Adding to the English wisdom on this topic is Luke McShane in his Sunday Express column. Included in his fine notes to Pashikian-Bacrot, Aeroflot 99 is this remark.
Sometimes when you're tempted to agree a draw, it's worth considering this question - would you, in your opponent's shoes, earn a comfortable draw against Kasparov?
If Kasparov could create problems for you, it's likely that you could create problems for your opponent. If that's the case, it's worth carrying on looking for chances. But staying circumspect is essential, as you also don't want to get overambitious and lose.
I always recall Nigel Short's saying something along the lines of "whenever your opponent offers a draw, stop for a moment and figure out why he thinks he's worse." I do like McShane's formulation of trying to not only put yourself in your opponent's shoes, but to imagine a sort of worst-case scenario. You could equally say "would this be an easy draw against a computer?" If the answer is no, lay on, MacDuffShane. (And, of course, ban the draw offer!)
Such role playing, with Kasparov or any other name that might inspire you, is a good motivational trick if you profit from such things. During a recent ICC Chess.FM gig we did together, Peter Svidler, self-deprecating as always, said how "even I would play on here" to indicate how even a relatively peaceable player like him thought there were sufficient winning chances to go on. The point is that he's aware of that "get it over with" tendency in himself, which means he can fight it. He added that while he was always aware that some players would surely play on in some positions where he took draws (or resigned), he just felt too miserable playing some of them out and avoiding that sort of feeling was worth more to his morale than becoming a die-hard.
On the other hand, El Svid has a famous draw in a forced winning position (Anand, Dos Hermanas 1999. Mate in 21 on the board!) and a well-known resignation in a drawn position (Kramnik, Corus 2004), though that can happen to anyone over a career, I suppose. As Mikhalevski noted about that 1999 game, even the sun has spots!
Chess is a fighting game, both sides being roughly talented, the player with the 'killer instinct' will prevail in the long run. This is not new and is a widespread trait in human areas which involve competition. This is hard to grasp for some peaceful-oriented players, but they can't do much as this behavior is dictated by their personal psychology.In the end it is about chance whether a player happens to be aggressive or defensive, because it depends on the many factors that have build his personality. A Leko happens to play like he does because it makes all the sense to his instinct and logic to play that way, and he probably doesn't, in his very self, conceive another better way to play, though he can sometimes fancy with extravagant play-from his point of view. The aggressive player on the other hand wants with all his will to defeat the other, in some cases not only for the game itself, but for the pleasure of domination of the other person. Sometimes the comments of the players give us a glimpse of the underlaying eroticism they experience when they dominate someone, words like "suffer", "dominate", "punish", "crush", "spank" are common in the chess lingo.
In a sense, it can be conjectured that many professional chess players have a bigger chance of owning human-sized plastic dolls, and in fact to frequently play chess with them.
Finally, to defeat a computer arises a very different feeling if compared with the human opponent. Pleasure can only be possible if the 'machine' is subjected to artificial personification by the human. To the people in the chess software business this represent a good strategy to differentiate their products, they could extend their programs with human personalities.
IMO Svidler is playing a game that is not as civilized as him ,that´s all.
"The aggressive player on the other hand wants with all his will to defeat the other, in some cases not only for the game itself, but for the pleasure of domination of the other person. Sometimes the comments of the players give us a glimpse of the underlaying eroticism they experience when they dominate someone, words like "suffer", "dominate", "punish", "crush", "spank" are common in the chess lingo."
"IMO Svidler is playing a game that is not as civilized as him ,that´s all."
Why else would sport be called civilized warfare?
The instinct to fight is a very deeply ingrained one. If not channelized in harmless pursuits, people seek to satisfy it through real warfare.
¨The instinct to fight is a very deeply ingrained one. If not channelized in harmless pursuits, people seek to satisfy it through real warfare.¨
You mean like in the roman circus?, or the beginning of the theatre?
Maybe you are talking about the ball games in the Maya empire, or is it american football?
It seems that there are one or two tinny exceptions for that rule, don´t you think?
"It seems that there are one or two tinny exceptions for that rule, don´t you think?"
The exceptions are just denial, that's all. The ingredients of human nervous systems are all the same. And the instinct to fight comes as a part of the package deal. You can no more have a part of it, than you can refuse to have apes for your forefathers.
The instinct serves a very useful evolutionary role. It helps in the survival of the individual and of the species. Humans just need to keep this instinct under control as the powers of harm at their disposal are much stronger than those of other animals.
Having the instinct to fight is nothing to feel shy about. In fact, it is well known that warrior tribes have far greater accomplishments in terms of civilization, decency and ethics than wimpy ones. Those tribes which do not have fight in them usually grow very rotten.
Read "no more not have a part of it"...:-)
¨Those tribes which do not have fight in them usually grow very rotten.¨
Unlike Germany in WW2? Or Sparta? Mmm i cant say im persuated.
¨Read "no more not have a part of it"...:-)¨
I will , thx.
"Without war human beings stagnate in comfort and affluence and lose the capacity for great thoughts and feelings, they become cynical and subside into barbarism."
- Fyodor Dostoyevsky
I will again refer to the Svidler interview on Chessbase, because now the right thread has appeared ,:) : Among other things, he said that he doesn't offer a draw that often, but somehow finds it very hard to refuse draw offers. Accordingly, he personally likes the Sofia rule, because it forces him to play on ... (another honest and self-deprecating remark).
On Leko, just four examples from 2008 that he doesn't always go for short draws:
Leko-Caruana (Dresden Olympiad, 1-0 93)
Mamedyarov-Leko (Elista Grand Prix, 1/2 110)
Leko-Adams (Corus, 1/2 98)
Leko-Zhigalko (Dresden Olympiad, 1-0 130)
"He who joyfully marches to music in rank and file has already earned my contempt. He has been given a large brain by mistake, since for him the spinal cord would fully suffice. This disgrace to civilization should be done away with at once".
"Heroism at command, senseless brutality, deplorable love-of-country stance, how violently I hate all this, how despicable and ignoble war is; I would rather be torn to shreds than be a part of so base an action! It is my conviction that killing under the cloak of war is nothing but an act of murder."
Albert Einstein
Finish it ,proloy, or go to war yourself.
Well said, Manu.
"Finish it ,proloy, or go to war yourself."
That's Svidleresque! Though, even Svidler says that he doesn't offer draws, only can't refuse the ones made to him!
Btw, if you didn't notice it, I was advocating civilized war. Or sport. Or fighting chess.
Something just reminded me of an incident at a scouts camp during college days. We were holding a chess tournament. One guy (who happened to be one of the funniest around) just got so disconcerted by all the tension that he just scrambled all the pieces, took the board away, and sat to play himself with another guy. We went for dinner. Having come back, we saw the two of them still playing the same game with all the pieces on the board. When checking, we saw that none of them would capture anything. The moment some piece was threatened they would withdraw it.
