Just getting the discussion thread up early this time. Will add the round notes later. Feel free to post results if you're watching live. Leaders face tail-enders today. What are your predictions for the second half? Can Kamsky keep up his amazing level? Will Topalov turn in yet another comeback? Anand, Kamsky, and Topalov all have three whites. Last year the players tired in the second half, perhaps partially due to the no draw offer rules. It's hard to imagine more decisive games than we saw in the first half, although a decline in quality wouldn't surprise. We've seen several tremendous games so far.
Update: Wow. Gata Kamsky turned in perhaps his best game yet, outplaying Ponomariov and then demolishing him when the Ukrainian made an ill-advised attempt to get activity with 35..d4. (I'll toss 29..Rc2 out as a more active defensive try.) Very impressive stuff. If you haven't drunk the Kamsky Kool-Aid yet, it's time to step up to the bar because it's all about Brooklyn, baby! I'm a little surprised Ponomariov didn't play the Sicilian. Kamsky's positional and endgame instincts have always looked fine; it was sharp stuff where he was having trouble during the comeback. On the other hand, tell that to Svidler.
Speaking of, El Svid got just the sort of sharp Grunfeld position he loves against Topalov and made a nice show of it, making Topalov and his novelty look pretty bad. That knocks Topalov, the defending champ in Sofia, out of contention. Bacrot played the Marshall against Anand, or, since White usually avoids it, we can say that Anand allowed the Marshall against Bacrot. A very complicated game along the usual Marshall themes. Black has good attacking pressure, White defends and hopes the initiative will break. It eventually burned out to a tense draw. Very high quality material today.
Predictions:
Topalov rebounds with solid 2.5/3 over the next three tournaments. He should win as he has the easiest schedule down the stretch (the three lowest-ranked opponents, 2 of them with white). Kamsky will continue to play well, but finish short of first place. Ruslan Ponomariov will remind us again that he is no Ruslan Bairachny.
I'll take a stab at predicting these Round 6 games: Kamsky-Ponomariov (with Black to play move 29) - it looks tough for Kamsky to hold at this point but the presence of opposite colored bishops makes me predict DRAW. Anand - Bacrot (with Black to play move 22) - this really seemed to favor Black a few moves ago, but I'd say it only slightly does now, so DRAW. Topalov - Svidler (with White to play move 27) - the Queens just came off, improving Svidler's chances, but now the doubled black pawns also become easier targets, so this looks like a WHITE WIN. Topalov - Svidler is easily the most dynamic game, in fact Black winning is probably more likely at this point than a draw. Of course, what do I know, I'm just a class player.
For what it's worth, here's what GM Shipov & IM Notkin are evaluating in their online commentary on chesspro.ru:
Topalov-Svidler - a clear advantage for black - extra passed pawn, bishop pair, that sort of textbook stuff.
Kamsky-Ponomariov - a slight advantage for white, black a and d pawns are weak.
Anand-Bacrot - too complex to call.
Thanks Alex! Sooooooo, I'm totally wrong on Topalov-Svidler (and I can't even blame it on my having a man crush for Topalov - 'cos I don't). Mostly wrong on Kamsky-Ponomariov. And I'll take a "draw" on my Anand-Bacrot prediction. So in predicto-chess I'm 0 wins, 1 draw, 2 losses - pretty much the same score I get at most one day tournaments. :) Hmmm, better dust off those chess books...
Update:
Kamsky-Ponomariov: black exchanged off one of his weak pawns, but at the heavy cost of allowing white to activate his pieces and take the 7th row. After 36.e6 white's attack seems unstoppable. I predict another street party in Brooklyn tonight.
Anand-Bacrot: Anand spurned an easy draw, which could have been achieved just by letting black sac back the exchange on e4. After 25.e5 and 26.h3 the tactics favored .... well, actually they didn't favor anyone yet, it's still anyone's game.
Topalov-Svidler: according to Shipov, Svidler, around moves 28-30 made one subpar move after another. Still, black's position remains much better. He no longer has the bishop pair, but his rooks are more active, and the c-pawn is not easily dealt with.
Wow, chesspro is an amazing site. Svidler still has the definite advantage in his game. Go, Petr, go!
nice game for Gata. I am becoming a believer.
gata nice play. I am with you all the way. lets win this tournament.
Just wanted to let you know, as a non-american, I'm pleased to see Kamsky performing so well. His games are a pleasure to watch, I hope he can keep up this level of play and I'm sure if he does, he'll be a contender for the WC again.
and Kamski wins again!
With Anand's draw he is the sole leader.
If this offends anyone out there, I don't give a ....! It's nice to see that the USA has the real deal again in chess. Now I remember how to spell it, "hoot"!
