Greengard's ChessNinja.com

New FIDE WCh Cycle Proposal

| Permalink | 93 comments

FIDE has published a proposal for a new WCh cycle concept, a two-year cycle finished with a long match. (It's in a PDF at FIDE, the entire thing is below.) The most striking thing about this is the delivery. A proposal? A proposal "made by Mr. Berik Balgabaev, Assistant to the FIDE President"? This is like "Genghis Khan would like to negotiate." Balgabaev's boss, FIDE prez Kirsan Ilyumzhinov, is the alpha and the omega of FIDE. He can simply declare this the new shining path and have the presidential board rubber stamp it next week. (This is also where the possibility of Bessel Kok coming aboard to run a professional chess wing of FIDE based in the Netherlands will live or die. Kok is not terribly optimistik.) Since FIDE is supposed to be run by the federations, I much prefer "proposal" to the usual "decree." A pity it wasn't used when they suddenly changed away from a long-match last time.

The proposal itself is a reasonable one. It's an accelerated "tournament ko" qualifier with 16 groups of eight and then all the winners form two more groups of eight. Then there is a short final candidates match, which I assume is between the two final group winners. The match winner faces the world champion in a match of "12-16 games" the next year. This brings back the long-match finish FIDE added, then banished from the 2007 cycle when San Luis turned out to be a hit. This is of course great news, and not just for reasons of tradition. Long matches have the potential to be far more successful in every aspect if supported appropriately. I was never a fan of the KO's and certainly won't miss them. Moving to a two-year cycle also feels like an appropriate step toward modernization. Overall, a huzzah.

Since I left my cheerleader outfit at home, let's go to the negatives. The main problem, and this is not to say it's worse than the problems of other formats, is how small groups often lead to ultra-conservative play until the do-or-die final round. The strong players pick a likely victim or two and are happy with short draws with the rest. I'd guess +1 would win a few of the groups and most would go to a +2 player. A majority of the decisive games could be the under-2450 outsiders getting smacked around. The proposal document leaves a lot of holes, but it seems there aren't any tiebreak games until the final, meaning the group winners are going to be decided on formula tiebreaks much of the time. That's even more of an unsatisfying lottery than rapid tiebreaks. Players usually hate them, at least when they find out at the last minute that somebody else's result in the final round just cost them a major prize. Plus, half of the games ending in short draws would suck.

The use of tournaments instead of matches and the format of the tournaments would also raise the specter of cheating since there's only one real prize and an extra half-point will be enough in most cases. E.g., someone with -1 meeting one of the leaders on +1 in the final round would have little or nothing to play for. In the second stage this could mean the difference between a few thousand bucks for one and a shot at a few hundred thousand, or more, for the other. This sort of thing is why FIDE went to matches at Fischer's insistence, although back then throwing a key game was less a concern than general collusion among the Soviet players. (Later revelations proved the old saw that even paranoids have enemies.)

We'll have to put in some time to consider solutions to these and other problems you might suggest. Making sure there is more of a margin at the top is one way, although it only reduces those issues instead of solving them. And of course the main reason the groups are so small is to make for a quicker event, so simply saying few groups of more players won't be received favorably by FIDE. That said, having 10 groups of 12 and also taking the top, say, four second-place finishers and having a single 16-player final would cut down on the conservatism, the accidents and the cheating potential. I know speedy smaller groups are popular with organizers but they open up all sorts of issues. Having four whites instead of three is another one. More as I come out of my Goombay Smash-induced haze.

[All below sic -ed] The proposal on the change of the World Chess Championship Cycle

The main circumstances of the current cycle which make it “bulky” are its complicity [a Freudian slip for "complexity" no doubt -ed]. It is a complicated formula (for example, Candidates’ matches) that is difficult to be presented and advertised for potential sponsors. This is also one of the reasons for the top Grandmasters’ unwillingness to participate in the World Cup due to its formula.

1. The following simple scheme is proposed: in odd years (November- December) to organize World Cup with 128 players, the winner of which will play a match of 12-16 games with the World Champion in even years.
- Every year FIDE will be organizing one of the most important events (which is significant for potential sponsors and for FIDE PR activities).
- The FIDE shares from the prize funds of these events shall become more regular in timing.
- A more harmonious calendar (now: The World Cup and World Championship are organized the same year during the period September-December).
- The leadership and FIDE employees can concentrate on the most important event and will not be “distracted” during the preparation and organization.

2. It is proposed to change the formula for the World Cup: Initial stage 128 players are divided into 16 groups with 8 players each on the following principle: group 1 (numbers 1, 32, 33, 64, 65, 96, 97, 128), group 2 – (2, 31, 34, 63, 66, 95, 98, 127) etc. Free days – after 4th and 7th games.
Then 16 winners are divided into 2 groups with 8 players based on their rating: 1, 4, 5, 8, 9, 12, 13, 16 and 2, 3, 6, 7, 10, 11, 14, 15. Free days – after 4th and 7th games. In the final, there is a match of 4 games and possible tiebreak.

- We shall not have criticisms from the participants and mass media in respect of first round: “knockout is roulette”, “lucky – unlucky”, “did not sleep well”, “one blunder and you are out” etc.
- The number of accidental results will fall – a total of 7 games in the first two stages is enough for the top Grandmasters to prove that they are the best. Among the 16 group winners, there will not be weaker chess players.
- All the games will be played with the same time control.
- The tie break might be necessary only in one case – if there is a tie in the final.
- There will be no necessity to have a large number of additional local arbiters for the tiebreaks during initial stages.
- Weaker participants have opportunities to meet with 7 different players from various continents, get substantial experience, and not leave after the 1st round as under the current format.
- The top Grandmasters will have difficulties in explaining the reason for their non-participation.
- No more confusion with the hotels and air-tickets, there will be a firm plan for the departures of those who are knocked out.
- Organisers will have more chances for the media coverage of the World Cup participants.
- It is easier to negotiate good prices with the hotels, as the number of the nights spent in the hotels will drastically increase.

These reasons will promote the increase of the status of the event with mass media, possible organizers and sponsors and therefore, encourage the participation of all leading Grandmasters.


The main drawback of the system proposed by Ilyumzhinov is that in 8 players' tournament it might become drastically different to have 4 Blacks and 3 Whites or vice-versa

Yah, mentioned that one too. The Dortmund 2002 groups had the relative advantage of being double round-robins and evenly balanced with the small exception of Lutz. But having one under-2500 player in each group isn't so bad, and that's what it would be like based on Khanty Mansyisk.

I'm interested in what players have to say about the use of system tiebreaks for what is basically a winner-take-all format. A majority of the groups would probably go to tiebreaks with so few rounds.

Kirsan does well to label this a "proposal". He can sift through the reactions (such as that of Mr. Shirov) and make the necessary modifications.

Democracy is nice, but how many of the world's top 128 have even a 10% shot, in a 12-game match, of knocking off Kramnik? (Bat-cave paging Jeff Sonas.)

Whatever the format, wouldn't it be better to halve the field and have more of the games played between GMs with a reasonable title shot?

Another major drawback is that there are only 4 games in the final. This way, the number #1 rated player gets 7 games to prove his superiority over players rated 32, 33, 64, 65, 96, 97 and 128, but only 4 games to prove his superiority over player #2! That doesn't seem to make much sense. Probably the sponsors of the World Cup would want some sort of a result - like, to have a single winner of the event that would either be the champion or at least the challenger. But can't the final be made a 10-12 game event on its own - perhaps in Wijk or Linares (surely, the sponsors of those tournaments wouldn't object). That way the candidate finalists will have a month or two to rest and prepare for one another after the World Cup.