His explanation: "We don't want any bloodshed."
In the meanwhile Kramnik does it again in Amber...
He just drew in 15 moves as white. It's an unrated tournament, they give them everything they want there, they have the best playing conditions, most players go on playing where they can as they have nothing to lose, and he chooses to rest early, what a weasel!
Kramnik taking such a quick draw with White is of course extremely rare these days. But if you want to think it is typical, I guess nobody can stop you.
Anand caught him with some prep, probably leftover from the match. As he had seen everything at home and the position was equal it didn't really make any sense for Kramnik to play on.
So if anyone, Anand should be blamed for going for a quick draw in the Nimzo-Indian - rather than trying to win with black in the Meran or anti-Moscow gambit !?
Of course this wouldn't fit with the philosophy of Kramnik-bashing, including (other threads) for things ["king of blindfold chess"] he didn't even say himself ... .
Kramnik risked nothing playing, the game wasn't rated, had he went into problems for playing that position there will be no serious damage to his ELO for Christ sake!
That's the difference between a player like him and players like Carlsen or Aronian, they try to play chess, always looking ways to complicate things, while Kramnik is nothing more than a 15-mover Drawnik.
Kramnik remains the king of blindfold, one loss notwithstanding. His performance over the years leaves no doubt about that.
In the rapids, it was Kramnik's responsibility to go for the win. Because he had to take revenge. Why would Anand refuse a draw with black with no advantage?
"Kramnik is nothing more than a 15-mover Drawnik."
Well that just Bowles me over.
Didn't I hear that name somewhere? Wasn't he world champ? Must have won some games I guess.
Btw, Thomas, it's hardly Black's responsibility to liven up a game at this level, nothing to do with Kramnik or anyone else- a draw with black is usually a good result against a 2700er.
¨... every time I get out, they keep on bringing me back in ...¨
Godfather part II
I think there's more to the Kramnik-Anand rapid game than meets the eye. Maybe Anand was counting on Kramnik playing Nf3 after Black's Nd7. In any case, if he wanted, Anand could have continued with Ng3 instead of retreating his light square bishop to g6. Once he did that, Kramnik's Qd1 essentially secures a draw.
Anand probably worked on Ng3 but didn't want to go into the wild complications in the rapid game. Maybe he's saving it for another day...
"it's hardly Black's responsibility to liven up a game at this level ..."
Generally I agree with you, but "Bowles and colleagues" also criticize Kramnik for playing the Petrov with black - and would have a field day if Kramnik tried the Najdorf and things went wrong ... .
Btw, if this was Anand's preparation for the Bonn match, when would we have seen it at that earlier occasion? If the Meran hadn't been (as) successful? Or was it leftover from an early stage of opening preparation before they found that -Bb7 idea in the other opening?
Kramnik must be the new "king of fold!", not blindfold. With the "fold" offer, half a point gone for Anand in rapids. :(
Should we have a rule like this? Unless the position is a forced draw, if you make a draw offer, you lose your right to make another offer to the same player in the same or another official competetion until your opponent makes you a draw offer?? I know it will be hard to keep track. But I like the young generation like Carlsen, Radjabov, dominguez etc. in not offering draws unnecessarily!
Personally, I've never understood those who criticise Kramnik's style. He plays the lines he's comfortable with. Its up to his opponents to prove his chess wrong. If he plays in too "peaceful" a style and that is "wrong" somehow, he will get what he deserves, in not winning tournaments, matches, invitations etc. If he secures tournament and match victories and also continues to get invitations to plum events with the style he has, surely that is a vindication of his chess? Either way, he will get what he deserves methinks. Just as Moro or Anand or anybody else reap the rewards of their Chess efforts, be it in terms of wins, ELO, invites or whatever. There may be people who dislike watching Kramnik's play and prefer say Moro's style. That is hardly Kramnik's problem.
On the separate issue of early draws, in this particular case, Anand played well with Black in the Blind game and this loss must have been devastating to Kramnik. I can fully understand that he decided to take a quick draw when the position was equal. Imagine how much more devasting a score of 0-2 would have been... However it does seem to shortchange the sponsors who expect a fight every game.
sorry meant to say: "Anand played well with White in the Blind game ..." not Black.
Also "I can fully understand that he decided to take a quick draw when the position was equal in the Rapid game."
It's frankly ludicrous to criticise Kramnik's play in Amber. He's always played entertaining chess and won the blindfold section and the overall tournament enough times to make a mockery of anyone who doubts his fighting spirit. In general, you don't defeat Kasparov over 16 games, or come back in a maximum pressure situation against Leko & Topalov, or even take the Anand match to the penultimate game without a will to fight. As d_tal said, his mature style might not be to everyone's taste, but that's a whole different issue.
Re: the Anand draw. Anand's the last person in the world you want to keep playing rapid chess against if you're down on time and don't particularly like your position. The mini-match means nothing, and Amber's a long tournament (effectively 22 rounds). Kramnik might be a bit rusty after playing no competitive chess for 4 months, but I'm sure he'll still try to push on and win if he can. He's only a point back.
Personally I find a Kramnik squeeze to be the most aesthetically pleasing style in the game. I also love a clash of styles e.g. Leko v Moro. Sacrificial King attacks aren't the only exciting games.
"Sometimes the comments of the players give us a glimpse of the underlaying eroticism they experience when they dominate someone, words like "suffer", "dominate", "punish", "crush", "spank" are common in the chess lingo."
Agree with the overall point, although frankly the only place I've ever heard the word "spank" used about chess is here on the Dirt...and the individuals who use it tend not to be chess professionals (an understatement).
The word, "squeeze", definitely should be added to that lexicon.
Here's how my teammate in the recent US Amateur Team East described one of his wins...achieved from an equal endgame after the rest of the team had gone out to dinner, leaving him to toil away alone at his board. (We three had all won, so his result didn't even matter to the match result.)
His game finally over, our teammate whipped out his cell phone and rang us in our diner to report: "I squeezed him until the pus exploded out of his head."
Although that win didn't matter much to the results, in a later match, that same guy again played on in a dead-equal ending, well after the rest of us had finished and gone out to dine. THIS TIME, the score stood 1 / 3 against us. So his game, which he duly won some hours later, was critical to our drawing the match - and ultimately winning a class prize in the tournament.
(Confession: If it were me, I probably would have taken draws in both games.)
I agree with d_tal. I don't care for Kramnik's style, but he should play the way that is best for him. I seem to recall that Kramnik made an effort to change his approach after getting grief for his "a painter paints" interview some years back; the change resulted in a number of poor tournament results. The guy won the world championship doing it his way; it works for him; so be it.
And don't forget torture!
