Proposal: A new World Championship format

by Frederic Friedel
11/4/2019 – When Paul Morphy played Adolf Anderssen in 1858, only two games of the eleven in total were drawn. Steinitz vs Lasker in 1894 produced four draws in 19 games, but when Capablanca played Lasker in 1921 there were already ten draws in 14 games. In the Carlsen-Caruana match last year all twelve regular games were drawn, so the title had to be decided in tiebreaks. But such tiebreaks involve rapid chess game, blitz and even the ominous Armageddon. Is there a way to decide the World Championship in classical games only? Problemist Werner Keym has proposed a format that does exactly this. What do you think?
YOUR PERSONAL CHESS COACH - Whether you’re taking your first steps into the world of club chess, or already playing at a tournament level: with FRITZ, you can train more efficiently, intelligently and with a more personalised approach than ever before.
FRITZ is more than just a chess engine – it’s a training revolution! Whether you’re taking your first steps into the world of club chess, or already playing at a tournament level: with FRITZ, you can train more efficiently, intelligently and with a more personalised approach than ever before.

Decide the World Championship by classical chess

The tie-break systems that are currently used in chess are designed to break ties between players who have the same total number of points after the last round. This is especially necessary when prizes such as titles, trophies, or qualification for another tournament cannot be shared. And of course in the World Championship, where there is a title to be awarded. But here the system used to break a tie between two players is either:

  1. the challenger needs to win; or
  2. rapid and blitz games, including the ominous Armageddon.

Nobody cheers when a World Championship takes course, it did last year.

Wikipedia describes how it continued: "Before the tiebreak, it was expected that Carlsen would be advantaged because of his prowess at rapid time controls. Although the rating gap between Carlsen and Caruana in classical chess was a mere 3 points, in rapid chess it was 91 points. However, Carlsen's showing in the classical games, and the way in which he offered a draw in a favourable position in game 12 led Kasparov to opine that Carlsen appeared to be losing his nerve. But this was the result:

  • Game 01: Caruana ½–½ Carlsen
  • Game 02: Carlsen ½–½ Caruana 
  • Game 03: Caruana ½–½ Carlsen
  • Game 04: Carlsen ½–½ Caruana
  • Game 05: Caruana ½–½ Carlsen
  • Game 06: Carlsen ½–½ Caruana
  • Game 07: Carlsen ½–½ Caruana
  • Game 08: Caruana ½–½ Carlsen
  • Game 09: Carlsen ½–½ Caruana
  • Game 10: Caruana ½–½ Carlsen
  • Game 11: Carlsen ½–½ Caruana
  • Game 12: Caruana ½–½ Carlsen

Tie-break games

  • Game 13: Carlsen 1–0 Caruana, 
  • Game 14: Caruana 0–1 Carlsen
  • Game 15: Carlsen 1–0 Caruana

Both players were gracious in the press conference after the match, and paid tribute to their opponent. Caruana said the results showed that Carlsen is the strongest player in the world, while Carlsen said Caruana had just as much right as he has to call himself the best player in the world in classical chess. Carlsen's strategy to draw game 12 and win the tiebreaks had been vindicated, a point he emphasized in the press conference. Both players recognized that the other was a formidable opponent. Carlsen said he was very happy for having overcome such a strong obstacle, and would work to get better in the future. For his part, Caruana lamented the fact that one needs to find one's best form to win a tiebreak, which he wasn't able to do, but he looked forward to making another title attempt in the future.

A proposal

Recently we received a letter from Werner Keym, who is a teacher (of French and Latin) and is one of the most creative problemists in the world. He is also a profound thinker in all things related to chess, and came up with the following interesting proposal:

Current rules

For Swiss chess tournaments for individuals (not teams), FIDE recommends — in an Annex to the FIDE Tournament Regulations regarding tiebreaks:

  • The result of the direct encounter(s) between the players (if any)
  • The greater number of wins
  • The greater number of games with Black (unplayed games shall be counted as played with White)
  • The Average Rating of Opponents Cut (the average rating of opponents, excluding one or more of the ratings of the opponents, starting from the lowest-rated opponent)
  • Buchholz Cut 1 (the Buchholz score reduced by the lowest score of the opponents)
  • Buchholz (the sum of the scores of each of the opponents of a player)
    Sonneborn-Berger

The Chess World Championship Match should be decided neither by rapid chess nor by blitz chess nor by Armageddon, at the end of the match, but instead by classic chess. For this the World Championship match should be split into of two parts: a prologue and the match itself.