Also, how big a chance do players rated below number 64 really have? If FIDE wants to stop the "lottery" way of conducting the cycle, perhaps it is time to admit that players below number 64 do not have enough of a chance to be included. They would have a chance in a KO lettery, but if there is a serious effort to make it a serious event, perhaps it is good idea to limit the field to those who actually have a chance to win - and give those players more games against one another so that the results are more accurate. And that will save a lot of money on the hotels, too.

Why isn't it more regularly considered to have an odd number of players in tournaments to equalize the colors? Sure, everyone has a bye each round, and it is a little disadvantageous to have a bye int eh first round (as it isn't really a "rest"), but surely the advantages outweigh this minor disadvantage.

Can someone explain the rationale behind this?

The idea is (or should be) to make the system rigorous enough to make sure that anyone who can get through it has a de facto legit shot at beating the champ in a match. This is why the KO's sucked, even as qualifiers. This system, as is, would be tough enough to guarantee that, I'd say, and I'd hate to see the number of players cut. Let them get their moment in the sun; the quest to play in a WCh event has enormous prestige.

The conservative play thing is a fan/organizer issue, not a player issue, I understand that. (Unless you take the long view that what is boring for fans is eventually bad for players.) The cheating thing might fall into that category as well, since you rarely hear players voicing concerns about such collusion these days, although there are occasional flare-ups in the big US opens. Still, when you only have one winner per group and the stakes are so high it's hard to avoid those thoughts.

How about this? 14 groups of 7 each (that's 98 total). Winners of those 14 groups, split into 2 groups of 7. Winner of each of those play a 6 game match.

You could get the whole thing done in a month...

1: Opening ceremonies
2: Prelim, rd 1
3: Prelim, rd 2
4: Prelim, rd 3
5: rest
6: Prelim, rd 4
7: Prelim, rd 5
8: Prelim, rd 6
9: rest
10: Prelim, rd 7
11: Prelim, tiebreaks
12: rest
13: Finals, rd 1
14: Finals, rd 2
15: Finals, rd 3
16: rest
17: Finals, rd 4
18: Finals, rd 5
19: Finals, rd 6
20: rest
21: Finals, rd 7
22: Finals, tiebreaks
23: rest
24: rest
25: Match, gm 1
26: Match, gm 2
27: Match, gm 3
28: rest
29: Match, gm 4
30: Match, gm 5
31: Match, gm 6


I think this system is good enough that a very good player will get through - probably a top 10 player. But as far as making sure that the best player is selected as the challenger - I have a lot of doubts. I can totally imagine a guy rated around #8 in the world winning it.

It seems that FIDE is considering a huge change - they go from the world champion competing in a KO or a tournament with others to a system where a tournament is supposed to only decide a challenger, like in the old times. But they way this worked in the old times - the system was rigorous enough that it made sure that the best player became the challenger pretty much every single time (I am talking about 70s and 80s here, though even the early FIDE cycles were good, too). I am looking at this proposal this way: if we are going to seed the World champion directly into the world title match, we better make sure that his/her challenger is the best player. The FIDE system of 70s and 80s provided that, but the suggested system is not rigorous enough, in my opinion, and it gives too much advantage to the champion, who not only doesn't have to play until the World title match that concludes the cycle, but also can get a relatively "easy" challenger.

Of course, this advantage of the champion will be somewhat offset by the fact that he will have to defend every 2 years instead of 3, but still I'd rather have the world title matches be between the 2 best players in the world.

Anyway, this is better than KO of course, and it is better than a cycle where the champion has to defend in a tournament rather than a match. So this is a definite step forward.

What happened to the idea of having the world champion entering in the semi-final? With Mexico coming up, now should be the right moment to cut the priveleges of the World Champion.

I think the KO has worked reasonably well, the main problem being that not all the top players have participated.
From a sporting pespective I think it has less drawbacks than round robin stages, where collusion theories will always exist.

> now should be the right moment to cut the priveleges of the World Champion.>

The World Champoin is already doing a favor if he signs with FIDE to play for the title the winner of such tournament.
He can solve alone this "problem" anytime by simply agreeing to play any top GM who has ELO&tournaments credentials and who can secure
a good money prize, and thus bypass any would be "WCh cycle".

First having 128 players in the beginning is not necessary and not good. If the winner turned out to be someone way down the rank then the winner will not have proper respect.

If the winner is from the top group then why bother starting with so many people. It also does not seem fair that a top member might be knocked out because of losing a game against a low rated player. We saw this happen to Nakamura in the US Championships.

Look at the prestige of the knock out champs who were lower down on the list. no one really liked it.

We need to have one of the very high ranked players be the challenger. That is good for chess. We saw what happened in the US Championship when the previous champ Nakamura had a hard time with some much lower rated players.

This is all about finding the very best chess player. The good ones are already showing themselves at the top.

the top rated 16 players would guarantee a good challenger. you can raise it to 20. Maximum is 32.

Lets look at a RR with the very top 16 players. There will not be anyone real low on the list. Everyone will have the ability to win games. Hopefully these players are high enough up the ladder that none of them will want to give up even a half point. But then the best strategy is still to play for the draw.

How about 16 players doing matches. that should get down quickly to one challenger. They need to take the top 16 players from a most recent rating list. There can be a provision that the player has played a minimum number of games over the previous year to qualify.

Overall there are far fewer players and so the cost should be much less.

I vote for a 16 player Match knock out system.

If anyone wants to be the World Champion then they will simply have to get out there and win games and their rating will move up. For example Magnus Carlsen is given all the opportunity to move up. He simply has to win games. This is ultimately true for everyone.

If someone does not make it this time then there is a new opportunity in only 2 years. So work hard on your chess and move up into the elite top 16.

"the top rated 16 players would guarantee a good challenger. you can raise it to 20. Maximum is 32."

During the last cycle that used round-robin interzonals like these (1987-90) a total of 54 players were invited and 51 accepted invitations. I really don't want to see the number of players at the number of players at that level going down, and since this format has fewer hazards than the knockouts did, I don't think 128 is *too* bad.

I'd definately prefer less, though. Something like 54 (six groups of nine) or 72 (eight groups of nine) would allow each player four blacks and four whites in the opening. The really nice thing about having eight groups is that right afterward you can have something like four-game matches in the candidate quarterfinals and six-game matches in the candidate semifinals. Then you can have a longer candidates final and a championship match later. Or, you can just make candidates a round-robin or a double round-robin, but that has the same problem Mig described in the post.

I know, I know...there's a chance that a highly-ranked player loses out to a lower-ranked player. But:

1) "Higher rated" does not necessarily mean "stronger." The current rating system does a pretty good job, but there's uncertainty in all ratings, and FIDE's rating list responds sluggishly to changes in strength.

2) There's a chance that higher ranked players lose out to lower-rated opponents in all formats. Look through Mark Weeks's page on the interzonals from the 1970s and 1980s. Often there was a top player stunk in the middle or bottom of the crosstable, or an underdog making it through.

I agree that 128 is a lot, and that less would be better, but the old (pre-1990) system had a lot of players, too. I hesitate to cut the number down just because some high-ranked players could lose to low-ranked players. (If they're losing to lower-ranked players, do they deserve to be candidates?) I think something just above the 1980s levels is desireable, but I think it's a lower priority to having a regular cycle, having a match at the end of the cycle, and fixing the white-black situation that Mig and Shirov mentioned.

