The MTel Masters turned out to be closer than we thought after Ivanchuk rocketed out of the box with 5/5 in the first half. Topalov made his inevitable run and came within a half point before even his formidable endurance faded in a spectacularly exciting 9th round loss to Bu Xiangzhi. Even after that Topalov had a chance to win the event headed into Sunday's final round. He would have had the better "most wins" first tiebreak over Ivanchuk had Team Bulgaria gone 2-0 in the 10th round. That would have required Topalov beating Radjabov and Cheparinov taking out Ivanchuk. Instead they got 0.5-1.5 when Radjabov's known (but apparently not to him or Topalov according to their post-game comments) queen sacrifice in the Sveshnikov held Topalov to a draw and Cheparinov continued his (self) destructive play and lost rather lamely to Ivanchuk.
That was Ivanchuk's first win of the second half and it moved him up to a remarkable +6 8/10 score. Even before the game was over -- actually by the time he was at 4/5 -- chatter was already rampant about the chance that Ivanchuk's performance would be one for the ages. More on that in a bit. For the first time, Topalov had to settle for a position other than first at this hometown tournament he helped create. His incredible fighting spirit provided entertaining games and a narrative for the second half after Ivanchuk's streak dominated the first. Topalov's 6.5 was enough to win for him in 2005 and 2006. This time he finished 1.5 points back with the same score.
Radjabov's inspired defense against Topalov earned him a +1 finish, a scenario that looked very unlikely after his turgid start. He'll even gain a few rating points. Radjabov's two wins, against Bu and Aronian, highlighted the unbalanced nature of the crosstable. Bu Xiangzhi and Aronian battled to the bottom, both finishing with -4. Cheparinov wasn't much better; in Bu he just managed to find someone playing even worse. His four losses to Ivanchuk and Topalov were as poor as anything Aronian and Bu managed to put together, if not worse. He was demolished by Topalov in the 8th round. Cheparinov's other three losses were all of a piece, over-optimistic attacking play he couldn't back up at all. Sunday against Ivanchuk was another example of this sac, sac, sac, resign formula.
There were exciting games in just about every round, though the quality was occasionally dubious thanks to the poor form of the aforementioned trio and the odd 40/90" first time control. Still, watching it day to day on Chess.FM it gave an exciting impression thanks to the energy of Ivanchuk and Topalov. We spoke on the air with the ICC's Macauley Peterson and Ivanchuk after the tournament. Chucky was as airily affected as ever, responding mildly positively to my question whether or not this success equaled his 1991 Linares win back in his early glory days. I honestly never imagined Ivanchuk would win another supertournament with guys like Topalov and Aronian in the field, but I'm happy to be proven wrong. Especially since we're the same age. Ivanchuk now joins Aronian and Anand as qualifiers for the Grand Slam final in Bilbao in the fall.
The discussion about whether or not Ivanchuk's 2977 TPR deserves a spot on the all-time great tourney performance list is largely an academic one. That argument is about ratings and inflation, not Ivanchuk's performance. Compare this win to that Linares 91 triumph, one of his three Linares wins. There he scored 9/13, an undefeated +6. His tournament performance rating was 2828 for clear first. He finished a half-point ahead of Kasparov and defeated him in their game. Also in the field: Beliavsky, Jussupow, Speelman, Salov, Timman, Karpov, Ljubojevic, Anand, Gurevich, Gelfand, Ehlvest, and Kamsky. That was a fair snapshot of the top 20 at the time. (Only Bareev, Short, Andersson, and Yudasin could have strengthened the field noticeably.) There were three players in the world rated 2700+ and Ivanchuk, not quite 22 at the time, wasn't one of them. (Kasparov 2800, Karpov 2725, Gelfand 2700.) Only 34 players were 2600+.