I think Kamsky's result has been great. I think that the no-draw format has really helped out, because Kamsky always plays to win, so he's used to it. It would be so great to see him go from this tournament to the Olympiad and perform well in both, though I expect we'll reserve him only for Russians and Ukranians.
Peace...
The Olympiad matches will likely be decided on Bd. 2 and Bd. 3. Kamsky, now that he is showing 2700+ form again, will likely rack up draw offers up top against the stronger teams, and Ibragimov should do well on Bd. 4. Nakamura and Onischuk figure to have the best battles throughout the critical rounds. It is good to know that Kamsky is quite capable of scoring the entire point against the elite players on the top board, so that if needed he can go for it all and people will not feel like we are therefore doomed to failure. Regardless, I expect a good result from the US, and with good fortune and excellent form, perhaps a medal is in the cards for us. This is definitely the strongest US team we have had in quite some time, and also one that has a number of fighters on it, so I look forward to following.
Hotep,
Maliq
Svidler is up a piece, but if Topo can exchange black's e pawn, it's a draw with the wrong colored bishop for a rook pawn.
It looks good for Svidler though.
It's over, Svidler wins!
I guess someone has caught on to Topalov's fraud.
Think about it:
You all live in some country where there is a top grandmaster who has reached 2600 and plateaued there for years. What are the chances of that grandmaster suddenly upping his play -- after ten or so years at the 2600 plateau -- and playing at 2700-strength? That's right: It just does not happen.
And try figuring this out: How come Topalov does not play well at Amber, in which speed and memory come into play? You would think he'd be amazing at Blitz, too -- like Anand. How about that play a couple of months ago? Lose several games in a row then win several in a row. That's balanced!
So maybe the rumors about Argentina were true: Topalov used a computer.
Of course, Mig will probably delete this because challenging the champion is generally frowned upon.
Maliq,
That's a good evaluation, and with Kamsky's current form, my optimism for a medal has risen. I hope his current form acts as a catalyst for the other US players on the team.
Chess Auditor, That is stupid.
Chess Auditor, Why stop with Mig? If you're out to cast yourself as a free-speech martyr, why not go all the way and warn that BUSH is gonna delete your comment? That, surely, would get you even more sympathy from fans of the mindless drivel that you specialize in.
Furthermore, we all know deep down that Kamsky playing at this level is much less surprising than Topalov's emergence as world champion. Kamsky has played at this level before, whereas Topalov was never in the Top-3 before the last two-three years.
I've never been a particular Kamsky fan (I'm also not American!) but this is very impressive. He now has the same number of points as he got at Corus, except he's played less than half as many games. Is this one of the greatest comebacks ever? To return to chess and start beating the world's elite must rival Fischer's achievement in beating a somewhat "past his best" Spassky in 1992.
Flyonthewall: I notice you did not call my mindless drivel a lie.
It couldn't take much effort to walk into a chess tournament "wired." I'm sure there are GMs who do it. But to accuse a specific player based on "rumors" or because, after a ten-year plateau, he got a lot better but played poorly at a rapid tournament is irresponsible.
Mig has a fault or two. But aside from obscenity, character assassination, and mentions of his 1836 chess rating, he shows a remarkably light touch with the censor button.
Chess Auditor,
Not that I agree with what you have expressed, but if Mig were to "delete" it, I would quit blogging here. Oh Oh?!
Anyway, I think his current less than impressive form has more to do with his new found fame. Not being use to that position and having everyone and their grandmother jumping on to your bandwagon so-to-speak, would be taxing at best. Oh please Mr. Topolov can you do this? Oh please Mr Topolov can you do that? I know there was a recent world champion that thrived on that, but I believe that "Toppy" is not of that mold. In fact he strikes me as a person that is contemplative by nature, more water, less fire. Although it doesn't necessarily translate that way OTB.
Chessbase (Mig?) says that "Anand had a very promising position against Bacrot but allowed the Frenchman to escape with a draw." TWIC says, "Anand struggles to a draw against Bacrot."
Any idea which is more accurate?
I still think that the idea of putting Onischuk on board 1 and letting Kamsky and Nakamura try to rack up points on boards 2 and 3 might be the best idea for the U.S. team. I just hope that Kamsky isn't totally drained after MTel and has no energy for the Olympiad. After all, he's not a young kid anymore.
Peace...