1. Prologue

1.1 This consists of four classic chess games.
1.2 The player with the lower Elo rating plays white in the first game.
1.3 The player who gets more than two points is the prologue winner.
1.4 In case of tie (2:2) the player with the higher Elo rating is the prologue winner.
1.5 We now have a prologue winner and a prologue loser.

2. Match

2.1 This consists of twelve classical chess games.
2.2 The prologue loser plays white in the first game.
2.3 If the prologue loser gets more than six points in the match, he will be the champion.
2.4 If the prologue winner gets at least six points, he will be the champion.

Comments
  • Lots and chance play no role
  • Conditions for the champion and the challenger are equal
    There are at most 16 classic chess games
  • The prologue will take at most four days
  • In the prologue the player with the lower Elo number has nothing to lose and must play aggressive for a win
  • The player with the higher Elo, however, only needs to play for a draw 
  • The championship match is decided in at most 12 games and there may be much excitement towards the end: in case of a drawn score in the 12th game the prologue loser must win, whereas the prologue winner must at least draw
  • The match will end by a fixed day — this is important for organizers, sponsors, media, and audiences
  • The process is similar to gaining the pole position in F1 racing — a favourite is found in advance, and gets an advantage due to his pre-race performance. But it can be taken away during the race itself.

Tell us what you think of the proposal of Werner Keym. Should World Championship matches in the future be held in this way? Please comment in our feedback section below.

Editor-in-Chief emeritus of the ChessBase News page. Studied Philosophy and Linguistics at the University of Hamburg and Oxford, graduating with a thesis on speech act theory and moral language. He started a university career but switched to science journalism, producing documentaries for German TV. In 1986 he co-founded ChessBase.

Discuss

Rules for reader comments

 
 

Not registered yet? Register

mc1483 mc1483 11/5/2019 03:16
Complications are unnecessary. Give draw odds to the champion, like it was done in the past, and the problem is solved. Maybe it's not fair, but there are no fair solutions anyway: if all games are drawn and armageddon is not good, someone must have draw odds, one way or another.
palometaman33 palometaman33 11/5/2019 02:50
The tie breaking system using rapid, blitz, and Armageddon time controls failed to prove that Carlsen is the best player in Classical Chess. Not to take anything away from his abilities in the faster time controls nor his obvious skill in the traditional game, but he did fail to beat Caruana. On the other hand, Caruana also failed to beat Carlsen.

This is simple: Occam’s Razor. Stop all the foolishness and play 12 games. The player with the best score after 12 is the champion. If no winner after 12, then play until one player or the other wins a game. There. It’s done. Why must we always succumb to the annoying tendency to unnecessarily complicate the mundane? The constant tinkering with the formats and the addition of tiebreaks using faster time controls renders meaningless any comparison of current players performance with the lifetime achievements of great masters past.
Nekthen Nekthen 11/5/2019 11:26
1. They may have been all draws, but my memory recalls some great fighting chess.
2. Prize money. A $50k appearance fee each and $1m for winning the series of matches.
Champion gets draw odds BUT the prize money is held over until the next match.
Draws will always be there. Imagine that one player wins the first game. He will defend the draw for the next games until he loses a game.
lajosarpad lajosarpad 11/5/2019 11:09
Let's formulate what the problem is: the problem is that the players had a safety-first approach. The reason was that in case of a tie, they can get away with this safety-first approach, unless they have blundered and if the other player blundered, then the match is won. Of course, Carlsen was very confident with playing the tiebreak and Caruana was playing safety-first as well, maybe he was intimidated by Carlsen.

What could solve the issue? Naturally, a system where safety-first is not paid off. Since the goal is the World Championship title, that's the thing by which we can motivate the players and specifically, the motivation must be to make sure that they understand that if they reach the natural result of consecutive safety-first approach on both sides (a tied match), then at least one of the players will feel to be the loser.

Who should that be? Champion or Challenger?