Now, I won't even get into whether or not Bu Xiangzhi 2008 is nearly as strong as Anatoly Karpov 1991 as non-adjusted for inflation ratings would indicate. Again, his current 2708 rating would make Bu #3 in the world in 1991, just 17 years ago. While chessplayers are getting stronger over the decades, such facile comparisons produce little more than frustration. (Although in this case I would ask who would favor Bu over Karpov in a set match now with a few months to prepare.) We know inflation exists. It's more relevant to compare the relative strengths of the fields for the period in question instead of the numbers. Having 4/10 games against players outside of the top 20 takes this MTel event out of contention for historical placement, though obviously it was a great showing from Ivanchuk. Topalov, Radjabov, and Aronian are all fellow top-tenners, even if Aronian played what he must hope will be remembered as the worst tournament of his life.
Nick de Firmian and I talked a lot with the Chess.FM audience about the various greatest ever tournament performances. Morozevich appears often on the lists thanks to his mind-boggling scores in relatively minor events. 8.5/9 in Kishinev 98 (category 12) stands out, and he put up similar scores in Russian Club Cup events around the same time and Pamplona. As TWIC's Mark Crowther said to me as we discussed Ivanchuk's win, despite the relatively weak opposition Moro faced, few elite players, if any, are capable of such overwhelming streaks. So it's hard to draw lines. We also have amazing open performances such as Sofi Polgar's famous Sacking of Rome in 1989, 8.5/9 and the highest open tournament TPR ever at the time. The teenage Morozevich scored 9.5/10 at his Lloyds Bank Open coming out party in 1994. And how about team events? Kramnik scored 8.5/9 in his first big international showing at the 1992 Manila Olympiad, a 2959 TPR.
Once you go beyond the numbers you can dig even deeper and get into the games, which is how the top players usually judge these things themselves. This is why you hear GM's talking about someone being lucky; they can tell the difference between great form and helpful opponents. They talk about great games, not numbers. Karpov's 1994 Linares, for example, is justly called one of the greatest tournament results of all time. 11/13 against the world's best, 2.5 points (!) ahead of Kasparov and Shirov. It's still hard to believe. But what to make of Bareev hanging a rook and mate in one in an even endgame? Of Ivanchuk absurdly dropping a pawn to a beginner's tactic and resigning? Beliavsky (who finished with 2/13) blundering repeatedly in the early moves of a Catalan and losing in 20 moves in the final round? To be fair, Karpov had a spot of poor luck himself, missing a strong winning attempt against Kasparov. Probably only the very pretty tactical demolition of Topalov (and maybe the smooth grind of Kramnik) would make a Karpov game collection. Anyway, this is probably the wrong example since I work with Kasparov and am always slagged whenever I mention anything critical of his top rivals, but it's only to prove the point about looking beyond the numbers, and that tournament has been analyzed to death. I don't doubt similar could be done to some of Kasparov's great performances, which is why we have the phrase "winner's luck" in the first place.
So what's next?
A couple of rapid events begin next week. The Carlsen-Leko match [Link fixed.] begins on the 28th. Leon starts on the 29th with Anand, Ivanchuk, Shirov, and Vallejo. On May 30, the Pivdenny Bank Cup stars Gelfand, Ponomariov, and Karpov. And Dirt favorite Mikhail Golubev, of course! The next classical biggie is the Aerosvit in Crimea on June 7. Carlsen, Svidler, Ivanchuk, Shirov, and Karjakin are the top seeds. Plus Onischuk, who will surely outperform his bottom seeding. Then Dortmund on June 28, where we'll finally see Kramnik in action again. Speaking of Onischuk, he's in the hunt for the US Championship, currently in progress in Tulsa. Yuri Shulman leads with 5/6, three rounds to go.
Looking forward to the Leko-Carlsen match.
Carlsen has beaten Kramnik in a classical game (and had him beaten in a blindfold rapid game too) and defeated Anand in a rapid game. If Carlsen learns how to beat Leko he could become invincible.
Somewhat tangential, but "winner's luck" made me think of Dennis Monokroussos' formulation, which I found very clarifying as well as practical.
I'm speaking now of non-professional players, but the thought probably is at least somewhat relevant to the elite events of this thread, too.
Anyone who loses a game is likely to say or feel, "He (my opponent) was lucky I missed the killer shot 25..Bxx" or whatever. Sometimes if the game is decided by a truly shameful blunder (like Shabalov's Round One collapse in the current U.S. Championship), even the winner will admit to getting lucky.
