The only real question at the start of this year's Biel tournament was by how much Magnus Carlsen would win it. He outrates his nearest competitors, Alekseev and Dominguez, by nearly 70 points. But of course we play the games for a reason and playing as the big favorite has its own pressures. After two rounds in Biel, Carlsen shares the lead with Alekseev on 1.5/2. Mighty Magnus picked up his win in the first round when Pelletier curiously discarded a pawn on move 42 of an opposite-colored bishop endgame. True, it took considerable skill to prosecute the impressive winning endgame zugzwang, but Black's position didn't look dire enough to give up material.
The first game is exactly the kind of position that Magnus seems to win lately though. Drawish looking stuff where he can't see himself loosing, so lets just play on. :) Me likes..
Carlsen just crushed Bacrot. Wasn't Bacrot playing reasonable-looking moves?
Playing in a tournament like this, where he's "supposed" to win, must be a lot of pressure on Carlsen.
Bacrot has a bad habbit of playing big blunders. 23. Nxc5 ? (Nd6 and black is at least equel.) was a mistake and 29. Rhe8 ?? is just loosing on the spot. Bacrot was in some time trouble so that might have something to do with it. This field, Aleksajev apart, is not good enough to give Carlsen a real test. Bilbao will be much more challeging for him. Difficult to compare Carlsen recent achivements with those of Kasparov in his early years. Gazza just blew opponents away right from the oppenings. Carlsen real force is, IMO, in tactical positions in the middlegame and excelent endgame technique. The future is looking bright for Carlsen. Would not be surpised if he is number 1 on the rating list october 1.
Could he beat Kramnik or Anand in a 12 game match. Maybe not just yet. I still feel his oppenings is not as good as Kramnik and Anand.
¿ What's going on with the live list ?, it hasn't been updated in a good while.
Bacrot does not handle time pressure very well. I've watched some of his games live and often the reason for his losses is time pressure (even if both players are under time pressure).
It's like if he was suddenly entering a blitz mode and he stops being careful.
I know it's easy to say from here.
I suppose even Frogbert has to go on holiday sometimes or whatever has happened. Word on the street (the ICC street, that is) is that Carlsen is 2796-something after today's game, and that he'll hit 2800 virtual rating if he wins tomorrow.
It is easy to calculate: Carlsen has virtual rating of 2796.4, and it will be 2795.0 after the next game if he draws (+/-5 for win/loss).
In general he is expected to score 6.4/10, so at the end of the tournament his virtual rating will be 2791.1 +10(score-6.4).
He needs +5 to break 2800 (he will be 2802.1 by above) so 3 more wins are needed (less losses).
Here is what his virtual rating will be at each round if he stays at +2 (towards 6/10). Simply add/subtract 5 for each win/loss...
R4 R5 R6 R7 R8 R9 R10
2795 2794.1 2791.5 2790.6 2789.4 2788.5 2787.1
eg if he gets to 6/8, he has +4=+2+2, so add on 2*5=10 to 2789.4 above to get 2799.4.
Now everyone can calculate for themselves whatever happens...
It doesn't recognise more than one space in a row here, so I rewrote the table...
RND4 ROUND5 ROUND6 ROUND7 ROUND8 ROUND9 ROND10
2795 2794.1 2791.5 2790.6 2789.4 2788.5 2787.1
I think Carlsen is the best chess player in the world right now. He has beaten too many players from drawish endings for this to be simply opponent error. The kid does not seem to need any opening advantage just a reasonably complicated middle game or an even ending.
Thanks REMIT, good job!
@remit, thanks. All this is setting up an amazing Bilbao Grand Slam. Of course Anand (is Kramnik playing?) won't be able to use his white opening prep, but still it will be a terrific event.
I am always wary of superlatives. I doubt Carlsen himself would say he is the best player in the world today. But Top 5 - of course!
And he will be World Champion sooner or later, yes.
If Carlsen finishes this tournament rated #1 in the world, the question will soon arise: Under what circumstances will FIDE allow its #1 rated player a chance at a match for the world champion title? It seems more interesting for the non-chess world press to have #1 and #2 going at it for the title, instead of the current situation where #6 (Kramnik) challenges #2 (potentially Anand) for the title, and next year either #5 (Topalov) or #18 (Kamsky) get their shot at it. We have a new title match system, and already a child genius is showing us that it is hopelessly outdated!
