12 draws: not really exciting

by Stephan Oliver Platz
12/6/2018 – The world was watching, but some people saw the spate of draws differently. Does the World Championship need reform? This is a debate that will continue for a while, perhaps until the next World Championship match in 2020. Contributor STEPHEN OLIVER PLATZ offers his opinion on Carlsen vs Caruana match as well as the World Championship format. The thesis? Blame the rules. | Photo: World Chess

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.


Some thoughts on the rules of the match

I watched the 2018 World Chess Championship match between Magnus Carlsen and his challenger Fabiano Caruana at home, entirely online. Unfortunately, I could not be in London myself this time, but the live commentary from the Saint Louis Chess Club team of GM Yasser Seirawan, GM Maurice Ashley and WGM Jennifer Shahade was interesting enough. Special guests like Garry Kasparov, Vishy Anand, and Vladimir Kramnik commented on the games via Skype, and most of all, I found Kasparov's assessments of the situation very impressive, because they often formed a strange contrast to the computer analyses presented by Maurice Ashley. 

But the twelve classic games that Carlsen and Caruana played seemed less thrilling to me, although I do not want to criticise one or the other player specifically for it. After all, the World Championship title and a lot of money was at stake! Considering these facts you can understand Carlsen's decision not to play the 12th game for a win. Why should he risk anything, if he was sure that he would be able to win the rapid games? The 3-0 in the tie-break proved him right. Nevertheless, a certain discomfort remains: 12 draws in a row — should that be the future of chess?


Whatever it takes | Photo: World Chess

Today's rules are inappropriate

My thesis is: Today's rules of a World Chess Championship in classical chess are not likely to produce profound and exciting games — I mean the games with long time controls, not rapid and blitz, which can, indeed, be very exciting.

From 1951 to 1972 and again from 1985 to 1993, we had a wealth of exciting, 24-game World Championship matches. Why were the rules changed several times since then? This affects not only the number of games to be played but also the introduction of the so-called "tiebreak" with ever-reduced time controls. Also, the time controls for the regular, "classic" games have been reduced more and more in the World Chess Championship matches over the years. Bobby Fischer and Boris Spassky still had 2½ hours for the first 40 moves. Later the players had had only 2 hours. Today they have only 1 hour and 40 minutes left with 30 seconds increment per move from the start. Did this constant shortening produce more profound and exciting games? I don't think so. And why should a tiebreak with rapid or even blitz games be used at all to decide who is the "classical" World Chess Champion? Must a marathoner (running 42 kilometers) be also the best 200 or 100 meter runner, if he wants to win the World Championships?

The negative consequences of shorter time controls

Because there is much less time available for the first 40 moves than some decades ago, opening preparation in top chess is becoming increasingly important. Bobby Fischer spoke in this context of "pre-arranged games" and said that chess is completely "dead" nowadays. Anyone who can rattle off the first 20 or 25 moves of a game out of memory will have a great advantage right from the start, not only because these moves have already been tested at home, but also because he has more or even much more time left for the rest of the game than his opponent, who perhaps had to start ten moves earlier to think for himself. 

This situation is exacerbated by the increasing strength of modern chess programs. While previously a whole team of top Soviet grandmasters helped Anatoly Karpov prepare novelties in the opening for his matches against Victor Korchnoi, today every halfway-technically savvy user can have such a grandmaster team consisting of "Komodo", "Houdini", "Fritz" and other chess programs working and analyzing for his own benefit. And why should one leave these technical possibilities really unused? No mathematician would even think of giving up a computer and working with a slide rule or paper and pencil instead. And yet, there remains the danger that top-chess may become too mechanical sooner or later due to the proliferating analysis of "opening" variations. How could you counteract this trend?


That's a tough one | Photo: World Chess

Longer time controls + more match games = more exciting chess

Producing more profound and exciting games requires, in my opinion, two elements: On the one hand additional time for the two players and on the other hand more match games. Why do the top grandmasters need more time for the first 40 moves? Someone who, for instance, sacrifices a pawn for some compensation (a lead in development, open lines, more space, attacking chances), cannot afford to make a weaker move, because that might soon lose all compensation and the game may already be lost. But how can you find the best moves, if less and less time is available? Therefore, in my opinion, the time control for a World Championship match should be increased again back to 2 1/2 hours for the first 40 moves in this manner:

Each player gets 1 hour and 50 minutes for the whole game + an increment of 1 minute per move from the beginning.

This means that the players will have 2½ hours for the first 40 moves just like Bobby Fischer and Boris Spassky had in 1972, and after the 40th move, there remains 1 minute per move until the game is finished. Even if 100 moves had to be played, a game would last no longer than seven hours. And both the attacker and the defender would benefit from the extra time which is especially useful in the critical phases of a game. When the crisis is over and one of the players has a clear advantage the increment of one minute per move should be sufficient to secure victory. Profound and more exciting games will be the result.

24 games are better than 12

The number of games was halved from 24 to 12! Who's going to take risks in this format? The shorter the match is, the harder the first loss will be. If 24 games must be played, even a 0-2 backlog can not be prematurely decisive, as we learned from the World Championship match Spassky vs Fischer in 1972.

When asked, both players in London supported the idea of lengthening the match, although a range of 16-18 games was mentioned.