"Sometimes the comments of the players give us a glimpse of the underlaying eroticism they experience when they dominate someone, words like 'suffer', 'dominate', 'punish', 'crush', 'spank' are common in the chess lingo."
Sometimes the comments of bloggers give us a glimpse of their underlying obsession with kinky sex when they think of words like "suffer", "dominate", "punish", "crush", and "spank" as erotic.
"Sometimes the comments of bloggers give us a glimpse of their underlying obsession with kinky sex when they think of words like "suffer", "dominate", "punish", "crush", and "spank" as erotic."
Freudian hangover...?
I guess he used 'erotic' in the general sense of pleasurable.
acirce was spanked badly by Open Defense in a mail game xD
The point is that it's not working anymore for Kramnik. In the past two years Kramnik has probably the worst record with black in top 20: 13% loses all the rest draws. Overwhelming majority of people in top 20 have a plus score in the same time period. For example Ivanchuk and Wang Yue (of all people) win 30% of their black games, while lose only 13% and 9% respectively. Morozevich wins even more (35%), but also loses a lot more (26%). Kramnik maintained decent results due to the killing performance with white (43% wins 6% losses), which is second only to Topalov's (54% wins, 8% losses).
Anyway, I doubt Kramnik would do any worse if he played any reasonable opening where 3 results are possible. 13% losses and no wins is a disaster for a top player by any measure.
Btw, I'm not a stat maniac or something. Fide site now has color stats for every player - quite interesting. Here's Kramnik's: http://ratings.fide.com/card.phtml?event=4101588
ostbender,
We remarked on these numbers at the FIDE site a couple months ago, and I, too, find the GM-level black/white scoring disparities fascinating. I've been thinking about it some.
On one side, those who know the GMs remark how overwhelmingly they prefer the white pieces, having the initiative, and controlling the choice of play.
In contrast, the stats from Jeff Sonas's work showed, I think, an 8% scoring difference between white and black, 54% to 46%, and that a straight line with this slope (54/48) best described the results of thousands of games over thousands of rating-point differences; the same 8% was the color scoring difference, whether 2000s or 2600s.
If it is only 8%, then why the strong preference for the white pieces? Why the huge scoring disparities at the top level? Color only matters less than 1 point in ten games, about 1 point in 12 games, yet black is dreaded, passivity in the face of white's initiative is depressing, playing defensive chess avoided - black will even sacrifice a whole pawn to wrest away white's initiative.
Is this just a human or psychological factor, preferring the initiative and avoiding those positions described as "not worse"? Maybe computers have much to teach us about playing these equal but without the initiative games. Perhaps there is a whole book waiting to be written, Winning Without the Initiative. Maybe this would be inhuman chess, but our psychology blocks us from seeing it as more equal than it feels. I now tell myself with black, "It's only 8%, so cheer up!"
tja
The link to Sonas's 54%-46% white to black scoring is here:
http://www.chessbase.com/newsdetail.asp?newsid=562
Sonas had another article somewhere, I can't find it just yet, pointing out that about the time Khalifman won the knockout wc (1999), he had a very high % of wins with white. Sonas said that Khalifman might have been the strongest player in the world - with the white pieces. This allowed Khalifman to play the draw-with-black-and-win-with-white strategy to perfection.
Some of Kramnik's match play reputation also comes from this statistical disparity - at only 13% loses with black, and in a 12 game match with only 6 blacks, that may equal no black losses at all. Kramnik counts on the no black losses strategy, even if it means no black wins either.
Mig.
A question. If you do not know it (probably you won't)please ask your followers.
In the qualification in the recent European championship in Budva what happened with BOLOGAN and PAPANIOU from Greece, why they did not play the plays off ?then they said it was 22 spots and the clasified 25 players !!!?
Rafael
Bowles - Aronian is no stranger to very quick draws , he's good for 2 or 3 per tournament.
Rafael, I can answer your question (the same discussion already happened at Chessvibes): Bologan and Papaioannou (is that the correct way to spell his name?) "declined" to participate in the playoffs because they were already qualified for the World Cup from the presvious European championship (Plovdiv 2008).
Similarly, three out of the top 11 players were already qualified from Plovdiv (Grachev, Nyback, don't know the third one). Hence it _looks_ as if there were 25 qualification spots rather than 22 as announced.
"For example Ivanchuk and Wang Yue (of all people) win 30% of their black games, while lose only 13% and 9% respectively. Morozevich wins even more (35%), but also loses a lot more (26%)."
Guys, it's completely useless to consider these colour statistics _without_ taking the opponents and games played into account.
Who, of the current top 20 players, played the most games against sub 2700 (and lots of sub 2600, too, regarding Wang Yue) players in the period from january 2007 and onwards, when fide started to collect these stats?
The answer most likely is: Ivanchuk, Wang Yue and Morozevich (and possibly Jakovenko). At least all of those played notably weaker players than for instance Kramnik.
Go to fischl's site and have a look at average opponent average for
Ivanchuk
Wang Yue
Morozevich
for the year of 2007, for instance. Check the same numbers for Kramnik and Leko, for instance (or Anand, Topalov and Carlsen, for that matter)
Comparing apples and oranges is never a particularly good idea.
Thank you so much Thomas.
Persons like you confirm that "gens una sumus".
What is odd that www.chessvibes.com needs to explain this situation and what about the "oficial" site?
Your friend.
Rafael.
Nice work.
It is not only the opponents that should be taken into acount but a lot other factors like how many games per year are played.
In that context Ivanchuk puts his rating on the line most of the time , playing non stop chess like he do.
I find his case very interesting because he does well against the top ( last Mtel for example) and against weaker opponents too , but he plays ALL of them.
On the other hand a player who plays only 20 games in a year has more specific preparation in his pocket to compensate the higher rating of his opponents.
I dont believe Kramnik or Gelfand (to name a few) could match the pace of Chucky without losing a lot of ELO.
Nobody can match chuky, he is unique.
"I dont believe Kramnik or Gelfand (to name a few) could match the pace of Chucky without losing a lot of ELO."
I don't think that's true. You just have to look at the Olympiads or tournaments like the World Cup and the top players usually don't struggle to perform to their ratings against lower ranked players. Or, as another example, you could take someone like Michael Adams, not renowned for his devastating aggression at the chessboard, who still seems to defend his rating when he plays as easily the top seed in tournaments like the Staunton Memorial or European Cup.
It's just that there's not much point playing lower standard opposition - for lower prize money - unless you really have to.
p.s. not that I don't love Ivanchuk. He probably doesn't even know who he's sitting opposite half the time, never mind worrying about rating loss :)
And the Ivanchuk does it again, he just won his mini match with Kramnik. Poor fellow Kramnik, now he not only plays rubbish drawish games, but also loses pathetically... that's life.
¨It's just that there's not much point playing lower standard opposition - for lower prize money - unless you really have to.¨
I don´t agree at all , that´s why i consider Chucky a real chess hero.