>I think this system is good enough that a very good player will get through - probably a top 10 player. But as far as making sure that the best player is selected as the challenger - I have a lot of doubts. I can totally imagine a guy rated around #8 in the world winning it.<

Yep, Russianbear captures my thoughts exactly. Imagine Anand and Topalov finishing half a point behind world's number 32 because they have not blown out world's number 128 both games, only once. Then you have a four game match for the title shot: in modern times this is an equivalent of a one game knockout!

I can think of other problems, such as how Tal memorial tournament ended up with three players tied for #1--in this WCC proposal with only one of these player from each group going into the final match, that would make the tournament pointless.

The problem isn't that the system would not produce a worthy challenger. The problem is the system is going to produce a challenger whose superiority over the other competitors will not be proven in any way, whose win will probably come down to one or two games played and whose accomplishment does not make us think of him as worthy of a title shot.

I find it hard to get worked up about a new cycle. Whatever we think is the case now, will only get changed depending on the whims of Ilyumzhinov.

>I find it hard to get worked up about a new cycle. Whatever we think is the case now, will only get changed depending on the whims of Ilyumzhinov.>

yeah, the single best thing that FIDE can do is to stop producing "alternative" WChamps and thus nurture confusions and illusions.
The "WCh cycle" is the problem of the world champ and it is always solved in the end by his choice whom to play.

This format is complete garbage. I see very little difference between this and the original FIDE KO in 1997 (the one with Karpov seeded into the final). Half the time the challenger will end up being someone like Kasimjanov, which means the champion gets an even bigger advantage than under the old system.

I would adapt it as follows:

The champion (seeded 1) starts from the beginning like everyone else.

Make the first round a double round robin.

The 16 qualifiers split into 4 groups of 4, who play each other 6 times (18 games).

The four winners move on to a semi-final matches (16 games) and the winners play the final (20 games).

I'm sure this could fit in a two year cycle. Anyway I'm sure this won't be the last proposal FIDE makes between now and Mexico.

In response to GM Shirov's concern about the 4/3 colour splits, here is a suggestion. Make the lowest rated player always be #8. Flip a coin as to whether the colours are normal or reversed. Thus the strongest players 1-7 would always have 3-3 splits against each other.

I am glad that the proposal has 1 winner advance from each section. I have been advocating since 1988 (in the chief arbiter's report of the World Active/Action/Rapid Ch.) that FIDE abandon systems which have multiple qualifiers from a single section, e.g. a 100-player Swiss with the top 9 moving forward. It just encourages short draws among the contenders.

I share Mig's and Bobby Fischer's concern about possible collusion, a concern which also could have been mentioned in relation to the San Luis format. But with 16 events splitting the chess world's attention, and a wider mix of players, the chances of collusion are *more* than 16x as great.

These 16 tournaments could be spread out in space and time. In the old days, Candidates Matches were played all over in Europe, often in small cities. I wonder if FIDE can establish enough confidence / reliability / credibility in Europe that such a thing might happen again?

Has anybody asked Fritz to see if he's agreeable to this format? After all, he's the new World Champion.

Give respect where respect is due -even if it's a computer.

Re: Jonathan Berry. The lowest rated player will not necessarily lose all the games but should normally be more vulnerable as Black, so this idea doesn't solve the problem. Besides there are no weak players on the second stage at all.

I also think that the knock-out system is not wrong at all when you have to determine one clear winner - after all he beats all the guys who on their turn have beaten all the other guys.


I will write an official reply on my behalf to FIDE proposal in a couple days from now, but there are several things I think that need to be cleared.

Firstly, overall FIDE trend for the last decade was to lower world champion's privileges and lately champion had to participate in candidates semi-finals, or as in the current cycle in the final tournament of the eight. The unification match was simply an exception. At this day of computer information, it is simply unfair for the champion to sit and defend his title once every two years or so while reaping all the benefits of holding the champion title, while the rest of the players have to struggle financially. So there's no question but that the champion has to actively participate in the cycle and be seeded into at least a quarterfinal stage (similar to the rules for 2005-2007 cycle where he was seeded into candidates).
Secondly, the new format is not acceptable as it allows a possibility of collusion in each minitournament ( as already being pointed out), wherease knockout format determines a winner in each match irrespective of other player's result. The knockout formula in that respect is more fair than the proposed new one.

To allow the proposal stand in its current form is an attempt to stop chess progression for a more democratic and fair competition. We cannot turn back the time and let the world champion revert things as they were back hundred years ago.

Gata, you make some very good points, as did Mig and many others who wrote. However, I think there is one point that is the most important.

The running of the World Chess titled tournaments and matches should be done by an independent organization. (I, personally, would prefer one run by Bessel Kok as previously announced by Kirsan and agreed by his Presidential Board).

The Organization and Format for all these titles, including the World Chess Championship, should be worked out solely by the separate entity, not by FIDE. In this way, they would be stable, if nothing else.

A lot of thought has to be put in to these titles, and this includes the welfare of the professional players and also the ability to earn money for FIDE and capitalize on the popularity of chess around the world. (Chess is very popular, but I doubt if even 2% of those that play and love the game have heard of Vladimir Kramnik or Fritz, let alone Mig's Daily Dirt or FIDE).

While the organization headed by Kirsan Ilyumzhinov is in charge of any Chess Title, or even the Rating System, then there will be changes at his whim and many people will suffer.

> We cannot turn back the time and let the world champion revert things as they were back hundred years ago.>

It changed to FIDE control in 1948 with the death of Alekhine and it reverted back when Kasparov left FIDE in 1993.

Therefore in the history of WCCs we have
120 years (1886-2006) out of which for 75 years (1886-1948, 1993-2006) the title was controlled by the World Champion and only 45 years (1948-1993) when it was by FIDE.

>but I doubt if even 2% of those that play and love the game have heard of Vladimir Kramnik or Fritz>

what ??

Kramnik-Fritz match (and the result) was in the news of New York Times, Der Spiegel, Associated Press, Reuters, etc.

I doubt there is anyone who hasn't heard even he has no interest in chess.

To be credible, the WCC cycle has to crown "the world's best long-match player."

After his defeat in the 16-game London match, even Kasparov acknowledged that Kramnik was the new champion. But what if the London 2000 match had been a WCC quarter or semi-final and had been chopped off at Game Six or Game Eight? Kramnik would have been your winner, at 1-0. But sitting on the sidelines, Kasparov could reasonably argue that his short-match loss did not invalidate his claim to being the world's best long-match player.

Forced to sacrifice one of our priorities, fairness-to-GMs or WCC-title-credibility, we'd clearly have to sacrifice the former and allow the WC to defend his title in the final match. If, however, the quarter or semi-final consisted of at least twelve games, a WCC's defeat at that level would credibly extinguish his claims to being the long-match champ.

Well, how many people read the New York Times? From Circulation figures, probably less than 1%.

In the US, at least, the number of folks who read any newspaper has really gone down.

There are probably a few million people in the US who know the rules, and occasionally play. However, 95% of them are not members of the USCF, nor active in a Chess Club, or structured community. Thus, they may not be exposed to knowledge, such as being aware of Kramnik or Fritz's status.
>but I doubt if even 2% of those that play and love the game have heard of Vladimir Kramnik or Fritz>

what ??

Kramnik-Fritz match (and the result) was in the news of New York Times, Der Spiegel, Associated Press, Reuters, etc.

I doubt there is anyone who hasn't heard even he has no interest in chess.

Posted by: Ovidiu at December 9, 2006 17:19

If FIDE wants to give 128 players a chance to win the title, that's fine. However, those who are ranked below, say, 50th in the world, ought to first play in a Qualification event, which could be a large Swiss. Then, you could have 10 or 15 Qualifiers, who would then be placed in RR tournaments described above.

