World Championship – out of the box

by Frederic Friedel
9/19/2018 – The match for the highest title in chess is always very exciting. But it is somewhat spoiled if the result is a tie. In previous times it meant that the reigning champion kept his title; later a tiebreak was installed, with rapid chess games, blitz and then Armageddon. Not a satisfactory solution, with the championship not decided by classical games. Werner Keym, a chess problem composer, has an alternate solution, one that like most of his problems is out of the box. Tell us what you think.

Endgames of the World Champions from Fischer to Carlsen Endgames of the World Champions from Fischer to Carlsen

Let endgame expert Dr Karsten Müller show and explain the finesses of the world champions. Although they had different styles each and every one of them played the endgame exceptionally well, so take the opportunity to enjoy and learn from some of the best endgames in the history of chess.

More...

A proposal

Recently we published an article on a unique personality: Werner Keym, a teacher (of French and Latin) and a musician, who in 2010 was elected Mayor of the German city of Meisenheim. In 2014, at 72, he retired from that post to devote more time to his family — he has five grandchildren — and to his hobbies. The foremost of them is problem chess. Keym sent me a signed copy of his most recent English language book, Chess Problems out of the box, which has given me immense enjoyment ever since. 

Werner Keym is one of the most creative problemists I know. He specializes in chess puzzles involving castling, en passant captures, pawn promotion and retrograde analysis. For many years I have enjoyed his problems, which I often encountered. Many have the advantage of not being prone to instant solution by chess engines. They force you to think. Go buy a copy of Chess Problems out of the box — it's just €10 / US $12 (plus €2/$4 for postage). Outrageously good value for money. 

On page 184 of his book, Keym makes an interesting out-of-the-box proposal regarding Chess World Championship matches. It was (partially) brought on by the 2016 match in New York. In the final twelfth game Magnus Carlsen, to the surprise of many, allowed a quick draw with the white pieces. It was a clever decision: the reigning World Champion felt that he was stronger than the Challenger, Sergey Karjakin, in rapid chess, and he proved that this was indeed the case by beating the Russian in the following tiebreak. "I have no objection to this outcome," Keym says. "There has to be a decisive outcome, and that can be achieved with rapid chess, blitz, Armageddon, etc. But it makes me feel uncomfortable that the World Championship in classical chess is finally decided by non-classical games..."

So Werner Keym has made a proposal to modify the way the Championship would be decided: if the result is a tie, don't have the reigning World Champion retain the title, and don't have rapid, blitz and Armageddon games, in the end, to decide who gets the crown. There is another way (we quote directly from his book)...


The Chess World Championship match should be decided neither by rapid chess nor by blitz chess nor by Armageddon, but instead by classic chess. 

Proposal: The competition consists of two parts, prologue and match.

Prologue

  1. Who plays White in the first game is decided by lot.
  2. There are then 4 rapid chess games. If one player gets 2½ points, the prologue is over.
  3. Otherwise, the result is 2:2, and now 2 blitz chess games will follow. If one player gets 1½ points, the prologue is over.
  4. Otherwise, the result is 1:1, and now further blitz chess games will follow. The first win of a game will end the prologue.
  5. We now have a prologue winner and a prologue loser.

Match

  1. There is an odd number of classic chess games (e.g. 13).
  2. The prologue loser plays White in the odd-numbered games (1, 3, 5, ... 13).
  3. If the prologue loser gets 7 points, he will be the champion.
  4. If the prologue winner gets 6½ points, he will be the champion.

Comments

  • The conditions for the champion and the challenger are equal.
  • The prologue will take 2-4 days.
  • The advantage for the prologue loser is that he has white in the first and the last game.
  • The advantage for the prologue winner is that he wins the championship in case of tie.
  • The championship match is decided by at most 13 classic chess games and there may be much excitement towards the end: in the 13th game the prologue loser has White and must win, whereas the prologue winner has Black and must draw.
  • The match will end by a fixed day. This is important for organizers, sponsors, media, and audience.

Discussion

This proposal is interesting. In the prologue phase, a relatively minor condition for the main match is established, with rapid and blitz games. After that, the match is decided with classical games. It is easily possible that one of the players loses the prologue and gains the title by winning one more game than his opponent. A vaguely similar system was tried in the 2018 Altibox Norway Chess Tournament: it started with a blitz tournament which determined which five of the ten players would have one more game with White in the classic tournament. Fabiano Caruana qualified for the extra white game and won the tournament, ahead of Carlsen, Nakamura, Anand and Wesley So, all of whom had also got the extra White.

One possible modification of the Keym proposal: in closer keeping with the current system it might be better to start with just two rapid chess games and eleven (instead of 13) classical games. This could be the schedule:

  • Saturday: two rapid games, if necessary blitz deciders
  • Sunday: Rest day
  • Monday – Thursday: four games (1-4)
  • Friday: Rest day
  • Saturday – Tuesday: four games (5-8)
  • Wednesday: Rest day
  • Thursday – Saturday: three games (9-11)
  • Sunday: closing ceremony

The match would last two weeks, and in the final (11th) game, if it is necessary, White would need to win, while Black could keep or win the title with a draw. But the outcome would be decided in classical chess.

Naturally, it is of special interest to know what our readers — and chess experts around the world — think of this proposal. Please use our feedback section below to express your views.