I wish the whole top 50 would play as many games as he does , chess is a language that needs to be spreaded.
Greg, it was not meant as a service to Kramnik or anything, I'm just slightly annoyed when statistics are used "wrongly". I'm not pro-Kramnik, pro-Topalov, or against any of them or Moro or anyone, I just like facts and information to be presented as clearly and correctly as possible. At least in general.
Actually, I didn't check anything before posting previously, I just took things from memory (dangerous!) - but here are some numbers:
2008
Anand: avg 2754
Kramnik: avg 2732
Ivanchuk: avg 2719
Leko: avg 2718
Carlsen: avg 2707
Morozevich: avg 2671
Wang Yue: avg 2645
(other top 15 in the current live ratings with lower avg of opponents:
Jakovenko: avg 2670
Movsesian: avg 2622)
2007
Kramnik: avg 2727
Leko: avg 2725
Anand: avg 2718
Morozevich: avg 2693
Ivanchuk: avg 2681
Carlsen: avg 2678
Wang Yue: avg 2608 (!)
(Jakovenko: avg 2640
Movsesian: avg 2578 !)
---
Finally, average 2007 + 2008 followed by black score percentage (2007 to present):
Anand: avg 2736 - black score: W:23 D:66 L:11 = 56 % (53 games)
Kramnik: avg 2729,5 - black score: W:0 D:87 L:13 = 43,5 % (46 games)
Leko: avg 2721,5 - black score: W:12 D:74: 14 = 49 % (57 games)
Ivanchuk: avg 2700 - black score: W:29 D:58 L:13 = 58 % (138 games)
Carlsen: avg 2692,5 - black score: W:23 D:62 L:15 = 54 % (105 games)
Morozevich: avg 2682 - black score: W:33 D:41 L:26 = 54,5 % (75 games)
Wang Yue: avg 2624,5 - black score: W:29 D:62 L:9 = 60 % (133 games)
(Jakovenko: avg 2655
Movsesian: avg 2600)
Both Wang Yue and Carlsen were clearly improving in this period, while Ivanchuk was being a jo-jo. Morozevich's play is both variable and double-edged, so when we compare his high number of wins with black to for instance Carlsen and Anand, we must also take into account that Moro's loss percentage is as big as Carlsen's + Anand's combined (26 versus 15 and 11). Wang Yue increased his rating in 2007 mostly by winning _many games_ against _lower rated_ players. He has a slightly lower win percentage with black compared to Moro, but his loss rate was only one third of that of Moro - and overall Wang Yue has the highest score percentage (60) with the black pieces.
But does the stats for Wang Yue say anything meaningful about who scores better with black against elite opposition, compared to for instance Leko? Not really, I'd say. Wang Yue played an average ca. 100 points lower than that of Leko, and that's significant.
In conclusion, I do think it's possible to say that Anand, Ivanchuk and Carlsen have the more "normal" score profile of top 15-20 players, also with regards to WDL stats, while Moro's total score also is "normal" in a way, but he played opponents rated 50-ish points lower than Kramnik and Anand on average, and his WDL stats are unusual.
Leko scores around 50%, and despite higher opponent average, it seems that he's slightly weaker with black compared to Ivanchuk, and clearly so, compared with Anand. The Anand-Kramnik stats are comparable in many ways, and there can be no doubt who the better black player has been for the past two years of those two.
Wang Yue has simply played so many games against lower rated players, that his stats aren't useful for comparison with the rest of the field here.
No problem, Rafael - it took me just two or three minutes to write that reply ,:).
BTW, this information is not from the chessvibes.com report, but "bits and pieces" from the discussion thread - maybe including inside information from people present in Budva. At another occasion, "Dimitri" posted on Chessvibes regarding the Radjabov-Smeets incident during Corus ... . And when Peter Doggers replied "I cannot argue, you were closer to the action blocking my view", I realized that this was GM Dimitri Reinderman playing in Corus B.
And it is not unusual that other sources are more useful than the official tournament website - though Budva was an extreme case: They had live games (so far so good), "tourist information" (fine with me, and understandable from the organizer's point of view), but already pairings and results were outsourced to chess-results.com. As another example, the Linares website is also not that useful (at least if one isn't fluent in Spanish) - rather than putting my basic Spanish to the test, I informed myself from other Web sources.
I completely agree that winning percentage doesn't tell the whole story. I quoted it simply because that was the only stat available to me. It's absolutely clear that the quality of opposition matters, so you are really breaking an open lock in your post. Calculating performance ratings with black would do the trick.
Still, from Kramnik's winning, oops, scoring percentage it's absolutely clear that his performance rating with black is going to be quite dismal for a player of his strength.
Ok, so I'm also annoyed when people write novels where a simple number would do. So here are some performance ratings with black (assuming that average opposition ratings given by frogbert above are correct).
Anand: 2779
Kramnik: 2683
Leko: 2714
Ivanchuk: 2757
Carlsen: 2721
Moro: 2715.5
Wang Yue: 2696.5
I'll leave to frogbert the in-depth analysis of this data.
what happened in wijk aan zee between rajdanob and smeets (I am lazy writting names)and I was soooo happy that Navara eliminated Cheparinov in the last round.....I wonder if Chepa is going to shake hands with Navara next time.
Rafael
Ok, so I'm also annoyed with people who don't understand when performance ratings are misleading and when they are useful and informational. The annoyance also includes people who don't understand the limitations of there being only 3 possible outcomes in a chess game.
Since osbender doesn't seem to understand why in THIS case comparing Wang Yue and Kramnik by means of performance ratings is less than brilliant, and because he's allergic to anything but short claims, I'll just arrogantly state that his "simple number" won't do - they're more a proof of simplicity of mind than simplicity of the problem.
PS! The latter isn't mainly due to his "simple numbers" being wrong, but for correctness I'd like to point out that those opponent averages were for _all_ games, including whites.
I'm going to compile a statistical analysis to answer this question:
Which number is greater:
a) the amount of times Larry Evans gratuitously brought up Bobby Fischer in his Chess Life column, years after Fischer had played a tournament game, or
b) the amount of times Mig gratuitously brings up Kasparov on his blog, years after he's played a tournament game?
Also forgot to mention that
- the black score percentages also included lots of games played in 2009
- the averages were unweighted averages of two averages, not averages over all (black) games played in 2007 & 2008
Note that this only explains why those performances given by osbender were wrong, not why it's futile to compare such performances between say Kramnik, Moro and Wang Yue.
"What happened between Radjabov and Smeets?"
To make a long story short, it was a time-trouble incident during Corus round 10 and (relevant to the present thread ,:) ) an unusual way to draw a game.