There is no good reason to have such short tournaments, consisting of 7 games.
As has already been mentioned:

Resorting to Tie-breaks to determine who advances is problematic.

Color inbalances take on added significance.

Collusion/results fixing is more likely.

The other problem is that while it is possible for a Top seed to come back from an early loss, it is still quite unlikely.

The fact that Tie-Break scenarios can be readily calculated with 3 or 4 rounds to play might mean that it is just as likely that a player (with poor tie-breaks) will be spurred to fighting chess, in an effort to win the section outright.

One possible improvement would be to make the 8 player events DOUBLE RR--but that would also entail playing an extra 7 games. By then, the section might have a runaway leader.

Maybe they ought to keep a KO formula for early rounds. The "Luck" factor can be mitigated by making the KO part of the event Double Elimination. That would involve having a "Repechage" bracket, and an extra round of matches. But there would be a clear winner, and no incentive to throw games.

FIDE could utilize KO with Repechage to get to the 16 "Candidates", and then have a RR event when you have winnowed out the chaff.


I am very interested to hear thoughts from Alexei Shirov, Gata Kamsky, and any other top player who is reading this, regarding whether you like having some sort of tiebreaker rules, such as most wins, best score with Black, combined score of people you beat, whatever, rather than a rapid/blitz/armageddon tiebreak. My personal feeling is that having tiebreaker rules allows you to adapt your strategy accordingly during the last few rounds of classical games, and thus the winner really can be legitimately determined during the regular part of the tournament, just in classical chess, without needing to resort to rapids except in very extreme cases. But I'm sure there is an argument that tiebreakers are too hard to figure out early enough in the tournament to do anything about it, or that players prefer to resolve it over the board even if they do have to play faster, or players with certain styles get a big advantage from those tiebreak rules, etc. That's why I would like to hear where various top players stand on this question. Because if they are fine with the rapid tiebreaks, then it seems more logical to have something like a double-elimination knockout tournament made up of four-game matches, rather than these mini-round-robins.

It is common ground that the object of the candidates cycle is to produce a worthy challenger who has, moreover, proven his/her superiority over the other candidates. There are therefore two requirements that the challenger has to meet, the ‘worthiness’ requirement and the ‘superiority’ requirement.

It is not necessary, IMHO, if it is indeed possible at all, to regiment the ‘worthiness’ requirement, let the cream float itself to the top, so to speak. I suggest to meet this ‘worthiness’ requirement through pre-selection restricted to the top ten (or so) of the rating list, if only because of a lack of a better measure of ‘worthiness’. (Pre-selection here basically means that the higher rated player has the right of first refusal to participate in the world championship cycle)

The ‘superiority’ requirement can then be met by KO matches between eight candidates so pre-selected. The challenger could then claim to have beaten the man who beat the man who beat the man.

I would therefore vote for three rounds of such KO matches to determine the challenger in, say, the odd years, with the ex-champ, if there is one, be seeded as one of the eight candidates, but with no further seeded candidates. The world championship could then be held in the even years. All matches ought to be 12-16 games long. Tie-breakers in the candidate matches only, the champ has draw odds.

The big problem is how the heck could such an utterly incompetent FIDE possibly carry out three rounds of KO matches every other year? But the above scheme does produce a worthy and superior challenger, and, IMHO, will even most likely produce the most worthy and the most superior challenger. This, in turn, will bolster the prestige of the world champion (which is another big problem according to a conspiracy theorem according to which Kirsan wants more prestige than the world champion)

Alexei Shirov

You are right. The suggestion of giving the weakest player pairing number 8 helps but does not eliminate the problem in the first stage, and does not help in the second stage.

Since numerical tiebreaks are envisioned for the first stage, the use of any of the following: most points with Black, most games with Black, most wins with Black; as the first or second tiebreak could eliminate or even overbalance objections based upon colour.

I too prefer knockout; in 1988 I suggested 1 v 128, 2 vs 127 etc. Since FIDE switched to that method, fewer GMs have objected to KO. But even with KO matches, luck can be a factor. For every Tal, there is a Lutikov waiting to be matched against him! Also, the tactic of two (or four) quick draws followed by a decision in Blitz is annoying to fans. Maybe a Sofia rule....

Jeff has anyone considered using as the tiebreak that the person with the highest elo moves on to the next round.

That would give some advantage to the players with the higher elo but that is basically what we want. We want the challenger to be the one with a very high elo.

This will also assure that the highest elo players move on giving the strongest opposition in the next round. This gives the higher confidence that the eventual challenger earned his postion.

With the tiebreak system so simple everyone will always know who might move on in the case of a tie at the top. This is not as silly as one might think. For example in a swiss system, the everyone is ranked by elo with the highest elo maintaining his position with those with equal scores.

Why is FIDE so insistent on having 128 players? Why not have just 32 players and play double round robin sectionals?

>Why is FIDE so insistent on having 128 players? Why not have just 32 players and play double round robin sectionals?>

Because they want to be "democratic", "fair", they
are into the business of "giving everybody a chance" as if we don't know the 10-12 top players
who would stand a chance against the champion in a match.

It is just ridiculous, it is working hard to solve imaginary problems.

QUOTATIONS from earlier posts, about the genuinely intolerable problem of giving some tournament players a higher White/Black ratio than other players:

[] issues. Having four whites instead of three

[] drastically different to have 4 Blacks and 3 Whites

[] Something like 54 (six groups of nine) or 72 (eight groups of nine) would allow each player four blacks and four whites in the opening.

[] In response to GM Shirov's concern about the 4/3 colour splits, here is a suggestion. Make the lowest rated player always be #8. Flip a coin as to whether the colours are normal or reversed. Thus the strongest players 1-7 would always have 3-3 splits against each other.

[] Why isn't it more regularly considered to have an odd number of players in tournaments to equalize the colors? Sure, everyone has a bye each round, and it is a little disadvantageous to have a bye in the first round

[] Color inbalances take on added significance.

[] One possible improvement would be to make the 8 player events DOUBLE RR--but that would also entail playing an extra 7 games.

[] tiebreaker rules, such as most wins, best score with Black

[] use of any of the following: most points with Black, most games with Black, most wins with Black; as the first or second tiebreak could eliminate or even overbalance objections based upon colour.

The much simpler "pie" fairness rule would make all these color imbalance problems vanish.
Yeah I know, rule improvements are changes, change is unthinkable, so change proposals are pie-in-the-sky.

>>At this day of computer information, it is simply unfair for the champion to sit and defend his title once every two years or so while reaping all the benefits of holding the champion title, while the rest of the players have to struggle financially.<<

Sure, Gata. It is completely unfair to reward the winner of the championshp cycle. After all, all he has done is proven himself to be the best player in the world.

Eliminating the champion privilege is equivalent to picking a tournament every few years and saying the winner of that tournament will be the next champion. It eliminates the history of the title and makes the winner seem like flavor of the month, since it in no way demonstrates his superiority to the previous champion.

This FIDE format would only make the GMs have to play two tournaments (including one in which most of his opponents will be ranked far below him and should not provide much difficulty) plus a four game match. I doubt that's really tiring over the course of two years.

Chess is the battlefield of titans. It holds our attention because we are in awe of the men's accomplishments and ability. King of the hill system, as practiced since the age of Steinitz, reinforced our perception of that greatness.

A few simple things to fix about this proposal:

1. Way too many players. Do we really need to give #128 an opportunity to play for the title every two years? It seems that you could cut the number of players in half and not have anybody complain.