No more tiebreaks, please

In addition, I think the challenger who wants to become World Champion should again be required to defeat the reigning World Champion in classic chess with long time controls, not in rapid or blitz games. In case of a tie, the reigning world champion should simply retain his crown. This requirement would significantly affect the match strategy. As long as a tiebreak with rapid or blitz games is guaranteed in case of a tie, there is absolutely no need to play for a win, neither for the challenger nor for the champion. This fact significantly influenced the course of the 2018 match!

Yet another aspect speaks against a tiebreak consisting of rapid and blitz games: The single-day shape of the athlete often proves crucial. Every chess player knows it from his own experience: Tired and badly rested you can't compete in rapid or blitz! Fatigue almost inevitably leads to frustrating losses. And the degree of fatigue will certainly increase during an exhausting World Championship fight. While less important in games with long time controls it can be decisive in rapid or blitz games. And, again, the World Championship match shall decide who is the best chess player in classical chess, not in rapid or blitz. For these two disciplines, we have separate World Championships starting in a few weeks.

Too long for sponsors?

The argument that a World Championship match with 24 games would last too long and no sponsors could be found nowadays does not convince me. If four games per week are played, such an event would end after no more than six (exciting) weeks. And we don't have or need a World Chess Championship every year. Every second or third year should be enough.

Magnus Carlsen is a worthy World Champion

Let me add one more thing, because I don't want to be misunderstood: Magnus Carlsen has, in my opinion, deserved to win. He would most likely have also been superior in the rapids to a well-rested challenger. Nevertheless, I would like to see the next World Championship match in a different, better format. This article is intended to initiate a discussion about this, not only among the numerous fans and lovers of the game, but also among the responsible organizations.


Stephan is a passionate collector of chess books and for yours has been successfully playing as an amateur for his German club. The former musician and comedian works as a freelance journalist and author in Berlin and in the Franconian village Hiltpoltstein.
Discussion and Feedback Join the public discussion or submit your feedback to the editors


Rules for reader comments


Not registered yet? Register

lajosarpad lajosarpad 12/11/2018 11:04

Mostly I agree with you, however, the match system where the champion is seeded is superior according to my opinion compared to the other approach. The reason is not reliability, but a semi-objective preference. I think that a guaranteed participation of the world champion in the finals makes the event more interesting. If at the final match no world champion is present, then we might have problems, like a 3-way-tie at the tournament, where the world champion is the third place, but having the same score as the other two. Consider the example when X is the world champion and he is shared first with Y and Z. If X scored against lower seeded players and Y and Z scored against higher places, or won more games with Black, whatever the tiebreak criteria, then the world champion was actually not weaker than the other two players, yet, he will not play at the final match.
Petrarlsen Petrarlsen 12/10/2018 03:38
@ lajosarpad:

About who is the strongest player, I think that there can be several different approaches: I agree that it is possible to consider that the strongest player is the player who can beat in a 1 to 1 match the strongest player besides him. But I think that it is also possible to consider that the strongest player is the player which can earn the greatest possible quantity of points when playing any random players. And, at first view, I don't think that any of this approaches are "false"; there are just different approaches to this question...
Petrarlsen Petrarlsen 12/10/2018 03:15
@ lajosarpad:

As you, I globally prefer the Champion vs. Challenger system, which has demonstrated to be a good system through many decades.

But I rather think that, nonetheless, a system, for example, with a World Championship tournament followed by a World Championship match between the two top-finishers from the World Championship tournament would give globally reliable results, because the "best player" has only to finish in the top-two of the tournament part to participate into the match. It seems rather unlikely that the best player couldn't be at least second in the tournament part, in my opinion...

By the way, if you have time for this and if you are still interested, I will return in one or two days to the "Champions Showdown: Improving the format" article (if this debate about the World Championship format stops rapidly; otherwise, when the current debate will stop), to answer you about the suggestion you made to improve one of the ideas I proposed about Chess960. I couldn't do it with the World Championship; this would have been to much at the same time, so I will return to it now...
lajosarpad lajosarpad 12/10/2018 07:53

Your point about the tournament element is being present even now is perfectly valid, however, this is normalized by the final match. If the challenger is good-enough to beat the champion, then there is no reason not to crown him. Yes, the challenger may have been lucky in the tournament, but in the match there is no third party. In the times of Steinitz and Lasker the method to choose who the challenger is was much worse, but Chigorin, Schechter & co. did not become world champions, because of the match. Of course, there can be better ways to determine who the challenger is, but in my view, if the challenger is better than the champion, that, in hindsight validates perfectly the challenger. If not, then we can wait for a better challenger. I prefer candidates matches to tournaments. The fact that there is a third person who might play lesser chess to different participants is too much of a white noise. If Karjakin would have beaten Carlsen, then he would have been a worthy new world champion, even if due to psychological reasons. Psychology is also important as a chess ability. Ivanchuk never became a world champion specifically due to this reason.

No system will be perfect, because we can never exclude the possibility that a player had bad luck or had no opportunity to participate. We need to have a system which gets near to perfect, but we will never achieve absolute perfection.

In a tournament Petrosian would have worse chances to defend his title due to the fact that his style is more vulnerable to the tournament problem. The person who is individually better than everybody else should be the world champion even if his dominance in comparison to weaker players is lesser than another person's dominance, who is lesser than the champ.
Petrarlsen Petrarlsen 12/10/2018 12:56
@ TRM1361:

"I would have 12 games and then if tied a coin is tossed to determine who gets white first and games 13 onward are sudden death. All at regular time controls, no rapid or blitz."