Links




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.
Discussion and Feedback Join the public discussion or submit your feedback to the editors


Discuss

Rules for reader comments

 
 

Not registered yet? Register

fgkdjlkag fgkdjlkag 9/23/2018 08:52
"If you introduce chess960, would it make drawn matches impossible? I doubt it."
"If chess960 would be introduced, then its theory would be developed as well. Supercomputers can analyze chess960 as well, nothing will prevent GMs from memorizing ideas. Yes, memorizing opening lines for 960 starting positions might be beyond the capabilities of the human brain, but if that's not possible, they will memorize patterns and ideas. "

Yes, chess960 would be expected to reduce the frequency of draws, which would make it more likely to have an undisputed world champion after a match (instead of it being drawn). It would not make drawn matches impossible. Memorizing opening lines of 960 positions to the extent as in classical chess, which is 1 position in 960, is absolutely impossible for a human. It does not matter that computers can do it. Topalov made this comment recently. And it is not so easy to memorize patterns and ideas. Kasparov himself just said recently that switching the position of 2 pieces that are next to each other in the starting setup changes everything. All the patterns. There is nothing that relates. He said it stronger than I am explaining it here.

@Petrarlsen, I never compared chess960 stakes to classical chess, I was responding to your statement that there has been "nothing at stake" in chess960. My point about draw frequency is not based on results of previous tournaments, but theoretical.

"
- "(...) in chess960 it is impossible to "kill as quickly as possible any tactical possibilities, and going as fast as possible in the direction of a drawn position""

All these are pure surmises ; you could be wrong or I could be wrong... But my opinion is that, at this extremely high level, if BOTH players are trying to "kill as quickly as possible any tactical possibilities", they will succeed in obtaining this. I agree that it would be MUCH more difficult if only ONE player was trying to get these types of positions, but if they tacitly agree on this, I don't really see a win coming for either of them...
"
You do not seem to know much about chess960. Of course if both players want a draw in classical standard chess they can agree to it after 1 move, if it is not prohibited. This is not different to chess960. But you grossly overestimate the players ability to reach a drawn position. I'm guessing you did not watch the Nakamura-Carlsen match or the champion's showdown 2018 st. louis? Look at how unbalanced the positions were. The players had to solve serious opening problems right from move 1.
fgkdjlkag fgkdjlkag 9/23/2018 08:50
@lajosarpad

Maybe I did not explain sufficiently. Take an example. Let's say that Botvinnik defeated Alekhine to become world champion and that Botvinnik ,the challenger, was the stronger player at the time of the match. Next Bronstein qualified and played Botvinnik to a tie in a 24-game match. My point is that it is unfair that Bronstein did not become world champion when he is at least as good as Botvinnik. One cannot say that it is fair because Botvinnik had to beat Alekhine, because Botvinnik was stronger than Alekhine at the time (as an example), and Bronstein is equal to Botvinnik. Bronstein was unlucky to be born when he was so that when he contested the world title, the world champion was equal to him in strength. So the world championship should be designed so that the best player wins, no advantage to either side (as in draw odds to the champion).

@Petrarlsen, that was a mistype, it was supposed to be not "as much of an accomplishment as", the same as the example I gave above. Requiring the challenger to play better than the champion is not fair due to the historical time period point.

"Why drawn matches indicate that there are problems with chess? What is the problem with the game?"

"Your proposal is not guaranteed to work, see Kasparov-Karpov 1984. I do not say that now is more needed to have a limit than before, but we need a limit. Having the players to play the match for half a year is beyond the reasonable purposes of a match. And that's what happened in Karpov-Kasparov 1984."

You just answered your own question. There is something wrong with classical chess when a world champion cannot be decided in 5 months, no matter how you look at it. If you have blitz or rapid tiebreaks, then it shows there is something wrong with classical chess, because blitz and rapid are not classical. Giving one player an advantage (draw odds to the world champion), is only to prevent a drawn match, there is no other logic to it. I think this point is clear.

"I don't think that this is as simple as that. The 1984 Karpov - Kasparov match lasted for more than 5 months (and this seems to me to be MUCH to long), and, nowadays, more than a third of a century after, chess is still alive and kicking ! So it is possible to have an isolated problem for a given match, without this meaning that "Chess is Dead" ! And I think nonetheless that some precautions should be taken to avoid problems in the line of this 1984 match."

As above, the point is not that chess is dead, the point is that a sufficiently long classical chess match cannot determine a sole winner without introducing non-classical elements or giving an advantage to one player. It was not an isolated problem, it was mentioned here that it has occurred 3 times already.
Jacob woge Jacob woge 9/23/2018 08:24
“I have no problem at all with a tied match and I fail to understand what is the problem we intend to solve with a tiebreaks. “

Tiebreak was introduced in the Kramnik-Topalov “unification” match, because there was no challenger, but two world champions. Or at least, both regarded himself as such.

It has stayed with us since, for reasons unknown to me. I find it unnecessary and out of place to mix formats (including random chess) and would much prefer a longer match.
Petrarlsen Petrarlsen 9/23/2018 05:16
@ lajosarpad :

"I think the approach of giving the match win in case of a tie to the winner of the first game as a tiebreak as you have also stated, but that tiebreak is not necessarily just or logical. If we look at the Anand-Gelfand match, for example, Gelfand won the 7th game due to Anand's mistake and Gelfand's alertness, but then in the 8th game Anand came back with a miniature. Would it be logical or fair to give Gelfand the title in that case, of course, assuming your proposed system would have been introduced before the match?"

What you don't seem to be taking into account is that, in my opinion, the players would play differently (and perhaps QUITE differently). To be the first to win would become a separate objective, and I don't think that the Anand - Gelfand and the Carlsen - Karjakin matches would have been the same, with this system (more still because, with the "draw odds to the Champion" rule, a match without any win wouldn't be undecided : it would mean a match victory for the previous Champion, so the players - or at least the Challenger - couldn't simply play draw after draw waiting for a rapid or blitz tiebreak, because there wouldn't be one !).
Petrarlsen Petrarlsen 9/23/2018 05:01
@ fgkdjlkag (2/2) :

3) About Chess960 :

- "It is not correct that there has been nothing at stake for chess960, there were several world championships, there was the high stakes nakamura-Carlsen match, and the recent champion's showdown st. louis had a prize fund of $50,000 for EACH match."