2. Everybody would like to see the top players make it to the penultimate stage. (second group)

3. Too many draws?

4. White/black odds--I think this was answered excellently above with the proposal for an odd number of players in the group. This would also give the players an extra rest day, no?

5. Final match too short.

Hey, how about this? Three new stages:

Stage #2 is a 9 man group. 5 invitations open to the world's top 5 players. 4 more slots open to winners of either this kind of group system (4 groups of 8 lets you have world's 34 with a chance to qualify) or a simple short knockout similar to World Cup or Libya. What I like about this format is nobody would really care too much about who the out of the players that are not in top five are in the second stage, but also it still allows some youngster or a low-ranked playing really well to make his way up.

With only one slot open in each group 2 in stage 2, I think there is too much of a possibility of two great players ending up in the same group while the other group leaves the audience not so excited. With all 9 players at Stage 2 playing each other, it will be easier to gauge relative strength. In this format, I also don't think the fans will complain as much if #2 beats out #3 on some stupid tiebreak or better draw--after all, the #2 will still have to play #1 and the chances of both 1 and 2 being unimpressive are rather small.

Final match should be 8 games. An 8 game match might not always determine the best player, but it will always show if one of the players is much better or if they are about equal. If it's the former, problem solved. If it's the latter, I would suggest it doesn't really matter if and which one is slightly better. This is to determine the challenger, not the champion and the loser will have another cycle in two years to try his luck in.

>Chess is the battlefield of titans. >

In the public's imagination. In reality they have to use the bathroom quite often and are not foreigners to mates in one move.

I too have a proposal. Let those few (10-12) GMs who have the career credentials (not just one lucky tournament) to first seek sponsorship and then challange the champ. The WCC is not under the obligation to play anyone, he must be given reasons to do it, not orders and not from a petty dictator (also an ordinary criminal) and "organizations".

I'm not sure I understand what people mean when they say that this format will, with some likelyhood, not turn out the strongest but only _a_ strong player.

What do you want? Is "strong" here supposed to be measured in Elo rating? Then why have a tournament in the first place if you're not going to respect its winner if he isn't #2 on the elo-list.

Tourney results are always influenced by luck, current form, how the opposition fares etc. And it will be relatively rarely that in a close field players finish just as their rating would suggest. So you can't have it both ways: If you want a tournament qualifier, then there'll be cases where the challenger is someone who had the tourney of his life.

If you don't want that, simply go by Elo.

I think the proposal is good enough. To bring back the old match tradition at the end is good news. Whether 128 is too much or not, is not the most important question to me.

I don't know where the cut off is but if you're not in the top 100 in the world then I really don't think you ought to be in a world championship event without pre-qualifying. It would probably look like top 16 + 16 qualifiers that should be plenty. (I'd probably favour top 8 + 8 qualifiers (for the format below) but I think that's an issue for the players)

Once you get the numbers down you can have longer and more classical formats.

But we're all trying to find a way round two problems. 1) Lack of space in the chess calendar which means that you have to keep the number of separate events down. 2) Lack of sponsorship for all but the most attractive of events and that surely doesn't include an event with Uncle Tom Cobley and all.

I would still say big swiss interzonal, candidates tournament plus either a straight world title match or match tournament that's just three events the last two of which should be relatively easy to finance. You could get that done every two years and people would I think respect the result.

The main problems with Candidates matches are that you need three or four separate spaces in the calendar and sponsorship has to be found for each stage between matches as people don't like sponsoring events where they don't know who is playing.

I understand FIDE's motivation for their proposals but I still think its too random. There are simply too many players in these events they propose. I suppose it must be a payoff for the votes they get from lesser nations. But if they want to promote chess in these countries then I suggest they have a scheme to get them experience world wide. The chess world championship is in too fragile a state to use to promote lesser nations chess in this way and I'm sure there are more efficient ways of doing it.

"In reality they have to use the bathroom quite often and are not foreigners to mates in one move."


Very cruel of you to remind Bulgarians, in every thread, that it was such a frail and imperfect creature who defeated their hero.

"I don't know where the cut off is but if you're not in the top 100 in the world then I really don't think you ought to be in a world championship event without pre-qualifying."

The number is chosen because it is 2x2x2x2x2x2x2. The number below is 64. This is important for a ko-system that can be cut down to a final. You can also chose 64 if you want, but where's the difference?

A comment about the 128 number that people say it is too much.

The first round is 16 groups. This means that number 1 player has to win a round-robin tournament of 7 rounds with closest rival the number 32. If he can't, he does not deserve to continue... Only the players close to 16 will have `serious' competition and that is only fair(since no16 is not that better than no17).

It is most likely that in the second round the best players will proceed. The second round swiss will produce a good winner. Now he may not be the best, but then why was Topalov the best. He won a tournament as well. In all sports, all formats you may have the weaker player winning..(even in maches.How many times a single blunder determines the winner...)

Also having one only place for qualification will make the games to have less rather than more draws...

The only real and big drawback, is that at the final rounds there will be unmotivated players playing against people that may qualify and thus affect the result. Also prearranged results may also appear.

How does any of this improve on the old FIDE cycle?

Eight players (top six from interzonal tournament plus top two from the old cycle) play candidates matches.

"How does any of this improve on the old FIDE cycle?"

It doesn't but a series of candidates matches I don't think would work anymore.

I was being conservative with anyone outside the top 100, I think the figure is a lot lower than that. Of course everyone should have a chance but only through pre-qualifying probably through a big swiss. If you're going to do something along the lines they've suggested then 32 is that absolute maximum number of players that should be taking part. A sprawling 128 player event gets little or no publicity, its too hard to report. Right now there are probably three players Topalov, Anand and Kramnik at the top, then a second group in the rating list of about four down to the early twenties who are much of a muchness, all very, very good. Then a third group down to about 50 who on their day can be very good. You're probably going to have to draw an arbitrary line in the second group of players. Unless you take the top 20 + 12 qualifiers from an interzonal style tournament. Which might just work.

>Chess is the battlefield of titans. >

>>In the public's imagination. In reality they have to use the bathroom quite often and are not foreigners to mates in one move. <<

The mate in one move is the first time this happened to a world champion in my memory. But I understand your point. It's only when champions are able to never make a mistake and have perfect bladders that they may be considered great.


The idea here is that somebody ranked 6th or 7th (and presumedly not a top challenger in this format) could easily win in this format. It's true that tournament results are dependent on luck. But the point is that in this format it would be particularly easy for a subpar challenger: make it to the second round against nobodies, get an extra half point on tiebreak and then win a four game match.

If the funding of chess events is in so parlous a state, how does it make sense to try and finance a tournament with 128 players in it? (About 96 more players than is probably necessary to determine the identity of the best player in the world.)


Why do you prefer "parlous" to "perilous"?

Enquiring minds want to know.

The poet in me just likes the flow better...

(So enquiring minds will be disappointed)

Yes, it does sound better.

What about in the FIDE preliminary tournament (the 128), you have someone advance ONLY if they win their group outright? No tiebreaks needed... If there is a tie from the top spot, then no one in that group advances.

Too radical? Too brutal? Possibly, but you'll see a lot more players going for wins.

>If the funding of chess events is in so parlous a state, how does it make sense to try and finance a tournament with 128 players in it?>

Kirsan the Great in the main sponsor of chess nowadys and he tries to maintain his postion as "president of international chess" simply by paying.

It is a vicious cycle (sic). Serious sponsors recoil from chess rather than having to deal with such gangster and this in turn strengthens Krisan's postion, makes his money even more valuable for the unemployed "chess professionals" since there is no other choice.