I am perplexed, g, as to why you would possibly think that not putting the strongest player in the US on Bd. 1 is sound strategy. Kamsky can beat most Bd. 1 players (he has already beaten Anand twice since his return, and has dispatched of Svidler and Ponomariov quite harshly in this tournament), so you actually stand a chance of picking up that full point up top. Nakamura and Onischuk are dangerous regardless of whom they play, so I don't see the danger of having them sit on the center boards. Ask yourself, if the US needed a draw on Bd. 1, which US player would be most likely to receive the olive branch from a top player right now.
Hotep,
Maliq
RR,
I think both Anand and Bacrot "struggled to a draw". It was one of those games where, no matter how hard both players try, and what pyrotechnics they pull out of their, uhm, nose, the equilibrium is forever maintained.
"You all live in some country where there is a top grandmaster who has reached 2600 and plateaued there for years. What are the chances of that grandmaster suddenly upping his play -- after ten or so years at the 2600 plateau -- and playing at 2700-strength? That's right: It just does not happen."
Actually, this has been known to happen. Bobby Fischer seemed to have plateaued at one point, before making a sudden jump and becoming unbeatable. On a slightly more modest level, Kramnik did a similar thing.
No one has offered a plausible theory how Topalov could have cheated his way through so many winning tournaments over the last couple of years. "It just does not happen."
"What are the chances of that grandmaster suddenly upping his play -- after ten or so years at the 2600 plateau -- and playing at 2700-strength?"
Chess Auditor, he's been above 2700 since 1999 (at least) and in the top 6 in the world since 2001. check your auditor's license first, maybe it has expired.
Ok, so first thing : Kamsky's playing great chess. Probably he's not far from his level from 1995.
About Topalov, and fraud suspicion, I think that any chess player above 2000 can look at Anand-Topalov first game in San-Luis, and see that this game went at least two times from won for Topalov to total draw. A player using a puter wouldn't blunder this way.
Furthermore, not so long ago, Topalov missed, in the 2nd part of San-Luis (in which he was back at his level) an immediate tactical win against Kamsky I think (that Kasparov mentioned almost at once). Tactical lines are not the kind of thing a computer can miss.
Topalov has spent a few years between 2700 and 2750. And he's had a fantastic year. Yeah and? That's the same with almost anybody. You reach a certain level, then you'll wait (and work) a few years, then you reach another level, and so on...
In this tournament, I think that Topalov is already not playing chess any more. He has other goals : his match against kramnik. He shall not show any more novelty, he shall not show anything about what he intends to play in september. A world championship is much more important for him than this tournament. I think he plays M-tel only because he has contractual obligations (and because he's last year winner, because it's in Sofia, and because he has a good contract). But he certainly won't be willing to show his weapons right now. Remember, last year at Dortmund, Topalov had been almost hollow...
Chess auditor, you probably forgot that 35 years ago, a player suddenly went from 5th or 10th place in the world to a stratospheric first and to the world champion title. Hmmm ... so according to you Fischer was using Rybka as well?
Also, for those who were around back then, there was a period somewhere around '95 or '96 when Topalov was winning every tournament in sight. He was virtually unbeatable for a year or two, then seemed to drop back to a lower level for several years.
Those who believe that Topalov used computer assistance can regard his successes as evidence of computer aid and his relatively average results as evidence that he had computer aid for the successes and was playing on his own when he doesn't do as well. It's all a matter of what you believe and how you interpret the events.
Is it true that Brad Pitt will be portraying Gata in the upcoming made for American television movie that has been mentioned on Chessbase?
The no draw rule really helps out Kamsky psychologically. Pono clearly wanted to steer the game to a draw (he was Black, after all...), and the position was simplifying. However, time pressure approaches, and it is easy to make a superficial move. That is when Kamsky strikes. He is a real opportunist, and the other players are giving him the opportunities! It's not like he has blown Anand, Pono, Bacrot away in the openings. He did land a shot against Svidler, but that was not due to his opening preparation; rather it was a blunder on Svidler's part.
The swiftness of his comeback is quite impressive. When he builds up a reservoir of Opening Novelties, he could be challenging back amongst the Top 5, right up there with Topalov, Anand, Aronian, and Svidler.
Let's hope that he can keep it together for the final 4 rounds, and maybe improve on his Plus 3 score. Hopefully, he'll be able to win the mini-match vs. Anand, if he plays a decent game with the White pieces. He's due White vs. Topo, which ought to help.
In retrospect, it was probably a mistake to invite both Pono and Bacrot to the M-Tel event--although they have been the "chum" for the other players. Both of the tailenders may find invitations to Elite tournaments drying up, if they are not careful. Especially Pono--as The Ukraine does not have the capacity to host such Super GM events.
In retrospect, The FIDE KO World Championship where Pono won his title turned out to be more of a disaster for Ponomariov, who won the title, than it was for Ivanchuk, who choked his Golden opportunity to become World Champion right away. Ironically, Pono became a prisoner of his own exalted title, and has never recovered his strength, fighting spirit, or consistancy.