With Carlsen and Caruana, goodness knows how many dozens of games there could be with such a system before there would be a win, as they would necessarily be extra cautious, because of the title being at stake for each game.

And, as someone pointed out, if it was a match between Ding Liren and Karjakin, as they have a particularly defense-oriented style, it would probably be even worse...
TRM1361 TRM1361 12/10/2018 12:25
I would have 12 games and then if tied a coin is tossed to determine who gets white first and games 13 onward are sudden death. All at regular time controls, no rapid or blitz.
willembert willembert 12/9/2018 11:24
Agreed, the tournament format has its own problems. However, longish draw streaks is a problem intrinsic to the match format. The decision rule doesn't matter that much. That's why I'd prefer to have both.

Personally, I don't mind draw streaks. The main problem with matches in the past was that the Champion could keep the crown longer than they should have. The format is clumsy, takes a lot of time.It doesn't represent situations where a player dominates only a limited time.
Petrarlsen Petrarlsen 12/9/2018 09:15
@ lajosarpad:

About the "Champion vs. Challenger" system, I think that it is quite a coherent system and I agree with your explanations. But I think that a system in which each World Championship cycle is completely independent from the previous one (as SunriseK's idea) would give quite reliable results. (I don't see any significant reasons why it wouldn't, per se.) It is true that both the match participants would come from the Candidates (renamed, for example, World Championship tournament), but in a "Champion vs. Challenger" system, originally, the Champion was also selected by a Candidates tournament, and the Challenger comes directly from the last Candidates, so the "tournament element" is also quite present in a "Champion vs. Challenger" system.

The main difference, in my opinion, is that in the "Champion vs. Challenger" system, the idea of "continuity" is more present, as each Challenger has to beat the previous Champion to become the new Champion. But I think that the other solution, as for reliability, would nonetheless give good results.
Petrarlsen Petrarlsen 12/9/2018 05:37
@ lajosarpad:

One thing is sure, in my opinion: that the current system is not optimal.

Clearly, for example, with this system, Carlsen could be slightly inferior in Classical games, and be more or less certain to win thanks to Rapid and Blitz games.

And this wouldn't be such a problem if he participates in the match as the Champion (as, for all the reasons that we stated, the onus is on the Challenger to prove that he is better than the Champion in a "Champion vs. Challenger" system), but the big problem would be that, if he would lost his title at one point, he could quite well regain his title while being slightly inferior in Classical games, just simply because he is better than anyone else in Rapid and Blitz games!

And I wouldn't think that this would be logical at all. (In particular as there are Rapid and Blitz World Championships; what does it mean to become systematically Classical World Champion thanks to Rapid or Blitz games, when Rapid and Blitz World Championships exist? Plus the paradox that a player can be Classical World Champion thanks to Rapid or Blitz games, and fail to obtain either the Rapid or Blitz title - it is the present situation: Carlsen won the Classical World Championship in 2016 and 2018 thanks to Rapid games, but he didn't manage to win the Rapid title in 2017 as it was Anand who became World Rapid Champion this year...)
Petrarlsen Petrarlsen 12/9/2018 05:27
@ lajosarpad:

I will think more about your thesis...

For the moment, I rather think that you approach is logical and that the results it gives in terms of the systems which can be compatible with it are quite reliable, but that it isn't probably the one and only logical approach.

For example, it is true that a tournament win (your example with players X and Y, for example) can be rather unconvincing, but if Karjakin had won against Carlsen in 2016, it wouldn't have been very convincing chess-wise either, in my opinion (as I said, it would more or less have been a purely psychological victory, without Karjakin having demonstrated any kind of chess superiority over Carlsen).

And as for the limits of the tournaments, yes, these limits exist, but they can affect the Candidates tournament as well as a possible World Championship tournament, so it can as well be discussed if the Challenger is really the best player apart from the Champion, and thus, if he beats the Champion, it can be discussed if another of the Candidates' participants wouldn't have be more worthy than him of the World Title.

In fact, all and all, I think that there are several possibilities that are rather close, and which would give a quite worthy Champion.

As for the "match vs. tournament" question, I prefer clearly a match, but for another reason: because the World Championship matches permit us to have whole series of games (every two years, approximately) between players which are more or less the two best players of their time, and, in this dimension and for the "Chess Legacy", they are irreplaceable.
Petrarlsen Petrarlsen 12/9/2018 05:10
@ lajosarpad (2/2):

As for me, in this last example, I would find it perfectly normal that the Champion wouldn't try to win the game; what a match is about isn't about winning one particular game, but about winning the match as a whole, and a draw in the last game, with the "draw odds to the Champion" rule, means a win in the match for the Champion if the players were tied before the last game.

And thus, as, for Carlsen, obviously, a draw in the last game wasn't far from being the equivalent of a win in the match, I think that Carlsen's approach is normal; he could have done otherwise, but it was a logical and coherent approach to this game.

I don't say this to defend Carlsen unconditionally; I am not always quite a fan of his behavior in press conferences, for example (in particular when he has lost a game; he isn't always quite a role-model as for fair play), but, as for this 12th game of the 2018 match, I rather consider that his choices were a normal answer to the global match-situation (even if he could also have made another choice - I don't mean that his choice was obviously better than any other choice; only that it was a rational choice, in my opinion).
Petrarlsen Petrarlsen 12/9/2018 05:10
@ lajosarpad (1/2):

I am reading your new posts; I would just like to open a paranthesis about your phrase: "I have a problem with the world champion chickening out in the last game".