Lots of information on this here and elsewhere, videos on chessvibes and chess.fm, ... . And Dimitri's contribution was Smeets didn't say anything like "correct the piece" or "I want my bishop back!" but simply an annoyed "Jaaaaa".
And on the present black-and-white discussion (in several aspects), quoting osbender:
"Kramnik maintained decent results due to the killing performance with white (43% wins 6% losses)"
Isn't that also a non-negligible achievement? Your overall performance is the sum of your white and black games. So if Kramnik's black performance was a 'mediocre' 2683 [presumably still much better than Bowles, to name just one of the Kramnik-bashers], his white performance must have been close to 2800 or higher ... .
To give just one other extreme: Radjabov was frequently happy with a quick draw with white (at least at Corus), but - often successfully - playing for a win with black.
Maybe frogbert can give further details - anyway, good to have you back on this blog after a period of inactivity! I knew you weren't dead, in hospital or on an extended holiday away from the Internet because the live rating list was still updated regularly ,:)
@ Thomas, good post but you can't compare a fighter like Radjabov with a mere drawing specialist like Kramnik.
Another point, Kramnik-bashers? you make it appear as if there were a secret group attacking Kramnik as their main agenda pfffff!
The first part of my answer is simple: a 43% winning performance with white (second only to Topalov) is rather inconsistent with the idea of a "mere drawing specialist" (at the very least, the word 'mere' is wrong).
As wartime language has already appeared on this thread, Kramnik's style [on his better days] may be torture, making his opponent suffocate and die a slow and painful death starting from an almost equal ending. Topalov (and others) rather use an axe or a Kalashnikov to finish things more quickly, Anand is somewhere in between.
About Kramnik-bashing: Ask yourself if you would have commented on Ivanchuk's victory, if it hadn't been an occasion to put Kramnik down (and IMO the word "pathetically" doesn't do justice to the game, here Kramnik went wrong in a tricky but initially balanced endgame). And - of course you don't have to comment on everything - but you commented on Kramnik's 15 move draw against Anand, but "ignored" Anand's 17-move draw against Topalov .... .
"Note that this only explains why those performances given by osbender were wrong".
Ohh, nice, so now frogbert critisizes me for using the data to calculate performance supplied by no other than ... frogbert. Good game. Supply the correct data, I'll calculate you correct performance.
"Since osbender doesn't seem to understand why in THIS case comparing Wang Yue and Kramnik by means of performance ratings is less than brilliant, and because he's allergic to anything but short claims, I'll just arrogantly state that his "simple number" won't do - they're more a proof of simplicity of mind than simplicity of the problem."
That's a brilliant debating technique: giving a highly disputable claim, then find a lame reason
not to support it with anything. The most logical reply to that would be calling frogbert an idiot... Too bad that would mean going to his level and getting beat by his experience.
"Isn't that also a non-negligible achievement?"
I think you misunderstand my argument. I'm not trying to argue that Kramnik is a weak player, or something. On the contrary, I'm saying that he underachieved recently because of his ineffective opening strategy with black.
osbender,
you got what you asked for. If you had responded to my first posts in a nearly polite tone, you would've got a nearly polite answer back.
Naming me an idiot (which you just did, for all reasonable interpretations of your post) goes well along with your "debate technique" of 1) making arrogant comments about a post's length [completely irrelevant] and then 2) making use of the data that made up the majority of the lines in said post.
That you aren't able to comprehend and/or read the data I provided (including a clear description of what it was and where I got it), does not surprise me, based on your naive approach to interpreting your performance ratings (wrongly calculated, due to your own obviously wrong assumptions).
Well. Let's say I use one of my black openings, the Nimzowitch variation in the Sicilian, and 1) beat a 1400 player, and then 2) loses to Topalov (ca. 2800), then it would be interesting to ask osbender what the result would've been, if I used the same opening against a 2100-rated FIDE-player. osbender seems to think that he knows the answer to that question, based on the information I just provided.
I guess the reason for that, is that he posts before thinking, and never bothers to do any work himself to test his random assumptions.
@Thomas, I agree with you in the style definition, but that's what I mean when I say you can't compare exciting players like say Aronian, Carlsen, Morozevich, Ivanchuk with a 15-mover like Kramnik.
In fact for real constrictors see Karpov or Fischer, they weren't happy with short "equal" games.
In fact, those two could play also axe-chess. Make Kramnik play double-edged games, a Najdorf maybe? No way, he sticks with his ultra safe Petroff, you know why? Because he knows he would drop some 20 positions in the FIDE/frogbert lists! Naturally that's where he really belongs. Cheers.
Bowles, you talk such nonsense, it's painful to read. I hope you are trolling cos it's hardly possible to believe that you mean what you write.
"osbender,
you got what you asked for. If you had responded to my first posts in a nearly polite tone, you would've got a nearly polite answer back."
In my opinion it's you who's got what you asked for. Let me remind you your first post:
"Ok, so I'm also annoyed with people who don't understand when performance ratings are misleading and when they are useful and informational. The annoyance also includes people who don't understand the limitations of there being only 3 possible outcomes in a chess game."
If you don't understand that it's a direct insult, I'm sorry for you. Anyway, I suggest we take the hostility back and debate the issue instead of throwing mud. Your choice.
More food for Kramnik bashers: another 20-move boring game as black versus Carlsen...
"Let me remind you your first post:"
Let me remind you how counting goes.
1, 2, 3, 4, 5 etc.
This was in one of your replies to my 2 first posts about the subject:
"Ok, so I'm also annoyed when people write novels where a simple number would do."
Then in my 3rd post, I paraphrased your answer to me slightly, by saying:
"Ok, so I'm also annoyed with people who don't understand when performance ratings are misleading and when they are useful and informational."
Let's repeat, I've heard it's good for learning:
1, 2, 3, 4, 5 ...
If you would be less annoyed by the length of posts where I provide YOU data that you subsequently use (or you would keep your "annoyance" within your tiny little head), then the chance for a non-insulting reply from me would've gone up a lot.
Good luck with the counting - and leave more advanced mathematical exercises to someone else until you learn how to count to 3.
"I suggest we take the hostility back and debate the issue instead of throwing mud. Your choice."
Sounds like a good idea. I'm all for it.
Kramnik really firing on all cylinders. I wonder if he can blank Carlsen?
Any idea what exactly happened in the Anand-Morozevich blindfold?
A slight move order error :) 6 out of 6 decisive games in the blindfold...
An incorrect queen sacrifice as it can only happen in blindfold chess ... . BTW, 6 out of 6 decisive games included four wins with black, the other ones were rather well deserved.
frogbert, when I read you 10:17AM comment, I thought Carlsen-Kramnik must have been a draw - I was at work, couldn't check the game and thus didn't recognize the irony in your post.
Morozevich's win was generally deserved as he had a much better position. Kamsky, on the other hand, blundered in a good position (for a trap that could only work in blindfold - Kamsky would never assume Topalov just left the pawn en prise otherwise).