"In retrospect, it was probably a mistake to invite both Pono and Bacrot to the M-Tel event--although they have been the "chum" for the other players."
This comment is laughable. Pono is the #6 player in the world on the FIDE April 2006 rating list. If the #6 player doesn't belong in a 6-man tournament, then who does? Obviously he hasn't played at that level so far, but who could have predicted that?
"In retrospect, The FIDE KO World Championship where Pono won his title turned out to be more of a disaster for Ponomariov, who won the title, than it was for Ivanchuk, who choked his Golden opportunity to become World Champion right away. Ironically, Pono became a prisoner of his own exalted title, and has never recovered his strength, fighting spirit, or consistancy."
This is absurd. On the last FIDE rating list before Pono won the KO WCC, he was rated 2684, 20th in the world. On the first FIDE rating list after beating Ivanchuk, he was rated 2743, and #6 in the world.
Four years later, he is rated 2738, and #6 in the world. This hardly suggests a player who has lost his strength and fighting spirit. Inconsistency is perhaps the one valid charge, as Pono at one point had sunk as low as #20, before rising back up to #6. But since he won the WCC, he has been above 2700 in every quarterly rating list except save one (2695, April 2005). He's having a bad tournament, but Ponomariov belongs.
His one clear career mistake was failing to agree to terms for a match with Kasparov, an error apparently foisted on him by his handlers, and which he should regret for the rest of his life.
Please note that Ponomariov was a relatively last minute substitute for Aronian who declined the invitation to prepare for the Olympiad. So, cut him some slack.
>I am perplexed, g, as to why you would possibly >think that not putting the strongest player in the US on Bd. 1 is sound strategy.
Maliq:
First, I think Kamsky will be tired after this tournament, so he might not be at his best at the Olympiad and might be tempted to make quick draws if on board 1. Second, Onischuk is a solid player who did very well playing first board last time. Let's not also forget Onischuk is the U.S. champion and that deserves some respect also.
If Kamsky were to make it into the top 5 after an 8-9 year hiatis, that would literally be an an amazing accomplishment. I'm stoked about what he is doing. I wish the National Open was this weekend. In my mind, if he had continued to play during those "lost years", I believe that we may be talking about him in the same way that we speak of Anand. Just my opinion.
Re: the Olympiad, quick draws on board one aren't bad when you have a balanced team like the USA does. Your lower boards get the points and your top guy breaks even against the big boys. -1 has been a decent result for an American first board since Fischer left the scene since they haven't had anyone in the elite, although Seirawan was close enough for horseshoes.
Onischuk had a great +2 on board 1 in Calvia. Only loss was to Svidler, wins against Shirov and Carlsen. Draws with Anand, Ivanchuk, Gelfand, Akopian, Radjabov, Kasimdzhanov for a 2700 performance. With him or Kamsky and Nakamura on boards 2 and 3, +2 on board one would put the USA into serious medal contention this year. If your top board is 100+ points higher than your #2 you need him to score big.
As for tailenders, there are always tailenders. Pono did fine here last year and is an established force. Bacrot hasn't proven himself in supertournaments yet and is en route to another negative score. (Although he did well at the World Cup.) But he's hardly cannon fodder. He's only a half-point behind Topalov, the world #1.
Actually, I wrote: It was probably a mistake to invite both Pono and Bacrot to the M-Tel event.
Inviting one without the other would not have been so bad. It is one thing to have a tailender; it is quite another to have two players vying for the bottom of the cellar.
It is hard to argue with rating, but Pono's results suggest that he is somewhat overrated. I doubt that he'll be ranked 6th in the World after this tournament; indeed, I believe that Pono's rise back up to 6th is something of a Fluke.
Pono has clawed his way back to achieve a rating close to his personal best. Yet, ratings inflation has continued unabated and a 2738 rating today is probably inflated by a good 15-20 points.
Granted, Pono has survived Super Gm Elite Round Robins before, and usually finished within a point or two above or below Even. Not overly inspiring.
Pono should indeed get some slack, given that he was a last minute replacement. Even so, he is headed for a poor result, irrespective of the mitigating circumstances.
As for my observation that Pono would have been better off not winning his title match vs. Ivanchuk, I am not convinced by any counterarguments. Even if we accept that Pono is indeed back close to level that he achieved four years ago, as manifest by his 2743 rating, this only indicates 4 years of stagnation. Pono's earning of the KO title led his career being mismanaged "by his handlers". It is likely that those mistakes would not have occurred, if Pono was not Quixotically defending the honor of his Paper WC Title. If he (and his handlers) had been a bit more modest about the worth of his victory in a "Roulette" KO event, then maybe he would have come to terms with Kasparov, and gotten his match, and maybe the biggest payday of his life.