I think that Carlsen's logic, most probably, was more or less that a playoff was nearly the equivalent, for him, of a match-win.

And, in this case, this meant that a draw in the last game, following his logic (which has been demonstrated as being rather correct by the four-games rapid match of the playoff) more or less meant a win, globally, in the match.

So, in this context, it can be compared (even if it isn't exactly similar - but it isn't very different either) with a match using the "draw odds to the Champion" rule, in which the players would be tied before the last game, and in which the Champion would simply try to draw at all costs, to keep the title thanks to the "draw odds to the Champion" rule.
Petrarlsen Petrarlsen 12/9/2018 04:52
@ lajosarpad:

The post I answered, with my two last posts, was the first of your last four posts; when I wrote my answer, you hadn't yet written your three other posts...
Petrarlsen Petrarlsen 12/9/2018 04:50
@ lajosarpad (2/2):

And how does the current system (in very general terms - a Challenger who plays a match against the previous Champion) fares, compared to these possible systems?

A quite significant element about the current system is that it highlights the importance of the World Champion title: the World Champion is "apart from other players"; the new Challenger most prove that he is better than him to become the new Champion. Which wouldn't be the case with the other possible systems that I cited before in this post.

But is it as efficient to determine which would be the best player?

It would be closer to the last idea (who is the best between the two bests). But would it be less efficient? More efficient? About equal? I am not sure for the moment...

And the answer given to the question "Who is the strongest?" can be rather complex; for example, if Karjakin would have beaten Carlsen in 2016 (this would have mean, more or less: if he could have managed to be sufficiently solid to hold, for the last four games after his win in game 8), he would essentially have demonstrated a psychological superiority: he won the 8th game because Carlsen couldn't bear all these consecutive draws and "commited hara-kiri", and he would then simply have prevented Carlsen from winning any other game; the only superiority that he would have demonstrated would have been psychological (not even superiority in defense - Carlsen's only loss wasn't due to some weaknesses in defense; he simply couldn't stand anymore not to win a game, and lost all objectivity).
Petrarlsen Petrarlsen 12/9/2018 04:50
@ lajosarpad (1/2):

About your last post...: Food for thought!

It opens up, in my opinion, many questions...

What is being strong? Is it to obtain more points than every one else in average when you participate in chess competitions? (Then it would be more or less the Elo rating...) Is it to be able to be better than the other top-players on the planet? (Then, probably, the best way to determine this would be to have something like a 10-players round-robin - it would be a 18-games event; sufficient to have quite reliable results, in my opinion - between all the Top-Ten players of the Elo rankings...) To try to determine the two best players on the planet with an idea of direct comparison (it could be the result of the same "Top-Ten round-robin"), and to make them play one against another to see which is the best when you pit the two bests one against another? (With something like a 24-games match, probably...)
lajosarpad lajosarpad 12/9/2018 04:43
We could have a first to win N games with a limit, like the first to win 3 games will win the match, but at 12 - 12 champion retains. This would be logical and interesting.
lajosarpad lajosarpad 12/9/2018 04:43
We could use tournaments, like in 2005 and 2007, but in that case someone can become a world champion with a very unconvincing manner. Imagine the case when 8 players play a double-round-robin and there are two outstanding players, player X and player Y. X beats everyone, except Y by a margin of 2 - 0, but loses against Y in a manner of 0 -2, while Y beats X to 2 - 0 and everyone else by a margin of 1.5 - 0.5. This would mean that X has 12 points and Y has 11 points at the end of the tournament. Y has beaten everyone convincingly, yet, we crown X as the world champion, even though he was unsuccessful against Y. Another problem with tournaments is the possibility of some opponents offering less resistance to a player in comparison to the other. This could be a legitimate situation of pure luck, but could be deliberate strategy, like the one the Soviets had against Fischer. Tournaments will certainly increase the incentive to win, but they will add a huge randomity, which could lead to a never ending debate about whether the win of the champion was deserved or not.

We can have a tiebreak system, as we have, but since the total balance of situation, the current system is rewarding safe drawish play and in many cases punishing risk taking.

We could have a first to win N games without a limit. This would mean that players with exceptional defensive skills, like Karjakin could play virtually forever, not losing a game and waiting the opportunity to win one. If Karjakin would play Ding, the two super-solid players potentially would just wait for the other to err, which opportunity might come later rather than sooner. And if we wait for one to win 3 games, then a player, who has already lost two will play in an absolutely risk-free manner. If both won two games, then we will have a deadlock of draws, or a draw lock. Sponsors would not see much fantasy in organising a match which could potentially last for many months, not seeing the end.
lajosarpad lajosarpad 12/9/2018 04:25
If someone is already accepted to be the best player in the world, then we need a very strong direct argument to start to doubt that. This is why we call the champion a "champion" and the challenger a "challenger". At the next match it is logical to presume that the champion is the best, because he holds the title and has not lost the title, so nobody has proved to be better than the champion. Presuming that the champion is the best does not mean that we already claim to know that he is better than the challenger. If that was the case, we would not have a new title contest. The very fact that we have a title contest even though we already have a world champion is that we still think he is the best player, but we also have a challenger, who proved to be worthy to have an opportunity to prove himself to be the best in the world. However, the challenger is not successful in proving that he is the new best unless he beats the reigning world champion, hence the logic in the draw odds to the champion.