It has always been difficult to make a living out of chess but this marriage of Kirsan with chess is like living in fantasyland. FIDE worked well in the 1948-1993 period because of another economic fantasy : most of the chess players of importance were soviets and in fact supported by the state, i.e., living not because some genuine sponsorship/economics mechanisms.

This illusion broke apart with the fall of Soviet Union but suddenly another one emerged to "save" chess in the person of the another sate sponsor :Kirsan's Kamlykia.

Not until these artificial structures disappear the chess will become realistically professional (even if this could mean less top players unrealistically hoping to make a good life from chess).

@Jeff Sonas: I wonder what the estimated probability is for the players ranked 1 to 128 respectively to win the new FIDE WCh format. Jeff, I would very much appreciate if you could give us at least a rough estimate based on the current elo distribution.

Does anyone know if this book will ever be published?

The Attacker's Advantage: How Life Imitates Chess
by Garry Kasparov

We shouldn't take for granted the assumption that the best format is a scientifically rigourous method of chosing the undisputable "best player". Attractiveness, drama, and entertainment are also criteria that can be legitimately considered. Especially if obtaining sponsorship for all or part of the cycle is required to make it a reality. How many of the teams that play in the FIFA World Cup have a legitimate shot at winning it. But just the journey to qualify for the tournement at all draws in a lot of fans form all regions of the globe. And fans draw sponsors. As far as tie-break formats .. change the rules to eliminate agreed-draws, and you don't have to contrive so many tie-break formulas.

>Does anyone know if this book will ever be published?
The Attacker's Advantage: How Life Imitates Chess
by Garry Kasparov >

No clue but Yermo may write a truly useful one :

"Playing the Latvian Gambit and yet surviving : How can you be a chessplayer and still make a living"

The only group of 9 that makes sense to me is 9 groups of 9 players or 8 groups of 9.

at the end of the first round there are still 9 players for the last round of 9. Or else the 2nd round has to be different from the first round.

for example 8 groups of 9 players. at the end of the first round we have 8 players who can then play Match elimination.

well I guess you could do 5 or 6 or 7 groups of 9. then the 2nd round might be a double round robin of the remaining 5 or 6 or 7 players depending on how many rounds you want.

double round robins of 5 players would be only 8 games. 6 players would give 10 games. 7 players would take 12 games for double round robins. all within working distance.

I still say take the top 16 elite players and do matches to determine the challenger. Simple and effective. The higher rated player gets draw odds. so we play 6 games. drops to 8 then 6 games drops to 4 then 6 games drops to 2. these 2 might have a 10 game match draw odds to the higher rated.

This is more days but many fewer people and the cost should be much less. The official challenger will definitely be someone who everyone will consider is a good challenger. He will have won 4 matches to become the challenger. well I guess he could have drawn all 4 if he were a top seed.

It sounds like a good idea, but I agree with the people who think it's not challenging enough. Here is my suggestion:

A field of 124 players is composed.

It is separated into two groups of 64. (Players 1,3,5 etc. according to rating go into G1, players 2,4,6 etc. go into G2.)

In both groups, four-game matches are arranged between player 1 and player 64, player 2 and player 63, etc.

The following is done in both groups:

* When the group is cut down to 32, it is divided into two subgroups of 16, where all play all.

* The last 8 from both tournaments go home and the winner of tournament A faces player 8 from tournament B in six games. Number two from A plays six games against number seven from B, etc.

* There should be only four players left. They play in a quadruple-round robin against each other. Tie-breaking games are arranged if necessary to determine the sole winner.

By now, both groups of 64 have produced a winner. So, only two players left. A 12-game match between them takes place.

The winner faces the World Champion in 24 games.

That makes around 73 games in total, not counting preliminaries and tie-breaks. I imagine two years should be enough to play 73 games.

That's what I'd ideally like to see. But even if they stick to the "proposal" they've come up with and don't change it in the slightest, it'll still be an improvement on the present cycle, so I'm all for it.

Sorry, I just noticed a mistake in my previous post. I've obviously emitted a phase of the cycle I described. Between asterisks 2 and 3, there is another one:

* Eight players left. They are separated into two double-round robins. Two players per group go through.

Which makes the total number of games 79, not 73. Sorry again.

Americans, don't vote Paul Truong (member of Susan Polgar's team) into a position of power at the USCF. This man has been caught computer cheating on various chess servers in the past (including, but not limited to, Internet Chess Club and Chessnet). He abused his position of power helping US Chess LIVE to collapse. He has a history of broken promises and harassment. This is all facts, the information is available to those who seek it and ask questions. There are people who know what I'm talking about. Susan Polgar is pretending none of it ever happened. You need honest, up standing, right minded people in positions of power.. not egomaniacs that cannot be trusted and are allergic to constructive criticism.

After reading all the proposals in the above, the simplest would be a San Luis style DRR to determine the challenger. Just like Topalov was determined as the challenger of Kramnik in Elista. Mind you, don't go by the official labels of the events, rather, go by what actually transpired.

We should use Mexico as San Luis II 8^O !!!

Perhaps we should convince Kramnik to give
Topalov a rematch chance. Not for the sake of "fairness"--doing good deeds and granting a place in heavens-- but because it would give us something exciting to watch and Topalov can secure sponsorship.

Surely ‘parlous’ and ‘perilous’ mean slightly different things; the former something like ‘uncertain’ or perhaps ‘imperilled’, the latter more like ‘actively imperilling’. Finances can only be parlous; situations can be parlous or perilous; journeys, say, can only be perilous.

I find it unimaginable that anyone could think 128 players was a remotely sensible idea or that a tie-break system as opposed to a rapid (or preferably unrapid) play-off was anything other than stupid.

It’s not so unimaginable of course that those at the level of Shirov and Kamsky would prefer the champion to muck in with everyone else, but virtually no fans agree.

I don’t see that FIDE has ever improved on its first idea; a candidates tournament like Zurich or whatever Bronstein qualified out of in 1950 (was it Saltsjobaden?). As long as say the top six or so plus a couple from the last cycle are qualified into it, who cares how the rest get in? Who remembers how the likes of Stahlberg got into Zurich, and who cares? It could safely be left to the regions to find a method of nominating a couple of Euros, a token American or so, maybe even an African or Asian, or whatever. Maybe the European Champion, for example.

I’d have thought would give the best chance of sponsorship, an element of democracy, no colour problem, decent chance of an outright winner; in fact pretty much everything. The only argument against this system there ever was (the Fischer Curacao whinging) has gone now the Russians are no longer so dominant.

Candidates matches would be better of course, but if it’s right they are no longer practical, so be it.

Combine the contributions of those who'd support a Kramnik-Anand match with those who'd like to keep Danailov on the sidelines and you could fund a 100-game WCC match.

>Combine the contributions of those who'd support a Kramnik-Anand match...>

A Kramnik-Anand clash would be most interesting.
Nobody would doubt Anand's worth and right to an WCC match if it were to happen.

Maybe Anand can find amongst the nouveaux riches of Bombay a fanatic Hindu nationalist willing to sponsor him. Kramnik would then forget all this empty talk with "FIDE-cycles" and go for it.

I am assuming the size of the field is designed to appease the federations, as it increases the number of nations with at least one participating player. It's like the NCAA Basketball Tournament, which admits 64 teams, even though at least half of them have virtually no chance of winning. (The lowest-seeded winner in that tournament's history was #8 in its bracket of 16.)