Well, the Champion is already the best, that's why we call him/her the Champion. The Challenger has a right to contest, challenge that title, but the burden of proof is on his/her shoulder. So, let's give draw odds to the champion. Fair and square.

There were some ideas to make sure that the champion would be less comfortable with a drawn match than an outright win. I agree with this principle, but I disagree that the champion should be qualified. In case of a tie the World Champion would retain the title, but in the next match he would have a penalty in time controls. So, you have drawn all games with this time control. Let's see what you can do having less time than the challenger. Knowing this the Champion would have motivation to win the match as well.
BonTheCat BonTheCat 11/5/2019 10:52
Same old, same old. Just another solution which will skew the main match in favour of one player or the other. The simple solution is wildly unpopular in this day and age, I know, but if we really want to have a classical World Championship which is not decided by rapidplay and blitz or some ridiculous Houdini-like contortions, we should just return to the old format with 24 match games. The match would be long enough, both to allow someone to overcome an early loss, and to serve as a test of stamina.
Mr TambourineMan Mr TambourineMan 11/5/2019 09:11
pipopalazzo suggestion is bad as it ends up with no current champion. Even if it is only for some months you avoid this senario as this could end up very bad for examples with contact with sponsors, illness or anything else we couldnt imagin go wrong anyway goes wrong. Chess should have one champion. Even a mess with two or three champion is better than no champion at all. Why are we here? Problem is they'll had to abolish the draw odds for world champion in match Kramnik-Topalov as both was recognized as Wch. If I could imagine the stupidity in FIDE to throw in Rapid and Blitz into Wch match I would rather live with mess of two or three Champions at the same time. I remember I was asked "who is the current world champion" It was not funny to answer with two or three (Fischer) persons. But to have to answer "no one" is no no no.
besominov besominov 11/5/2019 07:35
Oh good, another proposal for the world championship format.

But seriously: this proposal is just another way to give draw odds, but in a bad way.

- "Conditions for the champion and the challenger are equal"

Not true. The higher rated player has draw odds in the Prologue. Which he can then turn into draw odds for the match.

If you're gonna do this it would make more sense to give the draw odds for the Prologue to the current champion because he has already proven himself before. (But you shouldn't do it in this way.)

- "In the prologue the player with the lower Elo number has nothing to lose and must play aggressive for a win,"

Also showing that actually the conditions are not equal.

All that being said I don't mind draw odds for the current champion.
(But for the whole match. Giving draw odds based on 4 games is just random.)

So just do it like they did in the past. This means the challenger has an incentive to win. (And yes conditions are not equal, I don't mind that. The current champion has already proven himself. A classical match should be decided in classical, pure and simple.)

(But not draw odds for the higher rated player, rating has nothing to do with a world championship imo, we're not looking for the highest rated player, for that we have the Elo list, we're looking for a match winner at the end of a fair and inclusive qualification cycle of slow classical chess.)
aljehin aljehin 11/5/2019 07:23
if you ask me. looking at the matches between two players in the last year. superiority win the match in case of a draw.
AidanMonaghan AidanMonaghan 11/5/2019 07:20
Solution: Experiment with shorter time controls for more decisive results. Elite level players are too equal in ability and too familiar with one another to generate more decisive results.