Dennis' contribution is that the most productive (in terms of self-improvement) and logical way to view luck in chess is this: It's reasonable to attribute your GOOD results to good luck. It's NEVER reasonable to attribute your BAD results to bad luck.
This is more than simple modesty or good manners/good sportsmanship. There is logic behind it.
When you win because an opponent clearly put up weak resistance, fate handed you an apple. In other words, the root cause that made your victory possible (and perhaps easy) was out of your control.
On the other hand, when you yourself lose due to a blunder, or generally playing below your strength, the reason for your loss WAS under your control. So it's illogical as well as ill-mannered to blame your loss on luck.
It is undoubtedly the case that there are more players with the hightest FIDE ratings than there were some time ago, but is this because they are genuinely better - having had time to learn from their predecessors and stand on their shoulders?
Or are the arbitrary decisions in the ELO rating system biased towards creating inflation rather than deflation?
g,
the elo rating system is calibrated against the average club player.
The 'rating inflation' is an indication that the
difference in strength between the average club player and the top players in increasing.
Regarding Bilbao, Magnus will also get an invitation to it, considering he actually shared 1st place there and the organizers never had any official tie break rules.
Congratulations to Ivanchuk, it's nice to see him on top of his game again. A less plastic member of the Top 5 would be nice (I'm a huge Anand fan, but his Slav games sometimes remind me of latter-day Karpov's Queen's Indians in their snooze inducement).
Bilbao should be interesting, since the winners of the major tournaments are all fighting players (except Aronian when he's bored or Anand when he's 2 points ahead of the field).
I've dreaded this day... During the MTel, I grew increasingly more attached to the video reports and the chiruppy Robert Fontaine informing me who had "won todae' with ze white pisses". I just can't get that incredible accent out of my head, it was just too damn charming.
What on earth am I supposed to do now that's it all over? The reruns will only hold me for so long!
I also think the ELO inflation argument is flawed. It's based on the argument that great players from today have better ratings than great players from a decade ago (and so on..).
Would it not be more sensible to believe that the advances in chess (e.g. by the computer)
simply has increased the skill level in such a way that this fact is reflected?
If somebody can travel back in time and freeze Karpov in 1991, i would bet some money that Bu would defeat him today - yes, even after defrosting, however without Karpov being allowed to 'catch up' in theory.
From that point of view, Ivanchuks performance *is* stellar, but it's nothing more than a statistical fluke. If he gets 800/1000 we can talk again ...
Just FYI: The Karpov Linares 94, with any caveats you care to worry about, was a TPR of 2985. To the best of my knowledge, this still has not been beat in any completed international tournament (streaks within a tournament have beat it; Fischers 6-0 whitewashes have undefined TPR, as conventionally computed; but for completed tournaments, it remains the record).
It's a pretty ridiculous tournament when Topalov gets to play 20% of his games against his own second and wins them both.
You have to wonder if Danailov sits up nights trying to think of new ways to make Bulgarian chess look bad.
@ wingit
Suppose you'll just have to rent "Borat". Know what you mean though, I STILL miss Daljit Dhaliwal on BBC News. Good Luck.
Actually, I don't think that the 2-0 internal Bulgarian mini-match score is a problem. Obviously, they probably have a tacit agreement not to play anything topical or theoretically significant, so a premium is placed on over the board middlegame play, an area where Topalov is one of the strongest players in the world (after, perhaps, Aronian and Anand - or after this tournament, perhaps just Anand).
Cheparinov, like other young players who worked as seconds for great players, has benefited greatly from the theoretical knowledge working on Team Topalov has granted him, and is a genuine talent in his own right. But if we assume that his 269x rating is indicative, at least in part, of incredible preparation for a sub - 2700 player, we must assume that he is a relatively weaker middlegame player than most 269x's. (The games from Mtel seem to support this).
Hence, a 2-0 whitewash by the stronger OTB thinker is far from an unlikely result, especially under the "Sofia rule".
Anyone have thoughts on the following list?
"Middlegame" Players: Anand, Topalov, Carlsen, Morozevich, Aronian.