Kramnik's continuing his clever WC plan of heading for the same position with white that he just lost badly with. At least you can't knock him for trying :)
"Kramnik's continuing his clever WC plan ..."
I am not sure if his loss yesterday was a direct result of the opening ... and at this moment he seems to be better against Carlsen (at least according to the live commentary by Larry Christiansen).
So if I understand correctly, he transposed the resignation and the queen sacrifice.
He just meant to capture on d5, first with the knight, then with the queen. Unfortunately he forget the first part of the plan, and took with the queen first.
Good reports here, as ever: http://www.amberchess2009.com/RoundReport10.html
You continue to be deliberately insulting in your post at 10:36AM. Then, replying at 10:54AM to the SAME my post you say that you are all for taking hostility back. What's going on, double personality, Jeckyll and Hyde?
"Well. Let's say I use one of my black openings, the Nimzowitch variation in the Sicilian, and 1) beat a 1400 player, and then 2) loses to Topalov (ca. 2800), then it would be interesting to ask osbender what the result would've been, if I used the same opening against a 2100-rated FIDE-player. osbender seems to think that he knows the answer to that question, based on the information I just provided."
I take this is your argument to support a claim that one should not compare the performance rating of Wang Yue because they played opposition of different strength.
First, this is not directly relevant to the original topic, which was poor Kramnik's performance with black recently. Throw out Wang Yue out of the table I posted and Kramnik's performance looks even poorer.
Second, here we are comparing success of players. You are talking about predicting power of ratings. This topics are related, but different.
Third, one might measure success in different ways: percentage of points scored, performance rating, money won, tournaments won, chicks scored, etc. All of this ways have advantages and disadvantages and which one to choose is certainly up to debate. The advantage of performance rating is precisely that it takes into account the level of opposition. If you can propose a "better" measure, please do so.
Fourth, performance ratings are as relevant as a measure of success as FIDE or live ratings are as a measure of playing strength. Since you argument that Kramnik's opposition was stronger than Wang Yue's is based entirely on ratings, you must accept ratings as a good measure of strength. Therefore, you must also accept performance ratings as a good measure of success. In other words, if you find Kramnik and Wang Yue's performance ratings incomparable, you also must find Kramnik's and any 2600 player's (say Khalifman's) FIDE ratings incomparable.
Fifth, I suggest that you part with the idea that you understand ratings better than me. To properly understand how the ratings work you need at least some mathematical background. For what it's worth I do have a Ph.D in math, and you probably don't (judging by the argument you put forward).
With that out of the way, and having established that your argument is not relevant to original discussion, let's nevertheless examine it and show that it doesn't hold water.
1. Your inclusion of specific opening choice (Nimzo sicilian) in the argument is irrelevant and undermines it. Players certainly adapt their opening choices to the opposition. Performance ratings takes care of that automatically.
2. If you want to predict your score in 1 game with 2100 based on 1 game against 2800 and one game against 1400, the probability that your prediction is spot on is not very high. Ironically, it's roughly the same as probability to predict the outcome of 1 Kramnik-Ivanchuk game based on 1 game Kramnik-Karlsen and one game Kramnik-Morozevich.
3. Ratings have excellent predicting power on AVERAGE. Therefore a reasonable thing to ask would be, for example, the following: If 20 randomly chosen players were to play a match of 100 games against 20 players rated 2100. Can I predict well the outcome of such match knowing the outcomes of similar matches against 1400 and 2800 players?
Unfortunately, nobody knows the answer, because there are not that many games played with such huge rating spreads, and so one can neither prove nor disprove statistically that ELO model works here.
4. However we are debating here Wang Yue and Kramnik and the rating differences are 100 points rather than 500. It has been shown numerous times that ELO ratings have excellent predicting power if the players are within 200 rating points. So if you were to chose 20 random GM's and have them play 100 games against 20 random 2600 players first, then you can predict extremely accurately their results against 2700 players. That has been proven.
5. Notice, although that I talked about matches involving 20 players. If I choose only one player instead of 20, in rare cases an individual style of that player could skew the results. That's why you have to average over players as well to get the most reliable prediction.
Osbender, just a quick question: Do you know "the other name" of Hans Arild Runde (the person behind the live rating list) ? ,:)
Yeah obstender , it is frogbert ..
So next time he insults you , it would be nice to hear you say something like:¨thx master¨ .
;)
frogbert said:¨or you would keep your "annoyance" within your tiny little head¨
See , he sees through the internet , he cuts through bone and tissue and actually sees how your brain is small and pathetic.
He is huge dude , dont ever test his wrath again .
Are you suggesting that frogbert is Hans Runde? That would have been a huge disappointment, even though I understand that good people are not immune to being wrong and overreacting...
Yes, frogbert is Hans Runde .... .
Implications: Of course he can still be wrong - but maybe it would have been better debating style to merely write "I disagree" rather than claiming that you are right, hence he is wrong. As far as 'overreacting' is concerned, both of you had their share ... both of you declared peace before continuing in the same style.
But my quip was mostly in response to your mention of a Ph.D. in mathematics (though this may also have been a response to frogbert suggesting that you first need to learn how to count to 3). I think neither expert knowledge of mathematics nor high over-the-board playing strength are necessary or even helpful to understand the ELO system [and its shortcomings]. I am (also) a scientist producing quantitative data, and using mathematical/statistical methods for interpretation - but one always has to worry if underlying assumptions of those methods are applicable in the given context. I am not an expert, but for example many statistical methods assume a Gaussian distribution of results - often a simplification with respect to reality.
Now back on thread: It is quite common to reach a 2700+ rating by playing mostly somewhat weaker opponents - simply because you need such a rating to get supertournament invitations to have a chance of playing stronger ones.
Next step is to confirm your rating against top 10 (or top 30) opposition. For example Carlsen (in a similar situation just two years ago) did so successfully. Wang Yue didn't quite ... yet - I am referring to Corus and Linares. [The Amber tournament is, IMO, just a nice and interesting "joke", not comparable to other major events in terms of defining the players' actual strength.] Jakovenko is the next one having to confirm his top 10 status - once he gets the required invitations.
So in conclusion, I would say frogbert has a point: a 2700+ performance against players rated 50-100 points lower isn't quite the same as the same performance against world elite players.
"Fourth, performance ratings are as relevant as a measure of success as FIDE or live ratings are as a measure of playing strength."
This statement makes me seriously doubt your claim to a Ph.D in math.
Thomas, the fact that he is Runde doesn't have any bearing on whether he is right or wrong, whether he understands ratings or not (you can calculate ratings without understanding, the software is available) and certainly doesn't give him any more right to insult people. He does something useful though, and that certainly scores high in my book in respect department and raises the benefit of the doubt that I'm willing to give by a lot.