Yes, but, a) the stakes can perhaps be compared to what represents a "normal" chess tournament (Tata Steel, Sinquefield Cup, etc.), but a Classical World Championship is quite on another level altogether, and b) if I remember well, each of these competitions were Rapid of Blitz competitions, and the draw percentages are always much lower in Rapid or Blitz than in classical chess. So, in my opinion, it is impossible to draw conclusions with these elements as to what could be a classical time control Chess960 tiebreak for the Classical World Championship.

- "(...) in chess960 it is impossible to "kill as quickly as possible any tactical possibilities, and going as fast as possible in the direction of a drawn position""

All these are pure surmises ; you could be wrong or I could be wrong... But my opinion is that, at this extremely high level, if BOTH players are trying to "kill as quickly as possible any tactical possibilities", they will succeed in obtaining this. I agree that it would be MUCH more difficult if only ONE player was trying to get these types of positions, but if they tacitly agree on this, I don't really see a win coming for either of them...

4) "Let's also be clear on one point. The fact that three 24-game matches were tied in the past, and that all this discussion is taking place, does not mean that there is something wrong with the system, or with the players, but indicates a problem with chess itself! So if a long match cannot produce a winner, something fundamental has to be changed; replace chess with FischerRandom, Harper's proposal, or something else."

I don't think that this is as simple as that. The 1984 Karpov - Kasparov match lasted for more than 5 months (and this seems to me to be MUCH to long), and, nowadays, more than a third of a century after, chess is still alive and kicking ! So it is possible to have an isolated problem for a given match, without this meaning that "Chess is Dead" ! And I think nonetheless that some precautions should be taken to avoid problems in the line of this 1984 match.
Petrarlsen Petrarlsen 9/23/2018 04:57
@ fgkdjlkag (1/2) :

1) I think that a completely unlimited match is a problem (cf. the more than 5 months long Karpov - Kasparov 1984 match) ; in my opinion, there must be some sort of a limit somewhere (perhaps something in the line of lajosarpad's system - which he didn't like much, by the way ! -, implementing two elements : a) a different scoring system ; b) a number of points to obtain to become the Champion : for example, if we would have 10 points for a win and 1 point for a draw with the title given to the first player who would obtain 40 points, with 3 wins and 10 draws - 13 games... -, the title could be awarded, but, even without any win, the match would end with the 40th game - in my opinion, by applying the "draw odds to the Champion" rule... -, which would still not be far from a 2 months match). I also think that this would be financially very difficult to organize, but, for this, the simplest thing is to try to organize it, and see if it works... And, when you say : "A long match would create tremendous excitement", I don't think that the excitement would continue if the match would not end after one or two months...

2) "(...) it does not mean much for the challenger to defeat the world champion (...)" In my opinion, it ALWAYS mean much, to defeat the World Champion : if the system is satisfying, to beat the World Champion means more or less to beat the best player in the World in a match, and I don't see how to beat the best player in the World can be considered as something not-so-significant ! There are thousands and thousands of IMs and GMs in the world, and to beat the best of them all in a match is ALWAYS a tremendous accomplishment, in my opinion... And if the Challenger's Elo rating is very close to the Champion's rating, it simply mean that the Challenger managed permanently to bring his playing level to a very high level... which is another very significant accomplishment by itself !!
lajosarpad lajosarpad 9/23/2018 04:11
@BKnight2003

It is simple indeed, but we still have tiebreaks. I have no problem at all with a tied match and I fail to understand what is the problem we intend to solve with a tiebreaks. But, as I described earlier, if we have to decide ties with tiebreaks, then it should be played with classical time controls. So I do not quite agree with you, but I think your proposal is better than the current system.

@Petrarlsen

I agree with you that the possibility of implementing a 3-1-0 scoring system with a score to be achieved by the winner is not optimal, but would be better than a match with a number of wins needed to win the match, due to the fact that we see the end of the match. I think random chess960 positions are so unkown that players probably lack the knowledge to be able to play in a safety-first measure. However, without relevant amount of data I cannot back my opinion with facts. I understand your point about your system, but I do not quite agree. I think the approach of giving the match win in case of a tie to the winner of the first game as a tiebreak as you have also stated, but that tiebreak is not necessarily just or logical. If we look at the Anand-Gelfand match, for example, Gelfand won the 7th game due to Anand's mistake and Gelfand's alertness, but then in the 8th game Anand came back with a miniature. Would it be logical or fair to give Gelfand the title in that case, of course, assuming your proposed system would have been introduced before the match?

"About the idea of giving an extra-white to the Challenger as a compensation for the "Champion keeps title" rule, I would think it would be better to let the Challenger chose if he prefers to have an extra-white or an extra-black game. This would then be HIS choice, and no one else would be responsible for it ! "

I think you are right here.

@fgkdjlkag

I understand your point now. Yes, the difference in strength between the world champion and challenger varies over time. Sometimes the challenger is stronger than the world champion, sometimes the champion is stronger than the challenger and sometimes their level is not quite different. The question is: what implications should this have on the system in your opinion?

Your proposal is not guaranteed to work, see Kasparov-Karpov 1984. I do not say that now is more needed to have a limit than before, but we need a limit. Having the players to play the match for half a year is beyond the reasonable purposes of a match. And that's what happened in Karpov-Kasparov 1984.