Of course, Kasparov would have almost certainly outclassed him, but he would have been able to cash in his title.
As such, the fact that the Kasparov match fell through was only a loss of money. However, Pono's handler's had concommitant demands for the first 2-3 years after he won the title. Increased appearance fees, special consideration, cherry picking events, etc.
Of course, many organizers demurred: they already had plenty of big ego players, and (for that matter) plenty of World Champion title pretenders. So, the past 4 years is really a lost opportunity: Call it the "Opportunity Cost" of winning the KO WC title!
THIS is what damaged Pono's career. Pono is in a 50 % rut, and I doubt that there will be a 2nd peak that will bring him past Leko, Svidler, and Aronian. Likewise, Adams, Gelfand, Polgar, etc. are at least a match for him. Now Kamsky is back.
If Pono had a tactical style like Ivanchuk or Shirov, that would be one thing. But Pono's style is perceived as being dull, and that means that he's going to have difficulty getting invites to Elite Round Robins.
From the FIDE site, I note that he has gained most of his points from 2nd tier and 3rd tier events. It's not a surprise that he can still do so. However, it won't be long before he is too old for "Young Masters" events.
Ciudad de Pamplona Mag & Cuernavaca Young Masters
Besides, even if he takes the rating points (5 or 10 at a time) for edging to a 1st place, what's the use when he cannot really hope to excel against the sharks?
Doug: sorry, I don't see the logic. You seem to be presuming great things would have happened for Pono had he lost the WCC. That's assuming a lot. For all we know, he simply would have been part of the large scrum of 2nd-tier players that hang around, but never win anything serious.
It's certainly true that Pono failed to make the most of his opportunity. But had he not won the WCC, there simply wouldn't have been any opportunity at all. How does he benefit by that? Meantime, his #6 ranking is real.
It's true there's been rating inflation over time, but that's mainly a consideration across eras. In a given era, everyone is part of the same system. Pono plays a fair share of games against the top GMs, and usually holds his own. Will he ever be top-5? Maybe not; but maybe he simply was never destined to be.
As Mig pointed out, most tournaments have tail-enders. You can't always guess who they'll be. Before this tournament started, Kamsky was the one who looked like he didn't belong, both because of his rating, and because of his very recent humiliation at Corus.
As sportscaster Chris Berman always says, "That's why they play the games."
Unfortunately, Pono is very unstable for years. Nobody ever knows the level of play he will show in his next tournament.
For a decade most of the elite players made a living not finishing first. Only Kasparov, Anand, and Kramnik did that. Between's Topalov's spate of wins in 1995-96 and Kramnik's drop-off, only a handful of events were won by others when two of the KKA were present. Leko got a few, Adams one, not much else.
You have to get to the top ten by winning second-class events because until you're in the elite you don't get invited to the Grand Slam events. (Unless you are Dutch, or Paco Vallejo) More relevantly, you can't just keep inviting the same guys. Pono is having a bad tournament, somebody has to by definition. Bacrot is only on -2 with losses against the leaders, hardly a catastrophe.
Bad results happen. If Bacrot keeps falling he won't get the invites. But inviting so-called outsiders - preferably qualifiers like Dortmund uses - is essential. Sure they won't all be supertournament winners like Bologan and Naiditsch, but you have to keep the fresh blood flowing.
You can all squeal about lack from proof from here until doomsday. As a matter of fact, lack of evidence is what Topalov is all about.
Anyone who has ever studied chess ratings and their calculation knows that for any rating differential between two players there is an almost exact estimate of a ensuing result between the two. Topalov has defied that rule on two occasions: When he performed well above average as a 2700+ player and when he performed well below average for a 2800+ player. Thus, no matter how you look at it, Topalov never plays at his relative strength. I can only explain that by suggesting that some external influence -- possibly a computer -- was at his disposal.
"Anyone who has ever studied chess ratings and their calculation knows that for any rating differential between two players there is an almost exact estimate of a ensuing result between the two."
I don't understand that sentence. Can you expand on it? I'm assuming you've studied ratings enough to know what a single standard deviation for a tournament performance might be, and you can include that in your explanation.
Thanks.
Chess Auditor, for a whole bunch of potential reasons, players may perform above or below their expected rating.
For instance, given Vladimir Kramnik's recent performance, you'd have thought he was ill, since the formula says he should perform much better than he has done lately. Oh wait, I forgot. He HAS been ill.