"Endgame/Quiet players": Leko, Kramnik, Svidler, Mamedyarov.
"Attacking players": Radjabov, Carlsen,Cheparinov
"Opening prep giants": Anand, Topalov
"Middlegame" Players: Anand, Topalov, Carlsen, Morozevich, Aronian.
"Endgame/Quiet players": Leko, Kramnik.
"Attacking players": Radjabov, Carlsen, Morozevich "Opening prep giants": Anand, Topalov, Kramnik
"Special weirdo-style group:" Morozevich, Morozevich, Morozevich,..., Morozevich, Mamedyarov, Nisipeanu
You left Shirov and JuditPolgar out of the ''Attacking Players' list, Adams and Svidler out of the 'Middlegame Players' list and Karjakin and Ivanchuk out of the ''Opening Prep Giants''.
Morozevich had higher TPRs than Karpov's Linares 94 in at least one of the events mentioned above. Again, I obviously don't consider them superior performances as such, but the TPR was higher. At the Chebanenco Memorial in Romania Moro scored 8.5/9 for a 3027 TPR. But only two players were over 2600 (though Moro himself wasn't at the time), Sakaev and Bologan. There were three players under 2500 in the field. Not exactly Linares, which is why the numbers are a bit silly to follow slavishly.
We'd probably see more such things but 1) top 10 players rarely play in such weak events; 2) few players at any level play so hard to win with both colors or 3) to win every game regardless of the tournament score. At the Chebanenco, Moro made his only draw in the second round (with white against the weakest player, of course) and won a 50-mover with black in the final round. Weaker competition or no, that takes some serious fire in the belly along with the talent.
One of my favorite trivia themes is how many players have their highest-ever personal TPR without winning the event. It gives credence to how strong competition drives excellence. Adams and Shirov both scored 8/11 with ~2860 TPRs at Sarajevo 2000. I believe that's the best lifetime TPR for both. But they finished a half point behind Kasparov (2901 TPR). And Anand's tremendous +6 undefeated score at Wijk aan Zee 1999 was a 2839 TPR, also second to Kasparov. I think that was his best until the 2007 WCh in Mexico, which I think beat it by a few points, fittingly. Obviously these guys don't consider these non-winning events their best.
On the last comments above, it was interesting to hear Spassky on Chess.FM say he didn't consider himself a "universal player." That's always been the cliche about him and a largely misunderstood one. He could play any type of middlegame incredibly well, but as he put it, he wasn't nearly as good in the opening and endgame phases.
"Somewhat tangential, but "winner's luck" made me think of Dennis Monokroussos' formulation, which I found very clarifying as well as practical.
I'm speaking now of non-professional players, but the thought probably is at least somewhat relevant to the elite events of this thread, too.
Anyone who loses a game is likely to say or feel, "He (my opponent) was lucky I missed the killer shot 25..Bxx" or whatever. Sometimes if the game is decided by a truly shameful blunder (like Shabalov's Round One collapse in the current U.S. Championship), even the winner will admit to getting lucky.
Dennis' contribution is that the most productive (in terms of self-improvement) and logical way to view luck in chess is this: It's reasonable to attribute your GOOD results to good luck. It's NEVER reasonable to attribute your BAD results to bad luck.
This is more than simple modesty or good manners/good sportsmanship. There is logic behind it.
When you win because an opponent clearly put up weak resistance, fate handed you an apple. In other words, the root cause that made your victory possible (and perhaps easy) was out of your control.
On the other hand, when you yourself lose due to a blunder, or generally playing below your strength, the reason for your loss WAS under your control. So it's illogical as well as ill-mannered to blame your loss on luck."
-Posted by: Jon Jacobs at May 19, 2008 09:52
That does sound logical, and it is surely a better mindset to be somewhat overly self-critical rather than making excuses. But let's just pretend for a second that playing strength is a random variable that can be modelled like in the Elo system, i.e., as a symmetrical distribution (whether Gaussian or logistic) meaning you're just as likely to play above par as you are below. Same with your opponents. In that case aren't you equally justified in attributing your bad performances to chance as you would be attributing your good performances to the same? In other words, in a given tournament, it is just as likely that you ran into opponents who played above par against you (by an average of X points, say) as the converse.
