such ideas upset the comfortable status quo for the beautiful sport of chess and create a break in continuity with the past. Perhaps the most important objection. We can all understand that emotion, but to answer the objection I will quote the grandmaster who sparked this new movement:

"It will be good for our sport. ... People will try extremely hard with white.... Instead offering a draw in a slightly better ending in order to save energy and catch a movie, chess players will show their whole ability and will win these endings. As a matter of fact this will develop classical chess.... Chess will become a true sport. We'll wake up to win or to lose that day. We'll come to the board, ready to play chess. Every single day we'll see players like Aronian or Grischuk pressing with white, wriggling out of trouble with black, and showing some blitz skills to an ever-larger public. That is something I would like to watch and play." (GM Kasimdzhanov)

I support the general ideas of the K/S/S proposals and especially advocate the variation I described above as "Matchplay." I believe these formats should be experimented with and given a chance to show their strengths!

About the Author:
NM Tom Braunlich is an expert on both creating and promoting games for the mass market, having worked as an independent professional designer for over 20 years with companies like Hasbro, Milton Bradley, and Parker Brothers. He has written numerous books including The Art of Game and Toy Design, and was associate editor of Official Rules of Card Games by the U.S. Playing Card Company. He was organizer of the 2008 FKB U.S. Chess Championship.

Read or Write Comments on this Article
(32 so far)

October - Chess Life Online 2011

Announcements