If players routinely met their expectations, we could cancel all tournaments and a formula would tell us who won. It is because players routinely miss expectations -- in both directions -- that tournaments are exciting.
If Topalov were cheating, surely the one place he'd be able to get away with it is on his home turf, at a tournament he already won last year. Computers don't play the way he does. No computer would have recommended the exchange sac he used to beat Kamsky.
[Correction: "Lack *of* proof." See above.]
To specifically address the point made by some posters, that Topalov was above (not WELL above) 2700 for several years, I suggest that this is probably the biggest clue to Topalov's imposture. Had Topalov progressed steadily throughout the years, his success would have been easier to swallow; however, a sudden, meteoric rise to the top suggests that he gained extreme amounts of new knowledge and understanding of the game in a very short period of time. What does Topalov of 2005 know that Topalov of 1995 does not? Where did he obtain this newfound knowledge? From Cheparinov? From his manager?
[Jonas: There are charts available detailing the exact chance of winning, losing, or drawing a game, based on rating differentials between two players. I am sure that with a little diligence we can find them, though I am too tired to try right now.]
1. "Chess Auditor, for a whole bunch of potential reasons, players may perform above or below their expected rating."
2. If players routinely met their expectations, we could cancel all tournaments and a formula would tell us who won. It is because players routinely miss expectations -- in both directions -- that tournaments are exciting.
Both (1) and (2) are wrong. I do not have the list in front of me, but just off the top of my head, I do not recall a major tournament that did not figure Kasparov-Kramnik-Anand at the top.
"For a decade most of the elite players made a living not finishing first. Only Kasparov, Anand, and Kramnik did that. Between's Topalov's spate of wins in 1995-96 and Kramnik's drop-off, only a handful of events were won by others when two of the KKA were present. Leko got a few, Adams one, not much else." -- Mig Greengard.
See, you simply can not claim that. Strong players -- Bareev, Shirov, Polgar, Gelfand, etc. -- never came close to finishing first in which the top three participated. Of course, you can add Topalov to that list.
A description of the ELO system can be found here:
http://en.wikipedia.org/wiki/Chess_rating
The table I was describing earlier is as follows:
ELO difference Expected score (*)
0 0.50
20 0.53
40 0.58
60 0.62
80 0.66
100 0.69
120 0.73
140 0.76
160 0.79
180 0.82
200 0.84
300 0.93
400 0.97
____________________________________________
(*) http://gobase.org/studying/articles/elo/
Ridiculous. Players move around all the time. Topalov was in the top ten for years and displayed the raw talent long ago. His rise to 2800 was steady over two years and represents around 50 points over 1.5 years. This is not sudden. Topalov went to the top, plateaued and fluctuated in the top ten, then put on another burst to hit #1 over a nearly two year period. (2745 is well above 2700, by the way.) What, he should continue to 3000 or be accused of inconsistency or cheating? He was never the most consistent player. All of this piffle is about one amazing result, San Luis.
Player, even very strong ones, show improved performance. Adams, Morozevich, Ivanchuk, Kasparov, all had jumps of 30-60 points in less than two years long after they first hit the top ten. Leko gained 40 points in one year to reach his career high. Svidler gained nearly 60 in 1.5 years to reach his career peak in 2004. It wasn't enough to reach #1 which is why with Topalov it's a big deal. On the other hand, if Kasparov were still playing Topalov still wouldn't be #1 and we wouldn't be going on about it so much.
Topalov's jump was not much of an anomaly. He just had a higher point from which to take off, making it especially impressive. He (and Anand) are finding out how hard it is to maintain this level.
chess auditor, just let it rest. i've been watching topalov's games and he doesn't play like a computer at all. i know because i use my computer for analysis and know the kinds of positions computer-type moves tend to get into. for example, in the recent game against kamsky, he used an exchange sacrifice that computer programs wouldn't recommend. topalov's style seems very human to me.
anyway, kamsky for president. i gata go now!
COTDT: Fritz 9 running on a 900mhz Pentium III recommended 27. Rxe6 almost immediately as the only move that would allow white to maintain initiative.
Not that I think Topalov cheats... that suggestion is one I think is best left to the profoundly paranoid.
26.e6 is the choice to sacrifice the exchange. If you don't play 27.Rxe6 you've just thrown away a pawn for no reason.
Btw, IPs are logged and identity theft is not a joke. Give it a rest. Thanks.
Yeah but computers don't think in terms of sunk costs or what it should have done earlier. They only look at what they need to do now.
Kamsky has the black pieces I think against Bacrot so Kamsky might lose.
Chess Auditor, you must have forgotten the Dortmund (I think in 2001) in which Anand placed 6th and didn't even score a single win.