Now, we all agree that 12 draws is a problem. The problem is that both players were content with a safety-first approach and since both of them played it safe, we have seen wonderful defenses and missed attacking opportunities. I do not blame them for playing it safe, even though I have a problem with the world champion chickening out in the last game, but that's a different topic. We need to change the format if we want to avoid this from happening again.
lajosarpad lajosarpad 12/9/2018 04:11
Why do we have a World Champion title? It is to have a clear answer to the question who the best player in the world is. This is an important question, because if there is a World Champion, then his participation in a simul or tournament instantly raises the profile of the event. The chess world suffered due to the lack of a clear answer to this question between 1993 - 2006. As a result, we need to have a World Champion.

If we do not take seriously the title, then we weaken the power of the answer to the question of who the best chess player is. Élő points will not necessarily answer who the best player is. If a player has a safety-first approach and steers to a draw rather than taking any risks, then he might have a lower rating than someone who is very strong, but takes risks. When Topalov was number 1 in the world, he was not actually stronger than Kramnik. In their 2006 match after 4 games Kramnik led 3-1 and the match headed towards a total Russian domination, so Danailov and Topalov pulled out the toilet trick, a cheapo.
Petrarlsen Petrarlsen 12/9/2018 06:35
@ Leavenfish: An advantage of the "First to 3 wins" + 24 games' limit over the simple 24 games' match would be, for those who don't like too much the "draw odds to the Champion" rule, that the importance of this rule would be reduced, because, if a player would obtain 3 wins before the 24 games limit (which would be rather probable in my opinion), the match would end immediately. Whether, without the "First to 3 wins" rule, the match would continue, and could end up in a 3 wins to 3 wins tie (or 4 wins to 4 wins, etc.). So the probability of the Champion keeping his title because of the "draw odds to the Champion" rule would be lower (and, probably, rather significantly lower).
Petrarlsen Petrarlsen 12/9/2018 06:22
@ willembert:

"True, it is a huge privilege for the previous Champion to be seeded directly into the final. I can tell you the reason - simply tradition!"

It is true that this is deeply rooted in tradition, but, obviously, if a tradition is born at a moment, there must be some sort of a reason for it! (And it musn't be TOO absurd, or else, very probably, it wouldn't have survived - for example, if the "tradition" was that the World Championship match was to be played between the Elo World n° 9 and Elo World n° 10 players, this "tradition" wouldn't probably have survived; and for good reasons!)

And, in my opinion, there is only one possible coherent explanation (which can obviously be expressed quite differently) for the birth of this tradition: the idea that the World Champion is the King, and that to earn the Crown, you must dethrone the King; very simple and very intuitive. It would perfectly be possible to "build" a World Championship system around another idea (for example, SunriseK's system - the Champion is seeded into the Candidates, renamed World Championship tournament, and the two best players of the World Championship tournament play the World Championship match, the winner of the tournament having draw odds in this match), but I don't think that the present World Championship system can really be explained differently.
Petrarlsen Petrarlsen 12/9/2018 06:08
@ Leavenfish:

I didn't reproach you to use the 1984 match! (this would be more than strange, as I myself used it before!!)

What I meant was just that it is possible to use the 1984 match to argue about what can be a "first to X wins" match in general or a "first to 6 wins" match in particular, but that it isn't possible to deduce from this match what would have been the exact course of the match if the required number of wins had been smaller. (And I rather think this could have significantly altered the course of the match, but we will really never know...)

Perhaps Karpov was rather tired, but he wanted the match to continue, so he musn't have been very tired! Or he would have "jumped" on the possibility of having a match-interruption (all the more since he would keep his title in the meantime).

Per se, I wouldn't be opposed to a "first to X wins" match at the condition that a limit would be decided; it seems to me to be just the obviously reasonable thing to do. And the main element to decide which limit to choose should be in my opinion the resistance of the players; I think it would just be absurd to have a match being decided because a player will just finish to be completely exhausted and play like a patzer just out of sheer exhaustion.

Other problems are the sponsors and the venue; what would be the limit on that subject?

Globally, I don't think that more than 24 games would really be realistic.

So, in my opinion, a "first to 3 wins" + 24 games limit could be possible. But would it be better than just a simple 24 games match? Perhaps; for the moment, I don't quite know what to think about it...

One thing is certain, if there is an upper limit (even if this limit is rather high), it cannot have the effect of encouraging draws, because, at the end, one player will "end up in the wall" if he simply draws all his games.
Leavenfish Leavenfish 12/9/2018 01:44
Petrarlsen wrote:

"This is "rewriting history"; the players played with a "first to 6 wins" system, and to assume that, if the system had been "first to 3 wins", they would have played in exactly the same manner is just simply not possible."

Hey, you brought up the '84 match...I'm not assuming anything, just stating that when WINS are all that matter...wins happen and using your mention of that match to support my idea.

A problem with the '84 match was that too many wins were needed. I think Karpov lost 15 pounds and was physically unable to continue. Of course he was always frail...Carlsen vs Kasparov (of his age in 1984), now, that match would not have been called for that reason! But...6 wins is just too many in this day and age of advanced theory and computers.

3 Wins...I'm telling you, 3 is the magic number...
Petrarlsen Petrarlsen 12/9/2018 01:39
In fact, to say that a "first to X wins" format encourages the players to play for a win is a serious error in reasoning: with a "first to X wins" format, a draw doesn't change anything; objectively, it doesn't has any negative consequence for either player, so they can draw as many games as they want.