I think the fact that the WCC match in November would be between #6 vs #2 is irrelevant, because part of the reason Kramnik and Anand ratings are not that high is because of "not showing too much" in the tournaments they have participated lately and I think is safe to say that Anand would lose points in Bilbao for the same reason.
I think the winner of that match will certainly recover his place in the ratings next year once the matches are finished and the player can play without reservations.
On the Topalov vs Kamsky match, it is clear that looks meaningless right now (and honestly, it always looked like a burocreatic way to allow Topalov a second shot); I would prefer this match won't take place and start the cycle with a new World Champion (or the same one) in 2008 and the possibility to open the cycle starting the next year; clearly, allowing someone like Carlsen to get a shot for the WCC only until 2011 when he could be a top 2 player in the world right now is a nonsense.
Congratulations to Carlsen, BTW. He might not be my favorite player, but certainly he is showing his prowess not only with the 2600's but also the 2700's and the 2750's. I remember Kasparov mentioning on his tournament in Niksic about "squeezing" out of the position all that was possible and looking for winning chances in drawn positions; he mentioned how proud he felt of scoring 3/3 after three of these squeezes against Petrosian, Seirawan and Larsen. Carlsen is showing this Kasparovian attitude at a younger age, with a lesser dependence on opening preparation ... only a decline in motivation (too much pressure for a young boy) can stop him from becoming a WC in the future.
Live rating:
The FIDE does not use virtual rating of 2791 to calcuate rating for next list. FIDE uses the published rating of 2775 for next list. Carlsen is now 2775, the rest of the field has average of 2669. So Carlsen's expected average win for each game is roughly 0.65(or 0.644 to be more accurate). For GM, the K=10. That translates to the following for Carlsen in this tournament:
Win=+3.5, Draw=-1.5, Loss=-6.5
So for +2: Carlsen rating=2791+2x3.5-8x1.5=2786
for +3: Carlsen rating=2791+3x3.5-7x1.5=2791(even)
for +5: Carlsen rating=2791+5x3.5-5x1.5=2801
Hope this helps.
Damian, I don't think the WC system is 'hopelessly outdated'. Keep in mind that Magnus won his first tournament only one year ago, it was last year in Biel and he shared first with Onischuk. His rise has been faster than even the biggest Magnus fans expected.
on second thought, if they want give Magnus a shot at the WC until 2011 then it might as well be hopelessly outdated at that time. lol
Carlsen had his chance in the KO, he was knocked out , so it doesen't make sense giving him a shot just because of rating.
If Carlsen beats the snot out of Dominguez (25th in the world), Alekseev (26th), Bacrot (33rd), Onischuk (50th), and Pelletier (not in the top 100), he may take over the #1 ranking. But how do victories over those folks demonstrate an ability to stay with a Kramnik or an Anand in a long match?
Elo's rating system is only a very rough guide to quality, and at the top level it's doubtful whether a spread of even several dozen rating points is of any use in predicting the ability of A to defeat B.
The obsession with rating points is completely absurd.
Yes, the obsession is absurd; too much emphasis is placed on who is the champ, opening theory and rating points-its like a glossy magazine really. Elo is pretty accurate though in most cases. For example no-one under 2700 has a real chance of being champ; anand and kramnik are usually at or around the top of the rating list, etc.
"at the top level it's doubtful whether a spread of even several dozen rating points is of any use in predicting the ability of A to defeat B."
Plenty of truth in that...
greg and chesshire, keep in mind that most of Magnus points were gained against top guys in Corus, Linares, Baku and Foros. Biel is more like an after dinner snack.
"But how do victories over those folks demonstrate an ability to stay with a Kramnik or an Anand in a long match?"
Its difficult to predict the future, I know of nobody who has achieved this feat yet. However past results show that "staying with" Kramnik at any rate has been more the norm than the exception in matches, since Kamsky, Gelfand and Shirov managed to beat him, while Leko and Topalov drew, set against victories versus Kasparov and Yudasin. I think its reasonable to assume that Magnus will stay within touching distance in a match. Of course not having the ability of predicting the future myself, I would not want to predict the result of such a match.
as far as i remember kramik beat topal in a match
Hmm, Bacrot really didn't play very well there, did he?