If chess960 would be introduced, then its theory would be developed as well. Supercomputers can analyze chess960 as well, nothing will prevent GMs from memorizing ideas. Yes, memorizing opening lines for 960 starting positions might be beyond the capabilities of the human brain, but if that's not possible, they will memorize patterns and ideas.

Why drawn matches indicate that there are problems with chess? What is the problem with the game? If you introduce chess960, would it make drawn matches impossible? I doubt it.
Petrarlsen Petrarlsen 9/23/2018 03:10
@ cepfan :

In my opinion, Werner Keym's proposal isn't significantly better than the present system : a tie is still solved by Rapid and Blitz games (to place them before is mostly only a question of calendar...), which isn't logical at all, as Rapid and Blitz World Championships also exist.

Or else, it would be necessary to split this title in two : a real "classical time controls" World Championship, and a combined World Championship with an equal representation of Classical, Rapid and Blitz games.

Taken together, the whole of the three present World Championships, with this Classical World Championship featuring mostly classical time controls, but sprinkled with bits of Rapid and Blitz, is not very coherent...

As for your second proposal, it suits me quite well ; I would only prefer to give the choice to the Challenger between an extra White or an extra Black ; as it is supposed to give him a small advantage to compensate partially the "draw odds to the Champion" rule, I think it would be more logical to let him chose what he prefers...
fgkdjlkag fgkdjlkag 9/23/2018 02:58
2)

A lot of proposals give one side an advantage, either by having an extra white, being able to get the white pieces first, or something else. Why the need for it at all? Have a match with exactly the same conditions for both players, and let the best player win. Eg, as I proposed early in this thread, a match of unlimited classical games with first to x wins, winning by a margin of y.
The idea that we need a fixed number of games today more than in the past is silly; insurance markets are much more developed now. A long match would create tremendous excitement.
@Jacob woge just made the same suggestion

I am not sure I understand @Bruce Harper's suggestion regarding 1/3, 2/3 points - it sounds to me that players are playing for the first decisive result, and the score for that day will be 1-0 or 0-1, no limit on the number of games? So the 1/3, 2/3 points is not relevant? Very interesting suggestion, would love to see a tournament with that format. It would have the effect of fundamentally change the nature of chess by making time management become much more important.

@BKnight2003, I think your idea is acceptable, but only if draw odds are combined with bidding for time, so that it is fair to both players.

@Petrarlsen, in chess960 it is impossible to "kill as quickly as possible any tactical possibilities, and going as fast as possible in the direction of a drawn position". The positions are just too rich and there are too many possibilities. Imagine how someone in the classical starting position would try to force a draw 300 years ago before the French, Berlin, Queen's gambit, sicilian even existed. Openings that are known to produce draws today were not known that they could produce them 100 years ago, and openings like the Giucio Piano 100 years ago that were thought to lead to draws are what players are using now to fight for an advantage. But with no opening knowledge it is very easy to make a large error that gives the opponent an opportunity too good to refuse.
It is not correct that there has been nothing at stake for chess960, there were several world championships, there was the high stakes nakamura-Carlsen match, and the recent champion's showdown st. louis had a prize fund of $50,000 for EACH match.


Let's also be clear on one point. The fact that three 24-game matches were tied in the past, and that all this discussion is taking place, does not mean that there is something wrong with the system, or with the players, but indicates a problem with chess itself! So if a long match cannot produce a winner, something fundamental has to be changed; replace chess with FischerRandom, Harper's proposal, or something else.
fgkdjlkag fgkdjlkag 9/23/2018 02:58
my first double post
1)
I believe I agree in general with @BeFreeBusy, @rubinsteinak, and I think Petrarlsen who want longer matches.

Agree with @Petrarlsen that it does not make sense to use chess960 for a tiebreak, it would simply introduce too much randomness.

regarding
lajosarpad 9/20/2018 11:19
"@fgkdjlkag the previous opponent of the World Champion was an earlier World Champion, or at least the challenger. If the previous World Champion or the challenger was significantly weaker than the current challenger, then the current challenger has better chances, but nevertheless, we are speaking about the absolute best of the world, there should be no weaklings. If there are, then there is another problem with the system which needs to be fixed."

I do not think you understood my point. The logic people are using is that since the challenger always had to defeat the world champion, it would be fair to use that system now, because the current world champion was once the challenger, and was forced to defeat the previous world champion under the same conditions.

But the skill gap between #1 and #2 is very different in various historical time periods. Look at the difference between Fischer and Spassky vs. the difference between Carlsen and Caruana. If the skill gap between #1 and #2 is absent (which has been the case a couple times in history), then it does not mean much for the challenger to defeat the world champion, but if the skill gap is tremendous, then defeating the world champion is a very different accomplishment. So forcing the challenger to defeat the world champion creates vagaries dependent on the time period.
cepfan cepfan 9/23/2018 10:12
The score 3:1 may be interesting in a tournement, but not in the championship match.

I think that these two simple procedures work well:

1) Keym's proposal (with a prologue)
2) Keym's proposal without a prologue:
Odd number of classical games.
The challenger begins and ends as white.
In case of tie the reigning champion keeps his title.
Petrarlsen Petrarlsen 9/23/2018 05:56
@ jsaldea12 :

"As I said this proposed new scoring system can be applied to world chess championship (...)"

You are certainly not overtiring yourself by answering to objections ; everyone says to you that what you say is absurd (in particular, mathematically absurd...), and you simply continue to repeat over and over the exact same nonsense ! You could at least TRY to refute the objections, but no ! You prefer to continue to state that 2 + 2 = 5 and that the white cat is in fact a black dog !... (Or perhaps a pink elephant ?!? Who knows ?...)