I agree they ought to work on a format that allows an equal number of whites and blacks, and that minimizes the use of mathematical tie-breaks. In the second round, with stiffer competition, perhaps the top TWO in each group should advance, instead of just the top ONE. Then, there could be two semi-final matches (4 games) and a final match (8 games).

I also agree that the final match of 4 games seems too short. But part of the problem is that they need a format that doesn't fill up too much of the calendar. The process can't afford to go on for months.

Why not a 14-round Swiss System Interzonal => six candidates?

Then a San Luis-style event (six candidates and two seeded) to determine the candidate?

I have a few things to say about the arguments regarding selecting the best challenger.

First of all, we should all be ashamed of ourselves if we stick to the idiotic idea that ELO tells us who the best player is. A higher ELO means that one player had better results in the past, but it doesn't do much for telling us how he'll perform in any one game against a player who understands chess well enough to capitalize on his mistakes.

Second of all, let's not forget how this mess started in '93. Kasparov broke away from FIDE not with Karpov, who was then clearly the second-strongest player in the world, but with Short, someone who topped 2700 only years later (and was nonetheless incredibly strong, to return to my earlier assertion about ELO ratings). When it comes down to it, as the good folks on ESPN are fond of saying, that's why we play the games. There's nothing unfair about exposing the titans to the chance of losing to less-prestigious players. For all we know, they might only have such high ratings because they (like many on ICC) don't play people who they can lose points to.

Kramnik-Anand would be ok, but it would be nicer if Anand wins something in the next year. Since 2004, has he won anything but the Corus co-win this year? Seems like every time I blink he runs second to somebody.

I think Nigel might plausibly dispute that Karpov was the second-strongest player in the world in 1993 given that he'd just beaten him in a twelve-game match. But anyway I think it's more stuff like Bronstein losing to Cardoso people are concerned about. I agree with you that I don't really understand why, though.

By my count this is the third version of FIDE's "new" WC cycle since the original was announced in April 2005, not counting the snafu with the candidates' matches originally scheduled for this coming April.

So my question is, how is anyone to know that this will be the way it will be in 2007? As difficult as it must be to prepare for something like this, the constantly changing rules must make it that much more difficult.

The new comps rating list with DF-10 tested has appeared. After testing Deep Fritz 10 enters the CSS SMP Ratinglist on rank 2

Fritz has been performing quite well and he managed to maintain this high level of playing in the last 2 matches:

DF 10 - Deep Shredder 10 : 12-8
DF 10 - Rybka 2.2 mp : 9-11

Fritz won all matches with exception of the last one against Rybka.

[Hardware brain: Athlon 64 X2 4200+ (2x2200 Mhz)]

1.Rybka 2.2 --V. Rajlich (CZ)-- 3021
2.Deep Fritz 10 --F. Morsch (NL/D)--2938

3.Deep Shredder --S. Meyer-Kahlen (D)--2895

>o my question is, how is anyone to know that this will be the way it will be in 2007? As difficult as it must be to prepare for something like this, the constantly changing rules must make it that much more difficult.>

here is your answer Stonewaller


Dutch Treat
by Hans Ree

The Gardener and FIDE

If the members of the board of FIDE know their Dutch classics,
they will merrily recite the lines from the poem The Gardener and
Death by the Dutch poet P.N. van Eyck: "I was surprised when in
the morning I saw here quietly at work the man I was to fetch at
dusk in Ispahan.''

These words are spoken by Death. The poor gardener has seen
Death and tries to flee him by running off to the town of Ispahan,
but of course to no avail. He cannot escape his fate and the town to
which he flees is already noted in the Great Book as the town
where Death will find and get him. In the chess world the sad role
of the gardener is played by the organizers who had moved their
tournaments to August so that these would not coincide with the
FIDE World Championship. Dortmund for instance. Every year it's
one of the strongest tournaments on the calendar. Of course it is
impossible to have such a strong tournament in the period where
the best hundred players (except Kasparov) compete for the FIDE
championship. In Dortmund and elsewhere, the organizers fled to
August. They tried to escape Fate but of course FIDE caught them

Was that the action against FIDE which Karpov ultimately won, or at least settled upon a payment by FIDE?

I found that the WC in Kanty Mansysk was a winner. A lot of players in two games ko-system, but only one top seed (Ivanchuk) dropped out early. The other results were very accurate to the ELO-Listing. Imo this system works as an qualifier for a "semi-final" tournament in mexico where the winner faces Kramnik in a best of six match.

to correct myself *best of 12*

> When it comes down to it, as the good folks on ESPN are fond of saying, that's why we play the games.>

Well, if that is such a smart observation then why not having everybody interested playing few games against DF-10 ? Would be like taking an IQ test.
Who gets the best mark ( plays most best, or closer to best, moves) is declared the strongest, the champ.
This would be an objective, absolutely fair assesment and it would eliminate the method of solving the question "who is the best" by requiring the pretenders to play games between themselves.

As has already been pointed out by many people it's absolutely different playing a computer versus playing a GM of the same chess strength, when it comes to psychology, preparation and best strategy. The very fact that opponent feels no pressure and is unable to blunder alone changes the game sufficiently.

Yuriy, you must have forgotten your student years.
There is a lot of psychological pressure when taking an examination.Many students complain that their bad grades are due the high psych. pressure during the exam.

At any rate, my proposal would eliminate the fake ELOs, collusions, and the fake GMs that have debased the title in the last 20 (or so) years. Forget the tournament play, set the "DF-10 exam" as the "gold standard" and cut down the inflation of IMs and GMs in Asia.

Anand won Corus 2006 on tiebreaks.

I have some problems with FIDE's "plan" because I think that something must be done about Mexico City first. As a chess fan and an editor at Encyclopaedia Britannica, I have chosen to only include the sequence of Karpov-Kasparov-Kramnik in our table list of world champions because of the tradition of defeating the current champion in match play. While I applauded the recent unification match between Kramnik and Topalov, I fear that my elation is destined to be rather transitory. For what motivation does Kramnik have to agree to put his title up for grabs in a tournament? It's one thing for FIDE to insist that the champion play a FIDE-chosen contender in a match or forfeit (as Fischer did), and quite another thing to insist that he play on even terms in a tournament for the title.

Of course, FIDE has signed agreements with the Mexico City organizers that their tournament will be a World Championship Tournament, but what good will that do if Kramnik declines to play? I, for one, will view the result as illegitimate and will not add the winner's name to our table of champions.

I don't think that this is the way for FIDE to regain control of the title, and I certainly don't think that the world of chess should return to the pre-FIDE days of champions choosing when and who to defend their title against.

So, assuming that Kramnik refuses to play in Mexico City, what will FIDE do? Ex-communicate him like they tried with Kasparov? Somehow I think that Kramnik will still be able to command money for a match of his choosing and the rift will go on in the chess world.


You bring a good point on reminding people about Mexico City. I think everyone forgets that the world champion in question, in the proposed cycle, is for the NEXT cycle, after the new world champion in Mexico is decided, so it's quite possible that Kramnik will not win the tournament and someone else gets a title.
With respect to the traditional system of champion being defeated by the challenger, I believe that it has been flawed as soon as Kasparov chosed Kramnik (loser of the candidate match against Shirov ) over the winner as his challenger. So any tradition that was preserved before that point became hopelessly lost.
In any case, with regards to the Kramnik's participation in Mexico, it has been already stated by FIDE that both Kramnik and Topalov signed contracts before their reunification match to the effect of the winner participation in Mexico. Trying to "re-negotiate" the contract and not play in the tournament, in this instance, will be an ultimate act of selfishness imho.