If one is concerned about any resulting quality of play reduction, then eliminate time controls altogether or watch an engine play itself. Otherwise, a balance must be found between quality of play and a reduction of draws that lead to incompatible tie-break formats.
allytton allytton 11/5/2019 06:59
I like Yassar Serwan's idea. It's simple and does away with making it complicated. I also think to make rapid chess or blitz the deciding factor. For me, Classical, Rapid & Blitz are not the same game. Rapid & Blitz should not be used to determine the World Classical chess champion. If you need to have a winner that bad instead of playing rapid or blitz simply have them toss a coin. If there is a boxing match for a champion and it's a tie then the title holder simply retains his title. This was the way it was done in chess until the Russians lost the championship.
Saint-Nick Saint-Nick 11/5/2019 06:27
I agree 100% the tie-breaker should be played before the actual match, the format of the prologue can be discussed as long as one player start match with 0,5 and the other 0, Regardless of the score in the following games the the players will never be tied. My suggestion to add to the mix is that the player trailing ALWAYS PLAYS WHITE. He will continue to play white until he overtakes the lead, then the white pieces pass to the other player.
Babysplitz Babysplitz 11/5/2019 05:57
24 game match is best I think with the champion remains champion if tied.
OR
maybe 16 game match and if tied, then you have CO CHAMPIONS.This way the Present Chaampion would try harder to win to remain the ONLY CHAMPION. IF THE MATCH IS TIED THEN CO CHAMPIONS.
I think that 12 games is too short for the WORLD CHAMPIONSHIP. Thus 16 games.
Forget rapid and blitz for the world championship. You have seperate tournament/match for those titles. THERE IS NO PLACE FOR RAPID AND BLITZ IN THE TRUE TITLE OF WORLD CHESS CHAMPION.
powpow88 powpow88 11/5/2019 05:25
Interesting. (1) Punishing the champion for not winning outright (else he becomes just a candidate next time) and (2) monetary rewards for each game won are very good ideas, anything that shake up the current stagnant format is good. Keep in mind that (1) which forces the champion to take extra risk helps the challenger, and (2) helps the champion because fees and endorsements of the world title could be a lot more than the extra prize money, so he can play for safe draws, but the challenger is forced to take risk to get the extra prize money.
the vhiz the vhiz 11/5/2019 05:24
Still doesn't stop Carlsen from simply drawing 16 games to retain his title. Ding would have better chances than Caruana in the current system of rapid+blitz TBs, so the motivations of the author can be questioned too (is he paid by Carlsen to put this out?).
artegall artegall 11/5/2019 04:22
A scenario is this: The prologue winner scores 3-0 (or 4-0). The prologue loser scores 6.5-5.5 and becomes World Champion. Overall score: World Champion 6.5 - Loser 8.5 (or 9.5). What will we all think?
pipopalazzo pipopalazzo 11/5/2019 04:04
I think that a simple way to solve this situation should be like this:

If the match ends tied, then the reigning Champion retains the title, BUT (here comes the change) he must participate in the next WC cycle as if he were another contender with no special benefits. The only way the champ can have the right to wait for a challenger is by winning the match (not only the title).
Peter B Peter B 11/5/2019 04:03
The first problem with this proposal is getting players to commit to the time for 16 days of play. That's why the format was reduced to 12 games to begin with.

It's an interesting idea, but the 4 game prologue still seems "wasted". Imagine a decisive result in game 1 - the player who is behind might then bide their time until the games which actually count. I also don't like using rating as a tie breaker.

I think rapid playoffs are here to stay, though an argument can be made of having the tiebreak played before the match instead of after. But other sports manage just fine with the tie break at the end.
Leavenfish Leavenfish 11/5/2019 04:01
Too much of a gimmick.- and one could manipulate 'ratings'...

You want decisive chess?

1. First person to 3 wins...wins!
2. You pay based on results: Lets say a $1,000,000 purse, give each player a $50,000 'appearance fee'...and wins determine the rest.
Ex: 3-0 = $900,00 for the winner, ZERO for the the loser. 3-1 = 775,000 for the winner, $225,000 for the loser....and so on. Money talks! Draws should not count in this day and age.
This (potentially) short format would create excitement because of the possible snowball effect - 1-0 and you are essentially playing without house money. 2-0 and you are...and you put all the pressure on the player who is behind...there is not a bunch of games left (as in the traditional set number of games format) to 'stabalize' and play mickey mouse chess. You have to win!
TRM1361 TRM1361 11/5/2019 03:56
"Jacob woge: No fewer than 16 classical games. The challenger must win to become W.C. The reigning W.C. must win to secure next match, else he/she enters the next cycle at candidate level."

I like it but I would say 12 classical games. Then sudden death. Flip a coin to see who gets white first then alternate. It must end with a win. If you can't win it in the first 12 you have to be a candidate next time.