You would probably not succeed in earning anything else, but the price for perseverance is for you at 100 % !...
jsaldea12 jsaldea12 9/23/2018 03:09
Continuity: Why the proposed chess scoring system of 3 points for win and 1 point for draw is best answer to “draw problem”, said draw problem is brought to Arkady Dvorkovich, Candidate for President of FIDE. Finally the cat is out: “draw problem” is real. Why such proposed scoring system will reduce “draw problem? Because as stated the ratio difference is big: 3 to 1. compared to present outmoded scoring system of 1 win to .5 draw” Because of that big ratio difference, THE TENDENCY OF CHESS PLAYER IS TO PLAY FOR WIN RATHER THAN DRAW., MAKE THEM MORE COMBATANT, AGGRESSIVE, ORIGINAL, NOT COMPLACID, NOT CONTENTED WITH BORING DRAW. The problem would to how to orient, circulate to the chess players of this new scoring system that is simple and easy understand. As I said this proposed new scoring system can be applied to world chess championship or other tournaments or recreational chess.
Petrarlsen Petrarlsen 9/23/2018 02:43
@ BKnight2003 :

Your solution is very close to a 13 games match with one more game with White for the Challenger, and the "draw odds to the Champion" rule, the slight difference being that, with your system, after 12 games, if a player is leading, he will be awarded the title, and the 13th game will not be played.

As for me - very slight difference - I would prefer to give the Challenger the choice between an extra White game and an extra Black game : I am nearly sure the Challenger will choose an extra White game, but it seems more logical to give him the choice (for example, if he has some serious problems with his White repertoire, it could perhaps be something negative for him to have an extra White game - this would be quite paradoxical...).

Globally, your system would quite suit me ; the interest of my system would nonetheless be that, as the importance of the "draw odds to the Champion" rule is much reduced in it, it could more easily satisfy those who don't like very much the "draw odds to the Champion" rule...
Petrarlsen Petrarlsen 9/22/2018 11:19
@ Jacob woge :

"(...) If one would have a 3-1-0 scoring system in a chess match, I think it might work using a combination of a fixed target and an unfixed number of games (...)"

Yes, it works ; in fact, lajosarpad explained the exact same system previously under this same article. I suppose that, as he explained it rather differently from you, you didn't realize it was the same system...

But the objective is not at all the same that when this 3 - 1 scoring system is implemented in a tournament ; in the system you and lajosarpad exposed, it is rather some sort of a mitigated "draws not counting" system ("à la Karpov - Kasparov 1984", for example) ; it is not meant to give an extra motivation to the players for them to win games...
Jacob woge Jacob woge 9/22/2018 07:24
Same as in the final, knock-out phase in football WC, the scoring system is irrelevant in chess matches. You go plus one by winning, minus one by losing.

The reason that you win a 24-game match by reaching a score of 12 1/2 is not the score itself, as someone suggested below. It’s that your opponent can no longer catch up, even by winning the rest of the stipulated games.

Nevertheless, If one would have a 3-1-0 scoring system in a chess match, I think it might work using a combination of a fixed target and an unfixed number of games:

A) First to reach a certain score wins.
B) No fixed number of games stipulated.

This means, no matter how much you trail, there will always be enough games left to catch up thru a (hypothetical) winning streak.

On the other hand, different from the six-wins requirement of the early KKK matches, there will be an end to it eventually.
Petrarlsen Petrarlsen 9/22/2018 04:51
@ lajosarpad (2/2) :

- "In that case the champion was beaten in a game, but not in the match." (about the system I proposed in a previous post)

In a way, with this system, there are two competitions played at the same time : the "normal" match, and a "sudden death" playoff, in which the first player to win a game also wins the match. I don't completely agree that the champion would be "beaten in a game, but not in the match" : if you take the "sudden death playoff" side of the match, it consists of a series of games ; for example, if the first win is the 8th game, as in the Carlsen - Karjakin match, the seven previous draws count also - the players played those seven games, neither of them succeeded in winning them, and the win in the 8th game wouldn't be an isolated game, but the last of a series of 8 games. In my opinion, this wouldn't be at all the same thing as a "classical time control Armageddon" ; for a "classical time control Armageddon", only one game would count, whether, with this "sudden death playoff", the draws preceding the first win count also. And even in case of a win in the first game of the match, the possibility of a draw would have existed ; the players wouldn't have been forced to win or lose this game ; they could have drawn it, continuing the contest with other games, so I don't think that this can really be considered as just an isolated game ; this is more than this.

- About the idea of giving an extra-white to the Challenger as a compensation for the "Champion keeps title" rule, I would think it would be better to let the Challenger chose if he prefers to have an extra-white or an extra-black game. This would then be HIS choice, and no one else would be responsible for it !
Petrarlsen Petrarlsen 9/22/2018 04:50
@ lajosarpad (1/2) :

- "The benefit would be that the audience will get more games if the contestants cannot decide the match."

Yes ; in fact, it would be some sort of a mitigation of the system giving the title to the player who would win a certain number of games, without counting the draws at all. This last system is in fact the same as the system you invented, but with a 1 (for a win) - 0 (for a draw) - 0 (for a loss) scoring system ! The advantage of your system would be that it would avoid "unlimited matches" ("à la Karpov - Kasparov 1984") ; in case of a great number of draws, the match would be longer, but would nonetheless come to an end at one moment. I agree with you (and with your arguments) about the fact that this isn't optimal, but I think that it is nonetheless better that the "draws not counting" system - the 1984 match showed quite well that this system can end in an absolute impasse in certain circumstances !...