The difference is being able to withstand psychological pressure is not a component of one's knowledge. On the other hand being able to play part under pressure is certainly part of one's skill in a competitive sport such as chess.

When did fake ELO, collusion and/or fake GMs make a difference in who held the title over the past 20 years? The DF-matches certainly would set a gold standard...of who would play best against DF-10. Not necessarily the same as how they would fare in face-to-face competition, where, for example, as I was pointing out, if you make certain kind of moves, your opponent may be more likely to blunder, or you may have done better in preparation, etc.


Perhaps he could also win the world championship on coin flips? I did point out he co-won that one , but he should do more if the champ is to handpick him as a challenger.

The classical world champion should have an unsurpassed claim to being the world's best long-match player. Regardless of the Shirov situation, that tradition was safely passed along from Kasparov to Kramnik.

Given FIDE's slipshod practices, the "contract" probably reads, "I, Vlad, promise to play in the Mexican WCC." No financial terms. No dates. No enforceable contract.


> Somehow I think that Kramnik will still be able to command money for a match of his choosing..>

Of course, granted that he chooses a strong, credible challanger.
Kasparov could not find sponsors for the match with Shirov because such match would have been of no interest, no excitement as its result would have been too easily predictable based on their previous encounters.

>.. and the rift will go on in the chess world.>

There is no "rift", the champ has the right to choose whom to play.

It is his title, he has won it. You or FIDE don't have the right to tell him whom to play more than I have it to force you to whom and to what price to sell your car.

Mr. Kamsky,
I see no reason why Kramnik would not participate in Mexico. It is only the 'FIDE title' that he is putting on the line there. Win or lose he would still eb the classical world champion until someone beat him in a proper match.

Koster wrote: "The classical world champion should have an unsurpassed claim to being the world's best long-match player. Regardless of the Shirov situation, that tradition was safely passed along from Kasparov to Kramnik."
Have you heard of the term oxymoron? Look it up in the dictionary, because it aptly describes your statement. How can the claim of being the world's best long match player be made by somebody who failed at the previous hurdle?

I also have a bit of a problem with the statement "The classical world champion should have an unsurpassed claim to being the world's best long-match player."

What about situations where no such claim exists? We were lucky because Fischer, Karpov and Kasparov each blew their opposition out of the water or beat it over and over again. But Botvinnik drew and lost to his opposition, Spassky had an uneven record and Kramnik's record in long-matches (and not even that long) prior to Topalov was +1-1=1. Do you consider those men not to have been true classical world champions?


"But Botvinnik drew and lost to his opposition..."
--When Botvinnik lost a match his opponent established a superior claim to long-match excellence and did become WCC. When Botvinnik drew a match his opponent did not establish a superior claim.

"Spassky had an uneven record..."
--Sure. When he lost, he wasn't WCC.('66, '72) When he won, he was. ('69)

"And Kramnik's record in long matches...prior to Topalov was +1, -1, =1."
--The correct order is -1, +1, =1. Kramnik's loss to Shirov gave him no claim to anything. Kramnik's defeat of the world's best long-match player in 2000 established him as the world's best long-match player. And Kramnik's draw with Leko in 2004 defeated that gentelman's claim to long-match superiority.

To clarify my difference with Mr. Kamsky over the impact of the Shirov debacle on WCC legitimacy:
--Are we arguing (as Shirov does not) that Shirov ever had a better claim than Kasparov to the title of world's best long-match player? If not, then Kramnik's 2000 victory over Kasparov established Kramnik as the world's best long-match player.


quick note before I respond to the substance: +1-1=1 is not any sort of chronogical order but rather the order in which wins, losses and draws are often listed in results tables, including boxing and mixed martial arts. You add commas to my comment for some strange reason...

When somebody loses then wins within a short period of time does it really make a claim that he is the world's best? To me it just shows that the two men are approximately equal and one was lucky the first time, the second the latter. Undoubtedly, one has an unsurpassed claim to being champion if you are the last one to win the match...but that is not the same as claim to being the world's best.

If you consider a single match victory to be sufficient to be established as the world's best long-match player, and a single match loss to no longer being the best, then surely one match draw is sufficient to say the two men are equally good long-match players. I guess Leko is Kramnik's equal.

We should acknowledge that while long-match is a deserving way to determine a champion, when it's close it does not always determine the best player and in fact, it may be impossible to determine who is. Champion does not necessarily refer to decisively best player but to the person who won the last "fair and deserving" competition for the title. A "fair and deserving" system is such that minimizes the possibility of a player on a level below that of the world's best getting that title.

Koster wrote: “Are we arguing (as Shirov does not) that Shirov ever had a better claim than Kasparov to the title of world's best long-match player? If not, then Kramnik's 2000 victory over Kasparov established Kramnik as the world's best long-match player.”

Ridiculous. Firstly whenever you are loath to concede a point, you build up straw men and shoot them down. Nobody ever argued that Shirov was the world's best long-match player. This is an argument you yourself built up so that you could shoot it down, thus avoiding the real point.

Secondly, from your second sentence its obvious what criterion you use for designating somebody the world's best long-match player. It is beating the incumbent “world's best long-match player”. Now nobody contests the point that Kasparov was the world's best long-match player until then, but how did he achieve that status? By a series of matches, not by a single match. And I don’t mean the multiple matches with Karpov, I mean the numerous qualifying matches that he won as a prelude to playing Karpov. This has historically always been the case. Never has ELO or personal recognition sufficed prior to 2000. And note that I don’t mean a lack of matches, I mean losses. i.e. Kramnik lost to Kamsky and to Shirov. How can Kramnik possibly lay claim to being the world's best long-match player or more accurately, how can you make that claim on his behalf when all he did was beat an out of form and ill prepared Kasparov? It’s the series of long matches BEFORE the final match that guard against just this sort of accidental WC. Imagine at this years Football (Soccer) WC, the world clamouring for an England-Italy final even though England went out in the quarters because by some criteria they had drawn up, England was the most likely to triumph against Italy. As I said, ridiculous


We seem to be in substantial agreement. As for the drawn matches, it's traditionally been the challenger's job to surpass the reigning champ. If he fails to do so, the reigning champ's claim to long-match superiority remains "unsurpassed."

That said, when the last three WCC matches have been 2-0, 2-2 and 3-2, draw odds are grossly unfair to the challenger. But all that's been discussed to death in other threads.


We seem to differ on one point and one point only.

The point is you seem to feel that a single victory over the world champion in a match format is sufficient to claim the person is world's best long-match player.

I disagree. I think it's purely sufficient to claim the man is the new world champion. If Sampras beats Agassi 16-14 in the fifth set tiebreak, if Agassi pulls out of the final match with a knee injury, if prior to beating Agassi Sampras lost three finals to three different players, if the two men exchange victories from one final to another, we would of course still give the winner's trophy to Sampras. But we would think twice before declaring him the world's best player.

BTW, for the record, I disagree that the draw odds you mention above are unfair, and certainly not grossly unfair, to the challenger. With a lower number of decisive results, any statistician, not just Jeff Sonas, will tell you the final outcome is less likely to be an indication of true ratio. But that's just life. It's no less unfair than a challenger winning by one game in a longer match, when perhaps the two men are equal in skill. The last two matches, the non-draw percentage was 29 and 45 percent, respectively. Until things get worse, I say chess world has bigger problems.

Twitter Updates

    Follow me on Twitter



    About this Entry

    This page contains a single entry by Mig published on December 8, 2006 2:11 PM.

    Vacation Sunset was the previous entry in this blog.

    Karpov in Brazil is the next entry in this blog.

    Find recent content on the main index or look in the archives to find all content.