Good thinking Jacob. Light a small fire under the champ :)
PendekarMustar PendekarMustar 11/5/2019 03:55
Worth to consider: Top player of Classical, Rapid and Blitz = World Champion.
adbennet adbennet 11/5/2019 03:52
In case of a drawn prologue giving the advantage to the higher-rated player is totally arbitrary. I enjoyed reading all the comments, people have clearly given this a lot of thought. "The challenger must win to become W.C. The reigning W.C. must win to secure next match, else he/she enters the next cycle at candidate level." This suggestion of Jacob woge is better than the one by Werner Keym.
PendekarMustar PendekarMustar 11/5/2019 03:47
Worth to consider: Top player of Classical, Rapid and Blitz = World Champion.
marifem marifem 11/5/2019 03:43
can they just have 16 or 18 classical games with the player with the first win getting the draw odds. the match may end at the penultimate round if the one trailing won't be able to draw. since the last game won't be necessary. still, only classical games without preference to the current champion. just to the stronger player.
but for the case of all games are drawn then the current champion retains the title. at least there are considerable amount of games to determine if the challenger can really dethrone the champion, not just 12 games.
PendekarMustar PendekarMustar 11/5/2019 03:41
ConwyCastle's idea is interesting.
PendekarMustar PendekarMustar 11/5/2019 03:38
ConwyCastle's idea is interesting.
powpow88 powpow88 11/5/2019 02:53
For the "purist" who say just play 12, 16 or 24 games, and the champion keeps the title if it's tie, the problem is you are getting and will continue to get 12, 16 or 24 draws in the world championships which is destructive to chess, very soon chess will be like checkers, just another boring game. The very top players today (at the level of carlsen and caruana) can almost draw at will if they refuse to take any risk. Changing the world championship from a two players match format to three players round robin will introduce a new set of game theory strategies, currently the strategy is "draw, draw, draw", and maybe lighting will strike (<5%) and the other guy would blunder.
chessaakash chessaakash 11/5/2019 02:37
I agree with ConwyCastle 7 : give the draw odd to the challenger instead of title holder. if you are champion then you should be able to beat challenger. If not then , its time for new World champion.

And ,perhaps play at least 16 games. No more rapid,blitz .
melante melante 11/5/2019 02:31
Anything that complicates thing is bad. Just play N games and, if the match is drawn, the current champion retains the title. Easy and straightforward. And it worked.
Jacob woge Jacob woge 11/5/2019 01:36
To express my opinion of this proposal, I suggest the winner of a tied prologue to be not the highest rated, but the tallest player. Rating is not a reward system.

Of course, you may run into identical twins.

No fewer than 16 classical games. The challenger must win to become W.C. The reigning W.C. must win to secure next match, else he/she enters the next cycle at candidate level.

So a perfect string of draws puts the title at jeopardy, only with a time delay.

When that candidates, including the W.C., is played out, the winner gets draw odds over the runner-up in the same way as above.

If the two top spots of such a candidates’ are a perfect tie, the winner is decided in the same manner as to-day. Same for a three-way. That situation is known and dealt with.

There may, in such events of drawn W.C. matches, be a period between Candidates and next W.C. match, without a World Champion.

Suggested variants of shuffle chess, whatever one chooses to call them, are different games. May I suggest, going down that alley, Double Bughouse instead. Two boards, two clocks, each player manages a black and a white game simultaneously. No opening theory, no draws. No need for even number of games.
Talhaunted Talhaunted 11/5/2019 12:54
Just about time that people realize there are far too many draws in Chess. When it comes to credit players for a gae, a draw should be called a loss with the only exception that it is king against king. The only thing that should count are WINS. Then decide how many are needed to prove the superiority of the the other player in a match. Draws as a rule should be treated as losses. The future of chess depends on that.
garyroe garyroe 11/4/2019 11:40
Keep it simple! The 'not so old' 24 game, winner by 12.5 points is still the most precise way to play a World Chess Championship. If all the games are drawn, so be it-nature of the game but that won't happen- then the Champion retains his title.
In this system the challenger must fight for a win, no blitz or Armageddon to fall back on, and the right amount of games to decide who's the strongest.
I believe the only reason 24 games was changed was because of $$$ money. Either lack of money or sponsors couldn't make a profit. Either way, take the money out of it if possible , and you have a good, honest, chess pure way to decide on a Champion.
So keep the reliable, fair and honest 24 game system in place, just find a way to fund it so everyone's happy . Don't destroy the integrity of a World Championship match by making it weird and unfair just so someone could make a profit.
I have some suggestions on that also.
wok wok 11/4/2019 11:14
This suggestion contradicts the FIDE tournament regulations. The player with the higher ELO has a worse ELO performance in case of a tie. Therefore the player with the lower ELO should win the prologue. In general the main problem with this system is that it gets ridiculous if ELO difference is small.