- "I disagree however about the chances of draws. Chess 960 is uncharted territory and players will not have too much knowledge about most of the positions, so they will make mistakes more easily, which increases the chance of a decisive game."

I think that it is VERY difficult to anticipate what would occur in such a situation, because the players would take into account into their general strategy a great number of elements, including, for example, the elements you stated ("Chess 960 is uncharted territory and players will not have too much knowledge about most of the positions, so they will make mistakes more easily"), and this can change vastly their approach to the game.

For example, perhaps, thinking that they are totally in uncharted territory - as you said - with the World Title depending on the result of the game, they would take an extremely conservative and "safety-first" approach, killing as quickly as possible any tactical possibilities, and going as fast as possible in the direction of a drawn position. And, at the World Championship level and with classical time controls, I think that if both players want this, the chances of a win would be absolutely minimal in this context. Globally, I think that Chess960 in such a context could be extremely different from what we have seen for the moment : most of the Chess960 competitions are more or less "exhibition competitions", with nothing significant at stake, and in rapid or blitz games. What would be a Chess960 game in classical time controls AND as a playoff for the Classical World Championship ? I think it isn't really fully possible to anticipate this...

- "However, if we make them drink a bottle of wine each, then we will increase the chance they will have a decisive result as well and even lesser minds will understand the drama."

Good one !!! But it would be necessary to chose a REALLY great wine, in accordance with the very high statute of the event !
Petrarlsen Petrarlsen 9/22/2018 03:57
@ jsaldea12 :

With your last post, you are either just plain wrong (if you mean that the 3 - 1 scoring system COULD change anything in terms of wins / draws ratio in a match - it is mathematically impossible) or completely of-topic (if your meaning is that the 3 - 1 scoring system is useful for tournaments - there is a time and a place for everything ; pages and pages of comments have been written on this question under other articles, and an article about the system for the World Championship MATCHES is not the place to discuss the usefulness of the 3 - 1 scoring system for TOURNAMENTS).
BKnight2003 BKnight2003 9/22/2018 02:52
@lajosarpad

Isn't my suggestion simpler, and still "keeps drama, has a fixed schedule, has an end and is chess"?

1 - Twelve games. Champion plays white in games 1, 3, 5, 8, 10, 12. Challenger plays white in games 2, 4, 6, 7, 9, 11.

2 - In case of tie, one more game with the same time control. Armageddon: Challenger plays white and must win.
jsaldea12 jsaldea12 9/22/2018 12:14
One of the questions presented by chess.com to Arkady Dvorkovich, Candidate for FIDE President is: “Chess is perceived to have a "draw problem." Finally it is out: drawn games is becoming monotonous and boring. To solve this problem the format of allotting 3 points for win and 1 point for draw is proposed. This makes the players more combatant, aspiring more for win rather than draw because the ratio: 3 to 1. Play this ratio and see thre difference with the present ratio of 1 to .5. Anyway it is only .5 why not play for draw..safe, just a little difference. But in the proposed ration 3 to 1, players becomes aggresive because the ratio difference is big
lajosarpad lajosarpad 9/22/2018 10:50
@Petrarlsen

The benefit would be that the audience will get more games if the contestants cannot decide the match. This means that if they play draws as a match strategy, then the match is longer and they risk to get tired and to make mistakes. However, if they play out their best opening novelties and try to beat their opponents at the start of the match, they will have better chances of achieving success. The other side of the coin is the fear of losing and playing in a safety-first approach, not taking risks. I am not a fan of this system, since I value drawn games as equally important as decisive games, therefore, a scoring system where the players split the point in case of a draw is logical, while a scoring system where the sum of points given to players is different in case of a draw than in case of a win is illogical.

I do not support the concept of using a tiebreak in world championship matches at all. But, if we cannot avoid having a tiebreak, chess 960 with classical time controls is a better fit in my opinion to chess (there is a 1/960 chance that they actually play chess, while in case of a rapid or blitz tiebreak there is no chance of that), but classical chess is obviously a better fit, so I think you are right about this. I disagree however about the chances of draws. Chess 960 is uncharted territory and players will not have too much knowledge about most of the positions, so they will make mistakes more easily, which increases the chance of a decisive game. However, if we make them drink a bottle of wine each, then we will increase the chance they will have a decisive result as well and even lesser minds will understand the drama.

Your system does not contradict the idea that the champion must be beaten, except in the case when the score is level and the first decisive game was won by the challenger. In that case the champion was beaten in a game, but not in the match. Due to this I am not especially in its favor, but it is by far the best system after the draw odds in my opinion. It keeps drama, it has a fixed schedule, it has an end and it is chess.

@conillet

Do we have factual knowledge that the perfect game of chess is a win for white? Do we know that black has less chances? We can only rely to empirical knowledge from the past, but that's just enough for a hypothesis. As about Jsaldea12 I think you were right and I were wrong. It seems that he or she thinks that changing the score for a win in a two-way match by itself would change anything. The reason of my mistake, was, as it appears that I try to find the rationale in everything and I often manage to find it at places where there is no such thing.
grimmlac grimmlac 9/22/2018 08:37
@fgkdjlkag, sorry but I beg to disagree with you when you equate white > black, where White may win. It's almost the same. If you think that first player may win with White on the first game. I would say that the 2nd player with White may win the also. If on the first game White who has the advantage then fight for a draw and wait for your turn to have White on the second game. Even though statistically, White has greater chances of winning than Black and there are greater chance of drawn games (but you know stronger player with Black pieces most of the time wins the game), it is still the player in best form who will prevail whether they have White or Black pieces.
If you would say that with 1st game, White has a chance to win against Black pieces, you are disregarding players strength and preparations, and summing up that chess games results is determined by color of pieces being played.
Petrarlsen Petrarlsen 9/22/2018 04:50
@ Masquer : Thanks !!
Petrarlsen Petrarlsen 9/22/2018 04:49
@ jsaldea12 :

"I am not speaking of world chess championship only. i am speaking of chess tornament, recreational play, etc."