If I knew he was Runde before hand, I probably would have reacted less aggressively. Still, his insult came first, despite him claiming otherwise. Scroll back to the beginning of that discussion to see it. I understand that it's easy to overstep the courtesy line on the internet...
Back to the rating discussion:
" but one always has to worry if underlying assumptions of those methods are applicable in the given context. "
Indeed it is though. It's been verified for chess, and in my lengthy post I went into some details to explain exact context in which it is applicable. See items 3-5.
"So in conclusion, I would say frogbert has a point: a 2700+ performance against players rated 50-100 points lower isn't quite the same as the same performance against world elite players."
Yes, of course, it is not the same. Moreover, winning Linares is not the same as winning Wijk (different set of players, different places, different prizes, etc). But if we follow that route we cannot compare say Kramnik's and Topalov's achievements as well, because nothing is comparable. People, however, like to compare such things. And in the context of comparing success as black, performance rating is clearly the best thing known to me. Perhaps, you can suggest something better?
"What's going on, double personality, Jeckyll and Hyde? "
osbender, again you are unable to see that I was just repeating your pattern.
You continued with the hostilities that you started, in 90% of your post, and then said - "but let's forget about it".
I returned the exact same deal to you, returned your hostile reactions and only afterwards said: "yes, peace would be a good thing". Look up your own post if you've already forgot what you did.
I've been copy-catting your behaviour from the start - and you complain about ME being hostile when I do. Maybe a course in self-insight would be good.
In my book, YOU started with irrelevant, negative comments by mentioning how "annoyed" you were with long posts. [And I simply responded to that comment with my "annoyances".] If you can acknowledge that, I can return to our "topic". I happen to know a few things about chess rating and rating performances - I've also got a couple of new features for my site ready to launch as soon as time permits. The major new feature is something I call "Performance Profiles" - which make visible some information that are VERY relevant for this debate.
"[being who he is] certainly doesn't give him any more right to insult people"
I've never claimed such a thing, and I haven't even once before on mig's blog even referred to what I do elsewhere, even in situations where that would've been tempting and/or useful.
It was Thomas, not I, that mentioned my "real" identity, and also in a clash I had with greg koster, I never mentioned "who I am" as if that would somehow be relevant to our debate.
If you're working as a scientist, I'm sure you don't jump to conclusions all the time in your professional time. Not doing so in your spare time either would actually be a good idea.
Copy-catting other person behaviour is frecuently seen in kindergartens, but the phrase ¨he started it¨ is also very popular among violent individuals of all ages.
I really hope none of your new features is an open forum .
"both of you declared peace before continuing in the same style."
Technically I didn't "continue in the same style" before _after_ you had written your post, Thomas. :o)
And I agree that osbender has been a little bit too eager to announce that he is right and I am wrong, at least if it's based on his understanding of my little example, which was rather a misunderstanding of my little, exaggerated point.
But I will get back to that, if he's able to acknowledge that his initial mentioning of his "annoyance" with people writing "novels" when "simple numbers would do" was the _completely dominating_ reason why I put on an arrogant face. Possibly he wasn't aware of the number of irrelevant ad hominems I experienced (due to _form_ and _length_ of posts) last time I engaged in some debates here on the DD, but as a Ph.D. he's surely able to understand that such comments are fallacious by definition.
The ball isn't in my court anymore.
Manu, you started to hate me because we _disagreed_ about something. I guess everyone here knows that you don't like me by now. Do you really need to continue your stalking behaviour making nothing but personally insulting comments about me?
I'm not interested in talking to you, and I leave you alone. Please be mature enough to do the same - you're not interested in talking with me either.
I dont hate u , dude.
You suggested to someone who disagree with u that he needed to learn to count or go to a course in self-insight (etc), i just find that kind of arrogance a little self-descriptive , that´s all.
But like the radiohead song , i might be wrong.
Frogbert, your first "annoyed" comment is time-stamped March 24, 2009 6:02 PM. My first "annoyed" comment is timestamped March 24, 2009 10:35 PM. Then, the insults flew.
My peace offer is still on the table, out of respect for what you do...
I guess you refer to this:
--I'm just slightly annoyed when statistics are used "wrongly".--
which was present in my answer to koster. I admit that I had forgot that I used the word "annoyed" there - even if I qualified it with "slightly" and put "wrongly" in quotes, both for weakening effect. The reason I probably forgot about it, was that it was _not_ directed at anyone in particular (for instance you), it's a reoccuring thing everywhere that stats are misused. But I understand that you interpreted it as a stab at you, which it wasn't. My apologies.
Your comment about me writing "novels" was I guess rather clearly directed at _me_, and like I explained, due to previous experiences here, I reacted sharper than I would've done with no such prior incidents. Then I guess the misunderstandings are dealt with.
If I don't remember wrongly, you expressed criticism of Kramnik's choice of black openings or more broadly his "strategy when playing black". That was part of the context for my little Nimzo-Sicilian example.
The simple point I had was that we have much proof that for instance Wang Yue's black openings work any better at the elite level than Kramnik's black openings. Without Kramnik changing a thing with how he plays as black, he might have outperformed Wang Yue (both rating performance and number of wins) if he instead of his very high average (only matched by Anand) had played the opponents Wang Yue was demolishing in 2007.
So
1) my "specific black opening" was a placeholder for "an opening repertoir"
2) beating the 1400 was "being successful" at one level, losing against 2800 was "being unsuccessful" at a very different level (exaggerated to make it clear)
3) and the point about the 2100 game was (of course) that we wouldn't have the faintest idea about how that "opening repertoire" would work out there, even if the "rating performance" of previous games was exactly 2100.
The latter point is a simplified but _relevant_ example of the shortcomings of rating performances - as my data (and calculations) clearly shows, different players perform at (very) different levels against different level of opposition. My performance profiles track this for all the players in the world top, over a significant number of games (currently running 18 month periods/windows).
These profiles are visualized in graphs where for instance two players can be compared, or you can watch a player's varying performances against different levels of opposition, compared to his own rating and his own average performance (only the latter is being calculated and reported by Stefan Fischl in his performance stats).
You asked for something better than (basic) rating performances - I give you performance profiles. :o)
"The simple point I had was that we have much proof that for instance Wang Yue's black openings work any better at the elite level than Kramnik's black openings."
Argh... editing and the mistakes it leads to...
I meant to say: "we have NOT got much proof..." of course.
Looking forward to the performance profiles... very interesting stuff. Strange that a game like chess which is so obviously given to deep statistical analysis of performance has settled for so few basic measures of performance, when you compare with a sport like baseball which has been analyzed every which way from here to Sunday. We still need a little Jamesian revolution...
frogbert,
Apologies accepted. I also apologise for overreacting. Indeed I interpreted your "annoyed" comment as a stab at me. It's funny how we treated each other as clueless arrogant trolls, because we've seen too many of them before.