Logically you're right. But as I recall, Dennis was (like myself above) talking about when people use the concept of "luck" to explain subpar play (either their own or their opponent's blunders) - rather than acknowledging an opponent's superior play and labeling it bad luck for themselves, which seems to be a less common behavior.
I don't recall what Dennis' purpose was in the post I cited - it might have been just for abstract clarity (he is pursuing a doctorate in philosophy and has blogged about issues beyond chess). Or it might have been a meditation about the mindset most conducive to working to improve your chess. If the latter, then the way one viewed an opponent's above-par play would probably have different implications than the way one viewed one's own (or an opponent's) sub-par play ... although I lack the philosophical depth to figure out the difference!
re rating inflation: I don't see why it's not possible that a) players today ARE better than those in 197X, and b) ratings are also inflated.
re TPRs: I am really glad Mig is discussing the context of the events, as TPRs can only tell you as much as its number, not the types of games played, the level of domination, etc. This is the only place where I have seen a well-rounded, critical discussion of TPRs. At least of the places I look and admittedly, I am quite lazy...
Looking at the mtel video reports I get the impression that Ivanchuck is a real nerd. More than everyone else.
I'm sure Ivanchuk would speak very highly of you, john.
Cynical Gripe etc. TPRs based on symmetrical distributions are flawed, of course. Mig and others intuitively recognize that a TPR in an event with many 2600 level players means less than a TPR in an event with all 2700 level players. Ivanchuk's Linares 91 performance was even better than this year's great M-tel performance, because the TPR compared him directly with more elite competition. Sofia Polgar's "Sack of Rome" was a great result, but only showed that her performance was far better than her opposition. The numerical extrapolation isn't valid far beyond the oppositions rating, because the statistics isn't adequate for that task (and neither is the phenomenology of a single event). Similarly, Ivanchuk's 3500+ TPR for the first halve of M-tel is only a proof that the TPR doesn't mean as much as some might think.
FrankM's remarks help explain why FIDE doesn't employ the standard TPR methodology (in which 8.5/9 against average 2400 opposition works out to a TPR in the neighborhood of 2760) to assign new ratings to previously unrated players.
Instead, if you make a plus score in your first FIDE-ratable event, they pretty much assign you the average rating of opponents as your initial rating - no matter how lopsidedly you beat up on them. All you get on top of the average is a measly 12.5 additional rating points for each +1. So in the hypothetical example I just mentioned, the initial rating would come out to just 2500, rather than 2760.
If someone scores N-0 (like Fischer) in a tournament, you can get a best 'floor figure' for their TPR by assuming they drew against the highest-rated of the 'N' players.
g
Ivanchuk is awesome!
damn carlsen is playin alot this year.
What was TPR of Tal in Candidates 1959? Tal himself probably had several higher TPRs, but that has to be one of the most meaningful.
Would people appreciate an effort to answer these kind of comparison/inflation? questions objectively? Subject to some of the same caveats as the Guid-Bratko study, but with over 100x the computational effort expended per move, and a better statistical model---? It's the needed "next step" of my anti-cheating work anyway...
Yes Ken: I think it would be a great idea! :-)
I think the better statistical model will be more significant than the 100x computation. The latter is always an option, given today's grid infrastructure.
Guy
Here is a correct link to the Leko-Carlsen match : http://www.lekocarlsen.hu/eng/
On the subject of great TPRs Gabriel Sargissian had a TPR of 3021 at the 2007 Ruy Lopez Festival, an 8 player Category 15 event.
TWIC #217 gives Morozevich's TPR at the Chebanenco Memorial as 2979. Where did you get get your 3027 TPR, Mig?
I had the good fortune to be there in 1994 when Morozevich won the Lloyds Bank Masters - it was my first international tournament. That victory can definitely be attributed to superb play by Morozevich, rather than poor play by his opponents.
I visited this page first time to get info on people search and found it Very Good Job of acknowledgment and a marvelous source of info......... Thanks Admin! http://www.reverse-phone-look-up.net