Whether, if the match has a limit to the number of games played without tiebreaks, each draw is NECESSARILY a "partial victory" for one of the players (and a "partial defeat" for the other one), because one of them is necessarily one step closer from the match victory (either the player who is leading the match or the player who benefits from the draw odds). And if all the games are drawn, the last draw means that the player benefiting from the draw odds wins the match, so this last draw is in fact a win for one player, and loss for the other, for match purposes.

If there is a Rapid and Blitz tiebreak and if one player is leading the match, a draw is a step in the direction of the Title for him (and necessarily a "partial defeat" for his opponent). And if the players are tied, each draw is favorable to the best player in "accelerated" (Rapid or Blitz) games.

So the "first to X wins" system is in fact the sole system which doesn't give ANY incentive to both players to play for a win (unless there are becoming too tired and want the match to end, but, in the 1984 match, neither Karpov nor Kasparov wanted the match to be interrupted, and this was after 5 months and 48 games, so the limit seems to be very high).

With the two other systems discussed in this post, there is ALWAYS one of the players which have something to lose by drawing a game, so, as for inciting the players to "play for a win", there are necessarily better than the "first to X wins" system.
Petrarlsen Petrarlsen 12/9/2018 01:11
@ Leavenfish:

In fact, the "first to X wins" system is WORSE and not better than the present system as for encouraging the players to play for a win: in the current system, the weaker player in accelerated games has objectively every interest to win the match in the "classical games" part; for example, if Caruana had assessed correctly the situation (which was certainly possible, in view of the respective Classical, Rapid, and Blitz ratings of both players), he could quite well imagine that he was running toward his own doom by drawing the two last games; he hasn't any reason to be happy with these two last draws, which, in practice, nearly ended all the chances he had to win the Title. Whether, if the match had been unlimited, these draws wouldn't have been a problem at all for Caruana; it would just have meant that the match would have been longer, and he would have had as many chances to win the match after them than before, more or less.
Petrarlsen Petrarlsen 12/9/2018 12:45
@ LLeow:

"if FIDE feels that the title should be decided in a match, the top two in the above tournament can play a short match (say six games), with the winner of the tournament wining the title if the match is tied. the last round in the tournament would be more meaningful and exciting. the match will also be more exciting if at any time at least one player feels the strong need to play for a win."

Why not; it is a coherent system (by the way, SunriseK already proposed it on another ChessBase page), but why would you want such a short match (6 games)? The tournament would replace the Candidates, and it would be simply possible to organize a long match (at least 12, but, if possible, 16 or 18 games) between the tournament's two top-finishers, and this match would simply replace the present World Championship match.

By the way, I am very much in favor of a match, because this gives a whole series of top-level games between two particularly strong players; a tournament which wouldn't be followed by a match wouldn't give this.
Petrarlsen Petrarlsen 12/9/2018 12:39
@ Leavenfish:

"When you say envoke the 1984 World Chess Championship against my idea of "First to 3 wins", you do not seem to remember (or maybe it just doesn't help your denouncement of my idea...) that Karpov would have won/retained the title after game 7."

This is "rewriting history"; the players played with a "first to 6 wins" system, and to assume that, if the system had been "first to 3 wins", they would have played in exactly the same manner is just simply not possible. Very probably, they would have taken into account the fact that the match would have finished much sooner, and wouldn't have played in the same way; anyway, no-one can know what would have occured in this case.

What is sure is that this match demonstrated quite cleanly that with a "First to X wins" system, endless series of draws are quite possible.

And this because, contrary to what you think, the "First to X wins" system is precisely THE system who encourages the LESS the players to play for a win: with this system, when you draw, nothing changes; you can make a draw, and wait for the next game... and do it for 17 consecutive games without any problem. With a limited number of games, this isn't possible. The only reason why the present system doesn't encourage the players to play for a win is the fact that there is a Rapid and Blitz tiebreak; at a certain point, the player can think that it will be better for them to wait until the tiebreak, and draw the last games; if the last game was REALLY the last game, the players would INEVITABLY have to fight, because, with this system, the end of the match is REALLY the end of the match, and not the beginning of a new (Rapid) one.
LLeow LLeow 12/8/2018 11:27
12 draws is a problem.

some may blame computers, but this does not seem fair. recent noteworthy computer matches stockfish vs komodo and alphazero vs stockfish show a reasonable number of decisive games. also some truly exciting games, rather than just games of attrition.

increasing the number of games from 12 to 24 would help in theory, but may deter potential organizers and sponsors. life today is much more fast paced in the internet age than it was 50 years ago, and the same is true for chess. time controls are faster. young players may wonder: what is an adjourned game? tournaments around the world include schedules with two classical games a day. let's not try to turn back the clock.

changing the format is the best option. the world championship tournament decided the title in 2005 and 2007 and should be revived. very simply put the best players in the world, including the title-holder, play for the world championship. there is a greater incentive to win in each game since +1 or +2 may not be enough.

if FIDE feels that the title should be decided in a match, the top two in the above tournament can play a short match (say six games), with the winner of the tournament wining the title if the match is tied. the last round in the tournament would be more meaningful and exciting. the match will also be more exciting if at any time at least one player feels the strong need to play for a win.
Leavenfish Leavenfish 12/8/2018 10:25

When you say envoke the 1984 World Chess Championship against my idea of "First to 3 Wins", you do not seem to remember (or maybe it just doesn't help your denouncement of my idea...) that Karpov would have won/retained the title after game 7. He even won 4 of the first 9!