Reminds me of Kramnik's splendid quote before the Topa match 'He has only encountered the sort of resistance one gets from, let's say, 2700 players....'.
Did Kramnik and Topalov play another match before the 2006 one? I don't recall that. Kramnik won the one in 2006.
OK the thing is:
The classical match was officially a draw. Kramnik won on rapid tie breaks.
Kramnik forfeited a game, so Topalov won one game less over the board.
Those are the facts, everything else is opinion.
To rephrase, Kramnik won the classical and rapid parts of the match, and Topalov won the non-chess part.
Anyway d_tal is trolling again, I see. "Staying with" Kramnik is the norm is easy because Kramnik lost some matches early in his career when he was the age pretty much no other player would be good enough to even make it to the stages Kramnik was losing matches at. Yeah, that makes a great deal of sense. One might as well say it was a rule rather than exception to "stay with" Kasparov in a match. Afterall, Karpov beat him once, drew another match, and stayed with him in 3 others. And Kramnik beat him (and drew him), and so did Deep Blue.
Calrsen has already been dealt match losses by Aronian (twice) Bareev and Kamsky. Granted, the very fact he even played these players in advanced stages of important events itself shows Calrsen's amazing talent. But that won't stop trolls like d_tal from bringing it up 10 years from now - even if Carlsen proceeds to become a 4 time classical/match world champ - as if these early losses will someone diminish Carlsen's future accomplishments. Trolls will be trolls.
greg koster: "Elo's rating system is only a very rough guide to quality, and at the top level it's doubtful whether a spread of even several dozen rating points is of any use in predicting the ability of A to defeat B. "
and neither does the rating system claim that it's able to predict precisely the outcome of a match between two specific players a and b based on their individual rating numbers. it indeed says _nothing at all_ about "quality" - what does "quality" mean, btw? (point is: quality is vague as a term, must be defined to make sense to discuss, and is almost impossible to measure across different playing styles and opponents)
what the rating system does quite well however, is to say the following about a and b:
how a did against a's opponents, expressed in one single number.
how b did against b's opponents, expressed in one single number.
the bigger overlap there is between a's and b's opponents, the more the ratings say about the relative strength relationship between a and b. with a "big enough" overlap, it makes good sense to assume that the player with the highest rating _in general_ performs better than the other player, against the kind of players both mostly play. in one sense, it then also makes sense that _in general_, against varied opposition, the player with the higher rating is the "strongest" player.
a match, however, is won by the player who loses the fewest games of the two. ranking players according to match results is obviously a very different measure. even among the top 20 players, the overwhelming majority of games rated (and played) are tournament games, not match games. hence, the quantitative evidence that might be used for match result rankings is so little - or alternatively so old - that any such ranking can't be made in a meaningful way based on statistics, that is, a quantitative measure.
the idea that ratings (mostly based on tournament games) should be very useful for deciding who will win a match (or even a minimatch over 2 games) between two players, is so strange to me that i even start doubting that people who think it necessary to argue against such a view actually understand what ratings are and when they are useful... ;o)
[my personal view is that as long as the rating difference is within roughly 40-60 rating points, then style, previous history between the players (match/game records) and current form are the most important factors if one wants to make an educated guess about the outcome of a match. in a single game, the white advantage obviously is most important, within the indicated rating difference.]
so, returning to the main reason for talk about "too much emphasis on ratings" - the live rating of magnus carlsen, currently 2796,4 - which gives an unofficial 2nd place in the world rankings, 1,6 points behind anand. now, does this sound very wrong, if we would like to measure the players' abilities to score well against the kind of players they typically play? hard to say, obviously, but instead of any long argument this or that way, i'll just list carlsen's results, starting with the world chess cup 2007. please list all players with a more impressive list lately - i'd be surprised to see carlsen outside top 3 in such a list...
wcc 2007: 3-4 of 128
corus a 2008: 1-2 of 14 (cat 20)
linares 2008: 2 of 8 (cat 21)
baku gp: 1-3 of 14 (cat 19)
aerosvit 2008: 1 of 12 (cat 19)
biel 2008: currently 1st (cat 18)
my view is that there are much "bigger anomalies" in the current top 10 in the live list than finding carlsen in 2nd place. :o)
For what it's worth, Elo himself said that his system did not pretend to be accurate to single digits.