I would like to suggest a different idea. Why not apply the Armageddon idea like this: 16 games match with classical time control, Player A gets 10 games with White and 6 games with Black. Player B gets draw odds to win the title. The reigning champion gets the right to choose whether he is player A or player B.

Alternatively I like the "asterisk" idea of powpow88. However I would apply it like this: Give the world champion draw odds. After defending the title by draw odds once he looses his draw odds for all subsequent title matches.
kychess kychess 11/4/2019 11:08
I think one of the solution, to the point raised in the comments that the prologue winner could lose the match badly later is to:
Combine the prologue score to the actual match score. i.e. it would be out of 16 games instead of 12 games and all other aspects of the proposal would still remain the same. So then the result of the first 4 games would have much more importance as that would decide a player's momentum and his chances for the tiebreak.
Player-1: 3 (now needs only 5 points to win title)
Player-2: 1 (needs 7.5 points to fight back and win title, needs a total of 8.5 since he lost the prologue)
powpow88 powpow88 11/4/2019 11:01
Basically like the currently system, the champion keeps the title if all draws (16 games instead of 12), but the challenger is "encouraged" to go all out (bold or reckless?) in the first four games in order to get draw odds for the main event. I would propose instead that if after 12 games and still a tie, the champion keeps the title with an asterisk, then after 3 years the world championship decided in a three players round robin tournament adding the second (newly selected) challenger, each plays 4 games against the other 2 players (total 12 games), tie scores settle by the usual methods.
Halflash Halflash 11/4/2019 10:53
Simple, old school, and efficient :
1) 12 classical games ( no less, no more)
2) if draw (6-6) ? Great, then the world champion remains the world champion.
tomohawk52 tomohawk52 11/4/2019 10:38
Offhand, I can think of two downsides to this suggestion if I understand it correctly:

1) It encourages the higher-rated participant to sit on their rating.

2) It could encourage rating manipulation, or the appearance of rating manipulation at least when the two players have similar ratings.
Ixion Ixion 11/4/2019 09:43
Chess needs to change. If we want new players the format for serious chess must be rapid games, like 30 min. Nobody outside chess understands why games have to be so long, and I know several players who have stopped playing because they don't have the time for classical chess.
Phillidor Phillidor 11/4/2019 08:55
It's always good to see proposals for changes. One one hand, no system can be absolutely fair, on the other hand, systems can become too predictable or a bit harshly said boring.

The proposal has two good points: firstly, it is not ridiculous, and secondly, it is divided into two zones - first is "risk free classical chess" and the second one is "pure classical chess". Winning the first zone is probably much more important than at first sight.

Now here's where I could stop. The proposed system is great for it's sacred simplicity and it's classical chess only.

But being a devil's advocate - who said that, all according to the current system, when the score being 5,5 - 5,5 one of the players cannot go all in? Rapid games, blitz games and the armageddon are just a way of determining the second criteria. Better than a coin toss? Surely, at least I think so. But better than the 4 preliminary classical games? A matter of taste. Probably I still like the armageddon better. It's like penalties. Creme de la creme. 10 milions people on the verge of heart attack. I got one due Vitiugov's painfull loss. But it makes one stronger, now I cheer for him even more. So all in all, current tie-breaks are still more memorable than the proposed ones, in my opinion. Nevertheless, I would love to see the proposed system being tried in other chess events. But for the world championship match, it is importand stick to tried and tested experiments. While this one looks pretty sound, it would be worth giving it a try in some tournaments.
luckyhappy luckyhappy 11/4/2019 08:52
Much better than the current system.
JPBaptista JPBaptista 11/4/2019 08:51
Why don't we return to the first K vs K format: first player to win 6 games (or 5, or 4, to make it not so long) wins, not counting draws.

Of course, you would have a longer match, but you could reduce the prizes with an inverted proportion regarding number of games necessary to find the winner. So players would be discouraged to make too many draws.

That way you have the match decided only by classic games and necessarily with wins.
Previous123Next