First to 6 wins IS too long...just like Fischer's first to 10 wins.

3 is a good number in this day and age of computer help - it would tackled the 'draw yourself to the tiebreaks' problem (which to be fair IS a real problem if one player is much better at the quicker time controls)...be short enough to encourage more 'adventurous' play by the person getting the early lead (to actually END the match).

Being up a win...or two even, lets them play 'with house money' so to speak... and at the same time encourage someone going down a game or (especially) two to try to get in the match. Sure, a few draws might then let them try to stabilize things....but the person who is just a win away also has a say in things and can steer the game to more critical positions because...as I said...he is playing with 'house money'. Currently....they can just ride out their game or two edge thru game 12.

I think our true GOAL is to make for more interesting/eventful games for the fans while keeping the basic 'Classical Championship' intact...this does it more than any idea anyone seems to have come up with.
willembert willembert 12/8/2018 08:53
True, it is a huge privilege for the previous Champion to be seeded directly into the final. I can tell you the reason - simply tradition! It doesn't make much sense from the standpoint to efficiently determine the best player in the world. It is just a lucky coincidence that Magnus Carlsen is arguably the strongest player in the world. As I said, there were times when the WC was far from being the best.

If the WC matches are not efficiently determine the best player, why FIDE sticks to it? Because the format as such is very interesting. Two opponents prepare against each other extremely well for months, the actual match is just the culmination point of a gigantic duel which last for almost a year. We can normally enjoy great chess games with opponents who know each other very well due to the preparation.

Put to the extreme, WC matches is a completely different sport than normal tournaments. It may be as far away as rapid chess from classical chess. If the players are also great in tournaments - fine! It doesn't have to be so.
Petrarlsen Petrarlsen 12/8/2018 07:31
@ fgkdjlkag:

"(...) all your points and examples are proving my point!" (about the possible respective levels of the Champion and the Challenger, more or less)

In fact, no, it doesn't prove your point; it just proves what we all already knows: that the World Champion isn't necessarily the player who is IN GENERAL the strongest player in the World (...Karjakin could perfectly well have become the World Champion, WITHOUT using the tiebreaks and thus, in classical games, and, clearly, he wouldn't have been IN GENERAL the strongest player in the World at this time, for example...).

But if the World Championship's purpose was just to try to determine who is IN GENERAL the strongest player in the World, it would held no interest: for this, we already have the Elo rankings...
Petrarlsen Petrarlsen 12/8/2018 07:22
@ fgkdjlkag:

- "First of all, you are speaking in circles: "The Champion is presumed to be the "best player in the World" until a Challenger beats him." By whom? By you, obviously (...)"

No; not by me: it is the only possible explanation why the Champion is directly seeded into the final World Championship match. Otherwise, I will let you offer us another more plausible explanation...

Once more, if the Champion isn't considered (for World Championship purposes, as the winner of the last World Championship match) as the "Best Player in the World", I will let you explain us why he have the enormous privilege to participate directly in the final match without playing any other competition of the rest of the World Championship cycle!!

So, following you, it would seem that we have a player, the Champion, who isn't supposed to be better than anyone else (...I don't see why, if he wouldn't be supposed to better than the Challenger, he would be supposed to be better than anyone else...) who directly participates into the final World Championship match! More than strange!!!...
fgkdjlkag fgkdjlkag 12/8/2018 06:34
@Petrarlsen, all your points and examples are proving my point! I'm not sure how you don't see this. First of all, you are speaking in circles: "The Champion is presumed to be the "best player in the World" until a Challenger beats him." By whom? By you, obviously, but I have pointed out an example and you have pointed out several more examples where the world champion is NOT the best player in the world at the time of the next championship. So why are you presuming that the world champion is the best player in the world at the time of the next match, with all of the historical examples (not to mention theoretical reasons) to the contrary?

"And if you consider that the situation can have significantly changed from one match to the following match, and that, because of this, the Champion cannot be presumed at this moment to be the "best player in the World","

There have already been a lot of examples of this, so you should consider it, not me.

"Logically, considering as you seem to do that the Champion cannot be presumed to be the "Best Player in the World" [ fgkdjlkag- this is a fact, not my consideration] at the moment of the following match and as there cannot be any possible reason to presume that he should be better that the 2 or 3 other best players in the World if he cannot be presumed to be better than the Challenger, then you should choose a format in which each player participates on equal terms in the World Championship cycle;"
Why should I choose a format in which each player participates on equal terms, why don't you choose it? You are basing your arguments over format on faulty assumptions.
Jacob woge Jacob woge 12/8/2018 06:34
“If a draw odds to the champion was used in this last match, Carlsen would probably just have played the Berlin every game as black”

That does not make sense. Carlsen obviously considered rapid play-off equivalent to draw odds in his favour. Not a single Berlin.
Petrarlsen Petrarlsen 12/8/2018 06:12
@ Leavenfish:

"I am...mystified by all this.

How? Eliminate the 'fixed number of games format'....go to the first player TO WIN THREE GAMES."

As it is the second time that you post this idea, I will just remind you my answer to you on this same page:

"- 'Draws....

Hmmm...the ONLY thing that can serve to make leaning on them pointless...what could that be?

SOLUTION: First person to THREE WINS...wins.' (Leavenfish)

This is certainly why the 1984 World Championship (I remind you that it was a "first to 6 wins" match) featured so few draws; there was only a series of 17 consecutives draws (...oh, and, I nearly forgot, mere detail, another series of 14 consecutive draws...).