Carlsen, Morozevich, Mamedyarov, and to a lesser extent Ivanchuk are probably highly overrated since they mostly generated their 27xx against much weaker players (e.g. - Bacrot) who in turn whipped up on 2600 GM's to get their own rating points.
The ELO system underpredicts results between serious rating differences; hence the ability of Morozevich to crack the top 3 on occasion despite never actually being able to gain a plus score against the top twenty.
frogbert--
Nicely done.
ditto to koster on frogbert.
Just read the brilliantly entertaining history of chess called "The Immortal Game" by David Shenk.
Recalling a quotation by one of Kasparov's vanquished super-GM opponents: "When you play Kasparov, the pieces seem to move differently."
I wonder if Bacrot was feeling this yesterday against Carlsen. The kid seems to have a knack for inviting his opponents to blunder (think Topalov's stunning help-mate). There seems to be a deep psychology underlying his approach; his wins seem effortless, his opponents don't seem to be able to put up their usual resistance.
Perhaps its time for Spassky's mind-control consultants to seek re-employment? ;)
"Carlsen, Morozevich, Mamedyarov, and to a lesser extent Ivanchuk are probably highly overrated since they mostly generated their 27xx against much weaker players (e.g. - Bacrot) who in turn whipped up on 2600 GM's to get their own rating points."
Excuse me? So players like Invachuk, who play everybody, are more overrated than Leko and Kramnik who usually never play people less than 2700+? Keeping a 2700+ is much easier when you don't risk it in opens, where you actually have to demonstrate that you are a 2700 with remarkable results. Please.
"Carlsen, Morozevich, Mamedyarov, and to a lesser extent Ivanchuk are probably highly overrated since they mostly generated their 27xx against much weaker players (e.g. - Bacrot) who in turn whipped up on 2600 GM's to get their own rating points."
That's a hilarious and ridicilous statement. Magnus has played so far this year each and every one of the top players in Corus, Linares, AeroSvit. He has beaten Kramnik, Morozevich, Ivanchuk, Aronian, Shirov and in 57 rated games so far this year with an average opposition of more than 2700. He has lost only 6 of those games and is currently unbeaten in the last 31 games.
yes, Tarjei
Carlsen is an amazing chess talent.
Yet, we must acknowledge that Anand, Ivanchuk, Morozevich, Topalov, and Kramnik are all great chess talents too!
Carlsen lost to Kamsky, so lost out on a match with Topalov.
He is young and will get his chances.
Match play is very difficult against the top six, but even 19th is hard as Carlsen found out!
Carlsen has played 57 rated games against an average opposition of 2722 so far this year. With the exception of Morozwitch he has, for instance, played rated games against all the 15 highest rated players in the world.
"With the exception of Morozewitch he has, for instance, played rated games against all the 15 highest rated players in the world."
Just out of curiosity, does anyone have the stats on how well he did against those 14 players?
I know, I could find out myself, but I'm way too lazy :D
Some classical games results this year:
Topalov: 2,5/3
Ivanchuk: 3/4
Kramnik: 1/1
Anand: 0,5/3
Shirov: 2,5/3
Aronian: 2/3
Mamedyarov: 1/2
Radjabov: 2/4
Karjakin 1/2
Kamsky 0,5/1
Overrated?
I simply HAVE to say this twice a year or so:
ELO was the Electric Light Orchestra.
Professor Elo (if you want to write his name as Élö, full marks to you) invented a rating system that bears his name (minus the diacritics, which would present a mental hazard to most people).
Whoever sees no difference is beyond redemption.
Sorry Knallo, but you got it wrong, ELO *is* the Electric Light Orchestra. They just toured my town a few weeks ago.