Or MTel 2005 in which Kramnik placed dead last with a -2 score. Your accusation of cheating is ridiculous.
Topalov 's rise can be easily explained; firstly he is talentes (as every GM over 2650 is...) secondly he is in a very good form (mainly it is a psychological effect) and thirdly...and more importantly ....his opening repertoire...is in a Kasparov-like level. For example dont you remember how he beat Aronian from the book by playing a fantastic (nevertheless prepared) exchange sacrifice ? ? ?
It is ridiculous and irresponsible to accuse someone when you havent any clues.
First, by my own admission, none of this can be proven, and everything that I am saying is conjecture only. But the charge itself is not "ridiculous": Anyone following baseball can surmise that Barry Bonds has cheated, even though he never tested positive for a steroid. Lance Armstrong's 54km/h time trials on hilly terrains (even though the hour record _on a track_ stands at 56km/h) are also highly suspect, which explains why French fans, well-versed in the Tour, were instantly alarmed. (In the end, they caught him using urine frozen from six years prior.) Second, there is at least one player from San Luis who brought up the same suspicions; so far as I know, most 2700+ players' minds are fairly lucid, and can hardly be considered "ridiculous."
While we are on the subject of computers, of which I know a fair deal, having studied them at the university level, and having taken a certain interest in chess programs, let us consider for a moment the history of the Computer Chess Championships. After Cray Blitz won in 1983 and 1986, Robert Hyatt released its code freely as Crafty in the mid-1990s, with various improvements. He did so knowing his engine would never be Chess Champion again. Many programs, including Fritz, benefited from Crafty code, and have asked permission of its author. (See http://groups.google.com/group/rec.games.chess.computer/browse_frm/thread/cc8520f40dc69198/6a1a0f5e00d2e358?q=crafty+copy+fraud&rnum=2#6a1a0f5e00d2e358 ). To attribute certain parts of a program to its author is progress; thus, if you borrow a piece of code from Hyatt and attribute it to him, just like scholarship, you are quoting a master and are improving on his work. But let us not kid ourselves: There are plenty of commercial reasons to want to not attribute code or ideas to someone. That is what plagiarism and cheating are all about. Similar to the above case, I do not think I need to detail what cheating has to offer; however, I must emphasize that cheating tarnishes and belittles the world championships.
Mig's contention that Topalov's *rating* improvement occurred over 1.5 years is correct but, with respect, his several wins in a row and 3000+ performance at the Fide World Cup event were the first warning signs. It was an event in which he of course lost both rapid tie-breaks, after beating both Kozul and Kharlov 2-0 in earlier rounds, they themselves having incredible performances.
More later....
The Topalov-cheats stuff is just a joke, but I feel it's important to set the record straight about one "stealth" issue raised in this thread, which might slip beneath some people's radar and lead to destructive consequences somewhere down the line.
I'm talking about the statistical inference concept as applied to Elo ratings (by the way it's "Elo" not "ELO" -- for reasons that will be immediately obvious in what follows).
Yes, the rating systems do incorporate certain statistically-derived assumptions about the probabilities of any particular result occurring, whether in a single game or a full tournament, based on the ratings of the respective players, and the rating differences between the players.
However, everyone who follows chess -- all the more so if you have any power, such as if you ever serve as a TD, club official, or federation official -- should take care to NEVER treat those assumptions as gospel. It's just a theoretical construct, not an iron law of physics.
Not only are humans human (i.e. not statistical automatons who always perform according to models), as other comments on this thread have already pointed out.
More important, experience has shown that where various natural phenomena are concerned -- and even more so for social and psychological phenomena -- so-called "outliers" (results that lie at extremes of the theoretical distribution of possible outcomes) usually happen more often than theory predicts. In finance, this realization brought a revolution in recent years, in which makers of option-pricing and other securites valuation models came to incorporate various so-called "fat-tailed" distributions that assume more-frequent occurrence of extreme outcomes, than the old familiar "normal" distribution (Bell Curve) allows for.
Now for the kicker, and my real reason for posting this.
Some time ago I happened upon a chess-related Web page that detailed an incident I found very sad, that should serve as a cautionary warning about both hubris in general, and the tendency to over-generalize from statistical models in particular.
The sad story is about how Dr. Arpad Elo, the statistics professor and long-time USCF Rating Statistician who invented the rating systems that bear his name (and which subsequently spread from the chess world to a number of other sports), penalized numerous deserving U.S. players on their way to achieving their inital FIDE ratings in the 1970s.