The subject, here, is not what the 3 - 1 scoring system can give in a tournament (there were pages and pages of comments on this subject under other articles), but the World Championship system. And, for a match, as in the World Championship, the 3 - 1 scoring system cannot change anything as for the respective numbers of wins and draws.
Masquer Masquer 9/22/2018 04:47
Well said, Petrarlsen!
jsaldea12 jsaldea12 9/22/2018 04:29
I am not speaking of world chess championship only. i am speaking of chess tornament, recreational play, etc. that allot 3 points for win and 1 point for draw is applicable. You want to know the advantage: try playing wth this format, and see the effect..
jsaldea12 jsaldea12 9/22/2018 03:04
Make chess composition/puzzle more exciting, uniform universal. Chess puzzle is different from regular chess games because it is puzzle. It does not have to follow actual chess rules of play. . It is puzzle. Thus, in chess composition, allow all kinds of position, even 7 queens, ALLOWED because it is chess puzzle., illegal and impossible positions OK. a because it is chess composition/puzzle, not chess game.
Petrarlsen Petrarlsen 9/22/2018 02:57
@ jsaldea12 :

"The reason for this is to make chess players more combatant, forced to play for glory: win or loss, or win, not play complacency boring draws."

Please explain how, in a match, the 3 - 1 scoring system could have for an effect to incite the players to play for a win and to reduce the number of draws. I would be VERY interested by your (certainly very convincing) explanations !!!
jsaldea12 jsaldea12 9/22/2018 02:28
CONTINUITY:
The proposal of Frederic Friedel, Editor-in-Chief of chessbase is OK. But the already proposed to allot 3 points for win and 1 point for draw is best. Whether in world chess championship, or classical chess, speed chess, blitz, or any tournament, the proposed 3 points for win and 1 point for draw is APPLICABLE. Simplified. The scoring system need not be revised too. The reason for this is to make chess players more combatant, forced to play for glory: win or loss, or win, not play complacency boring draws. thereafter, the proposals/formats of Editor-in-Chief Friedel and others how many games for classical, speed, blitz will come into play BUT ALWAYS BASED ON THE SAME PROPOSED APPLICABLE BASIC RULE: 3 POINTS FOR WIN, 1 POINT FOR DRAW . MAKE CHESS GAMES MORE AWAKE LIVELY AND EXCITING. See?
BKnight2003 BKnight2003 9/22/2018 01:22
Finally got to read all the messages, and @conillet brought up a proposal very similar to mine (a little earlier). But mine is much simpler, I believe.
Petrarlsen Petrarlsen 9/21/2018 08:46
@ conillet : "The idea of giving an extra white to the challenger is a non-starter. In the case of both players being of equal strength, the odds would then be weighted AGAINST the titleholder."

I am not particularly favorable to this idea, but this wouldn't give an advantage to the Challenger ; the extra-White would only be a (very partial) compensation for the "Champion keeps title" rule. It is more or less the idea that, for the last game (in case of a tie at this stage), the tiebreak is a game for which the Challenger has White, but must win the game. And it is much more easy to draw with Black (for the Champion) that to win (on demand) with White (for the Challenger), so the Champion keeps a quite real advantage.
Petrarlsen Petrarlsen 9/21/2018 08:38
@ lajosarpad :

- About the 3 - 1 scoring system for a World Championship match, I both agree and disagree !

Disagree, because jsaldea12's idea was obviously simply to implement the 3 - 1 scoring system, without any other specific measure (cf. : "Whether in world chess championship, or classical chess, speed chess, blitz, or any tournament (...)" : he seems simply to consider that this system is a "miracle system" for any given competition). So, in this context, the 3 - 1 scoring system wouldn't change anything in comparison with the present scoring system.

Agree, because, indeed, with your very specific system, the 3 - 1 scoring system would introduce a difference. But it wouldn't have at all the usual effect of diminishing the number of draws ; it would have an effect on the length of the match. I know that you wrote : "I am not a fan of the idea of such a scoring system (...)" ; as for me, I don't see what would be the use of such a system ? What positive effect would it bring to have a longer match in case of a long series of draws than if the players exchange wins, for example ?

- As for the system I proposed, as a preamble, I must say that I haven't anything against the "Champion keeps title" rule ; I would rather have thought that, in its pure form, this system wouldn't be approved by the majority of the chess public and players, but, in fact, seeing the commentaries under this article, my impression is much rather that this system does indeed suit many people, so to implement once more this system would quite suit me too...

I am not favorable to Chess960 as a tiebreaker, for two reasons : 1) Because this would use in effect game B (Chess960) as a tiebreaker for game A (traditional chess). 2) Because there is no certainty that in the context of a World Championship tiebreaker, there would be much less draws with Chess960 than with traditional chess : so why switch to another game, if the positive effects aren't so obvious ?