Here are his results in rated games against the top 15 in 2008:
Rank Name Games Results Score
1 Anand 3 0-1-2 0.5/3
3 Kramnik 1 1-0-0 1/1
4 Ivanchuk 4 1-2-1 2/4
5 Topalov 3 1-1-1 1.5/3
7 Radjabov 4 1-2-1 2/4
8 Mamedyarov 2 1-0-1 1/2
9 Shirov 3 2-1-0 2.5/3
10 Leko 3 2-1-0 2.5/3
11 Svidler 2 0-2-0 1/2
12 Aronian 3 1-2-0 2/3
13 Adams 2 1-1-0 1.5/2
14 Grischuk 1 0-1-0 0.5/1
15 Karjakin 2 0-2-0 1/2
33 11-16-6 19/33
Out of 57 rated games 33 was against the top 15, or roughly 58%. With the sole exception of Anand Carlsen has scored 50% or better against all these players so far in 2008. Score average, Anand included, is 58%.
Ok, that last post came out almost illegible. Guess we’ll just have to call it a reading exercise. :)
Icewalker:
"
4 Ivanchuk 4 1-2-1 2/4
5 Topalov 3 1-1-1 1.5/3
10 Leko 3 2-1-0 2.5/3
"
--
hm? something went wrong there i think.
ivanchuk, 4 games:
ivanchuk-carlsen 1/2 corus a
carlsen-ivanchuk 1/2 linares
ivanchuk-carlsen 0-1 linares
carlsen-ivanchuk 1-0 aerosvit
ivanchuk 2-2-0 3.0/4 or +2 in 4 games
topalov, 3 games:
topalov-carlsen 1/2 corus a
topalov-carlsen 0-1 linares
carlsen-topalov 1-0 linares
topalov 2-1-0 2,5/3 or +2 in 3 games
leko, 3 games:
leko-carlsen 1-0 corus a
carlsen-leko 1/2 linares
leko-carlsen 1-0 linares
leko 0-1-2 0,5/3 or -2 in 3 games
the end result is the same, though, +5 in his games against the other top 15 players - his two -2 records against anand and leko being the ugly points...
Yes, have to admit I sometimes type faster than I think.
i think it's relevant to point out the following as well, namely the distribution of white and black games in these 33 games against the 13 other top 15 players he's faced this year:
13 white games
C - Adams 1-0 (Baku)
C - Anand 0-1 (Corus)
C - Anand 0-1 (Linares)
C - Aronian 1-0 (Corus)
C - Aronian 1/2 (Linares)
C - Ivanchuk 1/2 (Linares)
C - Ivanchuk 1-0 (Aerosvit)
C - Leko 1/2 (Linares)
C - Radjabov 1/2 (Corus)
C - Radjabov 1/2 (Linares)
C - Shirov 1-0 (Linares)
C - Shirov 1-0 (Aerosvit)
C - Topalov 1-0 (Linares)
20 black games
Adams - C 1/2 (Corus)
Anand - C 1/2 (Linares)
Aronian - C 1/2 (Linares)
Grischuk - C 1/2 (Baku)
Ivanchuk - C 1/2 (Corus)
Ivanchuk - C 0-1 (Linares)
Karjakin - C 1/2 (Baku)
Karjakin - C 1/2 (Aerosvit)
Kramnik - C 0-1 (Corus)
Leko - C 1-0 (Corus)
Leko - C 1-0 (Linares)
Mamedyarov - C 0-1 (Corus)
Mamedyarov - C 1-0 (Baku)
Radjabov - C 1-0 (Linares)
Radjabov - C 0-1 (Baku)
Shirov - C 1/2 (Linares)
Svidler - C 1/2 (Baku)
Svidler - C 1/2 (Aerosvit)
Topalov - C 1/2 (Corus)
Topalov - C 0-1 (Linares)
--
score as white: +6 -2 =5 (only lost to anand)
score as black: +5 -4 =11 (2 losses to leko, 1 to mamedyarov and 1 to radjabov)
players with 50% against carlsen, possibly achieved because they played the white pieces in every game:
grischuk (1 game)
karjakin (2 games)
mamedyarov (2 games)
svidler (2 games)
--
colour balance in more detail...