And these were LONG games: a 13-moves draw, another with 14 moves, still another with 15 moves, etc.; hard-fought and endless games (a bit like the first Carlsen - Caruana game...; oh; my bad, it was 115 moves and not 15; not a big difference: only 100 moves).

- 'But the secret...is to actually reward playing for a win.' (Leavenfish)

The problem is precisely this, with a "first to X wins" format: you can play (as Karpov and Kasparov, in 1984) 17 consecutive draws: no problems, the situation stays exactly the same. While if there is a limit to the match-duration and no tiebreaks, each game means that one of the players is one step nearer the abyss... So, with such a format, it isn't a solution to draw game after game; one of the players will end up losing the match, proceeding like this..."
Raymond Labelle Raymond Labelle 12/8/2018 05:24
"I believe there is a very simple way we can avoid boring/ambiguous draws in chess. Every draw offer by a player can result in one of 3 responses by the opponent. (a) Accept (b) Reject (c) Switch sides and play.

Under such circumstances, the player offering the draw needs to be extremely certain of the position being a draw and should be willing to play it on either side of the board. Also, most press conferences show draw agreements made in positions where both players think that they are slightly worse. If so, here's the chance on the draw offer, switch sides and win!!"


Brilliant! Kudos.
Leavenfish Leavenfish 12/8/2018 05:21
I am...mystified by all this.

How? Eliminate the 'fixed number of games format'....go to the first player TO WIN THREE GAMES.

Everything else is secondary: Yes, slightly shorter time controls would aid in this...FEWER freakin "rest days" as well.

But the solution lies around ENCOURAGING PLAYERS TO PLAY FOR WINS. Simply adding more games to some 'fixed' number...would not 24 draws be as lame as 12? LONGER time controls?? No...theory continues to advance along with the strength of engines to 'prepare'.

Draws can count for ratings...but should not count for anything in the end result of the match or serve to get us closer to the end of some arbitrary set number of games. Fishers 10 is WAY too many (these days a anyway)...but the first person to WIN 3 games seems workable.
Petrarlsen Petrarlsen 12/8/2018 10:45
@ Nezhmetdinov1919:

"So, I guess before we even begin to discuss the tie-breaks and such, we would first need to clearly state what is the purpose of the classical chess championship. If its role is to determine the classical champion, then there is no denying rapid games can be done with. But if its main role is to determine the absolute champion (to the extent we can determine it), and the observance of the classical time control is a less important goal, then FIDE's format is just fine."

In my opinion:

1) As there is a Rapid World Championship and a Blitz World Championship, it would be very paradoxical that the only time control for which there wouldn't be any specific World Championship would be the classical time control, the time control used in most super-tournaments and high-level competitions. I don't think that a "combined" World Championship would necessarily be a bad idea, but then there should be a classical World Championship AND a "combined" World Championship and not only a "combined" World Championship.

2) IF the present Classical World Championship was considered as some sort of a "combined" World Championship, it would a very bad sort of "combined" World Championship, because, out of 8 matches with this system (2006, 2008, 2010, 2012, 2013, 2014, 2016, and 2018), only half of them featured Rapid games (2006, 2012, 2016, and 2018), and none of them featured Blitz games. So it would be a "combined" World Championship only half of the time, and never between the three categories of time control as one of them is never used. Contrary to this system, a good "combined" World Championship should feature equally Classical, Rapid, and Blitz games.
Petrarlsen Petrarlsen 12/8/2018 08:41
@ tom_70:

"Just make it the first to win 3 games. Draws don't count. Eventually someone will tire."

"Eventually someone will tire." Yes, certainly, but who? The players? The organizers? The sponsors? Or the public?

In my opinion, by "ranking order", the firsts to tire will be the sponsors, after them, the public, then, the organizers, and last, the players (...for the 1984 World Chess Championship, after 5 months and 48 games, the players didn't seem to be so tired as that, as they both wished the match to continue...).
Petrarlsen Petrarlsen 12/8/2018 08:32
@ fgkdjlkag (3/3):

Logically, considering as you seem to do that the Champion cannot be presumed to be the "Best Player in the World" at the moment of the following match, and as there cannot be any possible reason to presume that he should be better that the 2 or 3 other best players in the World if he cannot be presumed to be better than the Challenger, then you should choose a format in which each player participates on equal terms in the World Championship cycle; for example a Candidates tournament between the 8 or 10 best players in the World (following Elo rankings), the two top-finishers in the Candidates playing the World Championship match without any advantage for the Champion as being directly seeded in the Candidates or the World Championship match.

By the way, if you consider that the situation can have changed significantly as to the playing level of the Champion from one World Championship match to the following one, it would be also quite possible that the Challenger's level could have changed between the Candidates and the following match; the difference isn't so great as that: 24 months from one World Championship to the following one; 8 months between the Candidates and the following match - only three times less. For example, the Challenger could have serious health problems between the Candidates and the match and have lost much in terms of playing level. So I think that your approach is an impasse: the Champion is supposed to be the strongest player in the World at the moment of the following match, as well as the Challenger is supposed to be the next-best player at the moment of the same match. Or else, the whole system must be changed.

And if the Champion is supposed to be the "best player in the World" at this moment, then it is normal that the Challenger have to beat him and not only to draw the match; thus, the "draw-odds to the Champion" rule is fully justified.