When you say : "I think the world champion should be beaten by the challenger to get the world title", I think that this could apply to the system I proposed : the match would be at the same time a traditional match, and (for tiebreak purposes) a "sudden death contest" for which the winner is the player who wins a game first (and the Champion would have a very slight advantage, as he would chose the color for the first game). The players would know this in advance, and this would certainly influence their match strategy (for example, I am more or less certain that Carlsen wouldn't have "committed harakiri" against Karjakin in the game he lost against him in their World Championship match) ; to succeed in being the first to open the score would become an important element (in particular for the Challenger, because it would reverse the tiebreak in its favor). So, if the match was tied and the Challenger should win because he won the first win of the match, the Challenger would have beaten the Champion : not by the total number of wins, but by the fact that he would have won a game first - both would have known that this was important, but the Challenger would have been the player who would have succeeded in obtaining this...
conillet conillet 9/21/2018 07:39
The idea of giving an extra white to the challenger is a non-starter.
In the case of both players being of equal strength, the odds would then be weighted AGAINST the titleholder. Who on earth would want that?
conillet conillet 9/21/2018 01:02
Playing the tie-break (which may never be needed anyway!) beforehand is anti-intuitive, stressful for the players and also plain ugly.
Deciding a World Championship match by Blitz, let alone Blitz-Armageddon, is also a huge no-no.
A decider with rapid games only (open end, for as long as it might take) would be preferable by far.
Alternatively, a "classical Armageddon" (one single game!) might be played:
The White player gets normal classical time controls, and both players bid (blindly) for playing Black with draw odds. Whoever offers to play Black with LESS time gets to play it.
(Incidentally, I don't get why such bidding isn't done in ALL Armageddon games. It has been used in the US-Championships, but for some reason never caught on.)
lajosarpad lajosarpad 9/21/2018 11:45
@Petrarlsen

I think the world champion should be beaten by the challenger to get the world title. Your proposed system would possibly give the title to the challenger if the challenger would achieve the very first win, even if the champion levels the score. If we want a fight, then we need to cope with the case when all games are drawn. If that is the case, then declaring no winner and the champion loses the title without the challenger winning it and then deciding who the new world champion is in a tournament seems to be a possibility. As you already know I am not in favor of tiebreak games in general, but I prefer classical chess or chess960 with classical time controls if one wants to have a tiebreak at any cost. In fact a match with a level result, where the champion retains as a possibility is quite fascinating for me. It would mean that the challenger was not weaker than the world champion, but was not quite there, it was very interesting to entertain this thought after the Kramnik-Lékó match. As a Hungarian, of course, I was very much rooting for Lékó, but I did not consider the result to be a loss for him. In case of a win by a tiebreak for Lékó I would not have been convinced that he is really better than Kramnik in chess, nor a Kramnik win in a tiebreak would have convinced me. Since both players were playing very well and none was significantly better than the other, a tie was a fair result and I did not feel the need for a tiebreak.

I initially thought the same thing about the 3-1-0 scoring system in a world championship match as you described, but then I realized that this kind of scoring in a match could have gains. If the winner of the match has to collect a certain amount of points, let's say 15 points and 3 points are earned for a win, 1 point is earned for a draw, then a match with more draws would last longer than a match with many wins, with a few exceptions, where only one side wins, but there are draws as well. If all games are drawn, then 15 games will be played, as each player will get 1 point per game. If all games are decided, then on average each player will earn 1.5 points per game, so a match could be decided in 5 games if a player wins all games and the longest match with only decisive games would consist of 9 games. So, in case a score is to be reached, then rewarding wins with more points makes a difference. I am not a fan of the idea of such a scoring system, but this is the reason I avoided to write a comment with similar ideas as yours.
Masquer Masquer 9/21/2018 11:39
Forget the prologue -- the champ retains the title in case of a tied match, but here's the wrinkle:
the challenger gets an extra White! Simple, and no prologue or tie-breaker would be required.

EDIT: I've just noticed that cepfan has suggested the very same thing in the previous post. lol :-/
cepfan cepfan 9/21/2018 10:12
A great number of games does not matter. In 1951, 1954 and 1987 there were 24 classical chess games with a tie at the end and the reigning champion kept his title.

The best procedure is the one by Sonshi (9/20/2018):

15 classical games
The challenger starts and ends as white.
The reigning champion keeps his title in case of tie.
Pieces in Motion Pieces in Motion 9/21/2018 05:42
The 24 game format is still the best and should the result be a tie the world champion wins so the match can end. 24 games is long enough for the combatants to do their best and at the same time short enough to keep sponsors from getting worried and fans/viewers bored. And with the champ retaining the title should the end result be a draw it'll force the challenger to really go for the win hence making the champ do his utmost to retain the title. It's the format that contributed a lot to making the Fischer-Spassky and Kasparov-Karpov matches so special.
Petrarlsen Petrarlsen 9/21/2018 04:01
@ jsaldea12 :

"Whether in world chess championship, or classical chess, speed chess, blitz, or any tournament, the proposed 3 points for win and 1 point for draw is APPLICABLE."

A completely erroneous reasoning !

In a tournament, the 3 - 1 scoring system can theoretically change something (even if, in my opinion, its usefulness hasn't been demonstrated satisfyingly) : if a player wins half of his games and loses the other half, he will have more points than a player who would draw all his games...

But in a match, whether you allot 1 or 10 points to a win, at the end, the result is exactly the same : the player who wins the match is the player who wins more games !

For example, if the player A wins a game with a (supposed) 10 - 1 scoring system, he has an magnificent 10 points lead. But, if, for the next game, the player B wins, as he will also gain 10 points, the lead from the player A disappears as suddenly as it appeared previously ! And a draw is always neutral : each player gains the exact same number of point (be it 1 out of 3, 1/2 out of 1, or even 0 : this doesn't change anything).

I would remind you that the idea of the 3 - 1 scoring system is to "punish" players who draw too frequently their games... But in a match, if one player draws his game, the other player draws it also !!! So the 3 - 1 scoring system cannot change anything !!