carlsen has had "positive colour balance" against these 3 players:
anand: 2 whites 1 black (score: -2)
aronian: 2 whites 1 black (score: +1)
shirov: 2 whites 1 black (score: +2)
+1 in 6 white and 3 black games
carlsen has had neutral colour balance against these 3:
adams: 1 white 1 black (score: +1)
ivanchuk: 2 whites 2 blacks (score: +2)
radjabov: 2 whites 2 blacks (score: =)
+3 in 5 white and 5 black games
carlsen has had negative colour balance against these 7:
grischuk: 0 whites 1 black (score: =)
karjakin: 0 whites 2 blacks (score: =)
kramnik: 0 whites 1 black (score: +1)
leko: 1 white 2 blacks (score: -2)
mamedyarov: 0 whites 2 blacks (score: =)
svidler: 0 whites 2 blacks (score: =)
topalov: 1 white 2 blacks (score: +2)
+1 in 2 white and 12 black games!
another extreme example is that carlsen is +2 in 3 black games against kramnik and topalov. :o)
i guess it can be concluded that carlsen's rating is due to him "feeding" mostly on weaker players, under favorable conditions. or maybe not. :o)
Russianboor: "Anyway d_tal is trolling again, I see. "Staying with" Kramnik is the norm is easy because Kramnik lost some matches early in his career when he was the age pretty much no other player would be good enough to even make it to the stages Kramnik was losing matches at. Yeah, that makes a great deal of sense."
You are such a troll, I must have a masochistic streak in me to continue this. Its not about sense, its about facts. Deal with with it.
"One might as well say it was a rule rather than exception to "stay with" Kasparov in a match. Afterall, Karpov beat him once, drew another match, and stayed with him in 3 others. And Kramnik beat him (and drew him), and so did Deep Blue."
How is this relevant? MY comment was in response to a comment above it. Read it.
I have to ask, are you for real? Interminable blah with no rhyme or reason, an inability to distunguish fact from fiction, just line after line of balderdash.
"OK the thing is:
The classical match was officially a draw. Kramnik won on rapid tie breaks.
Kramnik forfeited a game, so Topalov won one game less over the board.
Those are the facts, everything else is opinion."
Exactly, but be careful, you may be smothered by an RB essay.
Carlsen-Dominguez seemed like a rocking game, then it devolved into a quick draw by repetition.
Was that simply theory?
"OK the thing is:
The classical match was officially a draw. Kramnik won on rapid tie breaks.
Kramnik forfeited a game, so Topalov won one game less over the board.
Those are the facts, everything else is opinion."
The fact is that the match that ended in a draw exists only in imagination of pro-Topy anti-Kramnik band. The Championship match was won by Kramnik. Period.
Now if you want to get to the bottom of HOW it was won, the breakdown is also pretty simple: +1 for Kramnik in 11 classical games played, +1 for Topy in the toilet game and +1 for Kramnik in rapid tie-breaks.
It is grossly incorrect to say that Kramnik won the match against Topalov 3-2 in classical chess. The problem is that performance game 6 and on in the match is a function of the match score. For example, it is likely that both Kramnik and Topalov would have played games 11 and 12 differently if the match score was not even. So it is not correct to remove the forfeited game after the match was over. This is not to say, however, that Kramnik was not a deserved unified world champion. Kramnik won it in tie-breaks, consistent with the rules in place and was the unified champ.
frogbert, as usual you have refuted gmc's (and other "likeminded" posters') absurd statements. However, I am wondering if it is worth the effort. "Carlsen has gained his rating points by beating weak players". LOL. We must realise that gmc is simply joking and not take his rubbish statements seriously.
Yasser wrote:
"It is grossly incorrect to say that Kramnik won the match against Topalov 3-2 in classical chess."
If you noticed, all I'm saying is that Kramnik won the match. Which he did. Plain and simple. That +1 -1 story was just a little reminder to these guys who still can't face 2 year old reality, so they invent some "match in classical chess" between Topa and Kramnik that ended in a draw. Such match never existed.
Btw, yessir's argument works also against these guys since all the games would have been played differently if there weren't the provision for the rapid tiebreaks. So you either accept the result of the match as a whole, or you go into the dirty details and they are even less flattering to Topalov then the official result.
THe regular games ended in a tie, THAT IS WHY THE TIEBREAKS U MORONS
Carlsen is now only 2 points away from tying for the no. 1 position after beating Pelletier today !