Based on two players' seemingly anomalous results in tournaments that Goichberg submitted to FIDE, Dr. Elo (who by then had left his long-time USCF post to join FIDE as its all-powerful rating statistician) managed to convince himself that Goichberg was dishonestly manipulating the results to attain desired FIDE ratings (NOT TRUE!).
The late Dr. Elo therefore decided not to issue a FIDE rating to ANYONE on the basis of ANY tournament Goichberg submitted in the future. Apparently Elo never even told anyone about this, either at FIDE, the USCF, or Goichberg's organization. He simply filed Bill's tournament reports in the wastebasket, and did not rate them, without saying anything to anyone.
It took years to straighten this out. According to the narrative I found on one of Sam Sloan's Web pages, it was eventually repaired only after a series of international meetings bringing together Dr. Elo, his boss, then FIDE President (and ex-World Champion) Max Euwe, then USCF Executive Director Ed Edmondson, and Goichberg.
The needless suffering and aggravation and wasted time that Goichberg himself went through in dealing with this was bad enough. Far worse, over a period of years, at least a couple dozen rising young American stars were denied the FIDE ratings they had earned in "Futurity" and similar events Goichberg had organized in the US, and had properly submitted to FIDE. These players weren't even told why -- they simply waited for FIDE to publish their ratings, and when it didn't happen, they were in the dark.
I see this whole incident as an abject lesson in what happens when even the most statistically knowledgeable person in the world (Dr. Elo after all had personally invented the rating system he was implementing) overestimates the degree of reliability and authority that should be assigned to both his own judgments, and the output of his models.
"[Jonas: There are charts available detailing the exact chance of winning, losing, or drawing a game, based on rating differentials between two players. I am sure that with a little diligence we can find them, though I am too tired to try right now.]"
You missed my point on standard deviations & process variance, or you didn't understand it. With every mean comes a variance, and that's what you're talking about when you bring up performance that's different from expected value...
If you knew so much about Elo, you would have probably noticed that sawtooth performances are very usual and many players lose a good amount of points noly to recover them after a while or the other way round. Without going further, out of my last 20 games, I lost around 30 points in the first 10 to regain them in the last , which in terms of tournament performance might result in around 300 points difference, and suppose it is not a unique case...
Chessplayer, you're wasting your breath. I can already tell you how the ironically named "chess auditor" will manage to dismiss your comment (if he responds at all). He will simply say it's absurd to compare your personal experience with your Elo rating, with what can be expected of top GMs. As with tgg on another thread, actual experience means nothing to such people; they always find some means to explain it away, so they can go on repeating their endlessly self-confident points, without need to support them with real-life evidence.
By the way it should be a simple matter to extract enough actual performance-rating data from the Sonas web site (albeit doing so might require a paid subscription to it), to compare the observed fluctuations of Topalov's performance across several tournaments, with the fluctuations experienced by other top players, eitehr individually or in aggregate. In other words, rather than gauging Topalov's performance variance against some theoretical variance assumption, one could simply gauge it against the comparable statistic among other top players, based on their actual performances for the past few years (or longer).
As your story is without attribution, save for the web site of notoriously paranoid man, Sam Sloan, I will not bother answering it. But supposing you are right about Elo skewing the numbers, what is your point? That the system is wrong because Elo acted wrongly? Know this: If Pascal comes up with a theorem and then shoots his mother, the theorem still holds. That is because the theorem existed in the world order before Pascal came into existence and does not depend upon Pascal being alive or dead. By transitivity, Elo's system works regardless of how Elo himself might have misused it. Thus, since we must cut the weed at its root, any argument detailing Elo's actions themselves bear no relevance to this discussion. That and the fact that you are just an unthinking insect to me, Flyonthewall.
Unlike many here, I try to avoid labelling anyone a "troll." But Chess Auditor's last comment proves that's what he is (though I suppose that was evident even before that).
To "Chess Auditor" (again I point out the irony in his choice of that handle), everyone is "just an unthinking insect." That is why everyone reading or posting here, has already in one way or another, correctly dismissed him in print as unthinking.
The reason his last post in particular unveils his troll nature, is that it would have been obvious to anyone capable of thought, that my comment he claimed to be responding to, was neither aimed at him (nor sought a response from him), nor was meant as a criticism of the validity of the Elo rating system.
My comment was, quite simply, just what I labeled it in its initial paragraph(s): a cautionary tale about the sad consequences that can arise from man's tendency toward hubris, as well as from the temptation to place too much faith in abstractions and models as predictors of real-world behavior. People who work with statistics are already aware of the problem and try to correct for it; I just wanted to call its attention to readers here who might not be as aware.
That's well known that cash makes people disembarrass. But how to act when someone has no cash? The one way is to receive the loan or secured loan.