"The simple point I had was that we have much proof that for instance Wang Yue's black openings work any better at the elite level than Kramnik's black openings. Without Kramnik changing a thing with how he plays as black, he might have outperformed Wang Yue (both rating performance and number of wins) if he instead of his very high average (only matched by Anand) had played the opponents Wang Yue was demolishing in 2007."
My point is that it is irrelevant how Kramnik would have fared against Wang Yue's opponents and Wang Yue would've fared against Kramnik's opponents. They play who they play and must chose (adapt) their openings so that it maximizes the result. That's the whole point of opening preparation and opening strategy.
Rating profiles are certainly very interesting and useful as they allow a great deal finer analysis of tendencies than a performance rating. But if my question is "Does my opening repertoir work for me?", you better give me one number to look at, not a profile. Here, performance rating is the best thing I can think of.
Concluding, I'm sure Kramnik himself is not very happy with his black performance lately. I would be very surprised if he doesn't supplement the Petroff, for example, with something more complex in the very near future.
¨It's funny how we treated each other as clueless arrogant trolls, because we've seen too many of them before. ¨
I knew it wasnt your fault , i knew it.
"My point is that it is irrelevant how Kramnik would have fared against Wang Yue's opponents and Wang Yue would've fared against Kramnik's opponents."
I think I'm mostly in agreement with you there, but note that comparing Wang Yue's performance with black against his opponents (the ones he played in 2007 and most of those he played in 2008, except the Grand Prix events) with Kramnik's black performance, is about as irrelevant. At best it shows that Wang's "black approach" was working for him against _his_ opponents at the time.
Earlier in this thread it appeared to me as if for instance Moro's and Wang Yue's high percentage of black wins was used as evidence for there being something wrong with Kramnik's approach to having the black pieces. But possibly we agree then, that those stats aren't very relevant for evaluating _Kramnik's_ game? Despite what the rating performances tell us?
Due to Kramnik and Anand playing mostly the same kind of opponents, I find that comparison much more to the point.
"I knew it wasnt your fault , i knew it."
The last comment from me to you in a while: I don't know about osbender, but if I or someone close to me would consider it a good idea for me to get psychological and behavioural evaluations from someone, I think I would choose to contact a professional and talk things over in a good, friendly environment.
If you have no intention of debating chess issues with us, I consider it more appropriate if you would look after yourself for a while, instead of adding your irrelevant personally oriented "observations" after each and every post of osbender and mine. Your services aren't wanted to put it bluntly.
I thought your last comment was the other one..
The moment you started posting your insults and overreactions in a public thread , the moment you start getting comments that you may not like.
That rule works for you and me and everyone ,get used to it.
Of course im not interested in debating anything with you , just take a look at this thread and count how many agressions and insults you put on it.
Osbender, I go all the way back to your post yesterday 6:33PM:
"It has been shown numerous times that ELO ratings have excellent predicting power if the players are within 200 rating points. ... That has been proven."
Even if this was 'proven' (by whom? based on what kind of evidence?) to be _generally_ true, it may not be true for each individual player. I still maintain that the two ways to obtain a 2700 or 2750 performance rating (and eventually, to get or keep a FIDE rating of that order of magnitude) require different approaches to chess.
Approach 1: scoring well above 50% against weaker opposition
Approach 2: scoring roughly 50% against 2700+ opposition
[once you get that far, there is hardly option 3 - a bit less than 50% against stronger players - only in matches].
Particularly with respect to black performances, I think it is quite difficult to get winning chances against the strongest players ... .
In team competitions, the captain can, at least to some extent, maximize strengths and minimize weaknesses of each player: one is stronger with white, another one with black; one is better against strong opponents, another one more reliably beating weaker ones.
In individual play, each player is on his own ... and I think hardly anyone is equally strong in all aspects of the play.
Back to Kramnik (just one example BTW). He wants to (or can) spend only so many hours per day or week on chess, also not all of this time is devoted to opening preparation. He may well think that this time is more efficiently spent on searching for further subleties in the Catalan(an opening he already knows well and that suits his style) than giving the Najdorf a try (which doesn't suit his style). Nothing at all wrong with that IMO. There are no rules that the Najdorf is obligatory, or the Petroff forbidden - maybe this will change if Bowles succeeds Ilyumzhinov as FIDE president ,:) .
I wrote:
"It has been shown numerous times that ELO ratings have excellent predicting power if the players are within 200 rating points. ... That has been proven."
You wrote:
"Even if this was 'proven' (by whom? based on what kind of evidence?)"
Proved in numerous studies. Based on results of actual games.
You wrote :
"it may not be true for each individual player."
Yes, it is not necessarily true for each individual player. Players have periods of steady improvement and decline, style specifics, easy and difficult opponents, they are influenced by bunch of other factors that ELO model does not account for. If you average, the influence of such factors cancels out and you get the behavior well represented by ELO statistical model.
Kramnik, of course, is free to spend his time in any way he chooses fit, perhaps, not working on any openings at all. But all the evidence points out to the conclusion that his Catalan works great, unlike his black openings. It's a lot easier to improve your below average results with black to slightly above average than to improve you unbelievably good results with white to out_of_this_world level. So if I were Kramnik and decided to spend my time on openings, I would certainly work on the dark side, but what do I know...
frogbert wrote:
"Moro's and Wang Yue's high percentage of black wins....But possibly we agree then, that those stats aren't very relevant for evaluating _Kramnik's_ game? Despite what the rating performances tell us? Due to Kramnik and Anand playing mostly the same kind of opponents, I find that comparison much more to the point."
I agree that Kramnik v Anand is more direct comparison than Kramnik v Wang Yue, therefore more to the point. IMO Kramnik v Wang Yue is relevant, but I agree that degree of relevance is debatable.
It is well known that having white on average is worth about 50 rating points. I wonder if this number is the same across rating brackets. What would that number be if we consider only games between 2700+ players, then only games between 2600-2700 players, then 2700+ v 2600-2700, etc. Do you have such stats available? I think having these stats would help a lot in clarifying to what degree Wang Yue's and Kramnik's stats are comparable.
"Yes, it is not necessarily true for each individual player. ..."
Conclusion: Kramnik is not an average player ,:) ["typical" may be more pertinent in the present context]
"So if I were Kramnik .... I would certainly work on the dark side"
Here you have a point ... and maybe his rapid game earlier today against Leko was an indication that he "listened to" (or rather anticipated) your remark ,:). He played a Sicilian in a must-win (or rather "wanting to win") situation with black - however, I think the Paulsen was a wiser choice than the wild ultra-theoretical and tactical Najdorf.
Before anyone else points it out: Yep, earlier in the Amber tournament had stuck with the solid, conservative and (mostly) drawish Petroff.