The top 100 January 2006 FIDE rating list is out. I'm sure Mark will have his usual summary top list up soon enough. [ChessBase has some stuff up now, though without the handy gainers and losers info Mark usually has.] As expected, Topalov completes an amazing 2005 and comes in as the top active player at 2801, becoming only the third ever to crack 2800. These marks are fun but irrelevant. 2700 used to mean world championship class; now that would cover the top 19 players.
Aronian arrives at #5 in a big leap although Corus, starting in a few weeks, will be the true test of his status. Kramnik is #6 at 2741, a point ahead of Leko. Mamedyarov jumped past Radjabov on the list. Amazing to see names like Shirov, Adams, and Bareev barely clutching to the top 20. A new generation is coming up and there are still only ten spots in the top ten, which has proven far less exclusive in recent years than it used to be. Only Kasparov, Anand, Topalov, Kramnik, and Leko can really be called permanent members. Svidler looks like a rock at #4, but a few years ago even he was out of the top ten. Ponomariov is back at the ten spot after a few difficult years and he's again looking like the star that won the FIDE KO in 2001.
Not that 2800 is a piece of cake just yet. And it's fair to say that with Bareev as #20, "championship class" really does include 20 players. This is the way it has always evolved. It was almost inconceivable for someone outside of the top few beating Lasker or Capablanca in a single game, let alone a match. Now you've got a few hundred players who can take games from the elite and Sergey Rublevsky just finished ahead of Kramnik, Svidler, and Morozevich in a round-robin. (Naiditsch who?!) So the new age of parity has come in ways other than Kasparov's retirement. Now if only the rating system were more dynamic to allow even greater movement and more accuracy we'd really have some excitement.
I know a lot of people talk about "rating inflation" but I'd like to suggest another possiblity: that more players are stronger, on an absolute scale, than at any time in the past. And isn't ELO supposed to be an absolute scale? In my opinion, the 2700's of today would compete head-on with the (then-elite) 2700's of past eras. And the 2800's of today are stronger than the 2700's of past eras.
I agree. Think of a few analogies. No one talks about time deflation because todays runners are faster than those of the past. Or in a more intellectual realm, no one would compare Albert Einstein to Aristotle, even though both were equally ahead of their peers.
The body of chess knowledge has grown and continues to grow.
With a bunch of '85ers having gotten too old for the junior list, Karjakin and Carlsen, both born in 1990, are now #5 and #7 respectively. Radjabov topping, naturally.
http://www.fide.com/ratings/top.phtml?list=juniors
ComputoJon writes: "And isn't ELO supposed to be an absolute scale?"
I don't know if ELO was designed by Mr. Elo to be on an absolute scale. To me it has the looks of a relative scale. Anybody knowledgeable on this?
While awaiting the New Year, I found several sources which confirm ELO to be a relative rating system, as opposed to absolute.
Wikipedia says: "The ELO rating system is a method for calculating the relative skill levels of players in two-player games such as chess and Go."
Another source (with statistical backup) says: "ELO recognizes the impossibility -- or at least extreme difficulty -- of having an "absolute" rating for a player. Elo ratings are not absolute, they are relative..."
If Arpad Elo (as I recall his name) were alive today, I believe he would be unhappy about the apparent inflation of ratings. The superior play of today's players is not an acceptable excuse for the increased ratings. The increased number of chessplayers partially explains the inflation.
Happy New Year to all who gather here.
Looking over the list of top juniors, it might just be that in ten years we'll have no Russians in the world top ten or maybe even top 15.
Indeed, it's strange to see the megastars of the 1990's (such as Shirov and Adams) fallen to second-tier status, while still only in their 30's.
It inspires renewed respect in players like Ivanchuk and especially Anand, who have managed to remain at the top of their game (literally).
Sadly, they have not included Russian superfinal for Jan Rating list, which would have seen Kramnik falling to No. 7, Anyways, there will always be a bunch of Kramnik fans who will continue to justify his mediocre performance and falling rating with some "piece-meal" statistical arguments. One thing they can't show (nor even Kramnik) is a World Class performance.
Kramnik now-a-days daily watches the new movie "KingKong" to practice the chest-beating : "I am a World Champ and performance does not matter" :))
I think, I surely hurt a minority's (Kramnik fans) feelings, but guys his performance is the weak link in this discussion.
-Amit
Sadly, they have not included Russian superfinal for Jan Rating list, which would have seen Kramnik falling to No. 7, Anyways, there will always be a bunch of Kramnik fans who will continue to justify his mediocre performance and falling rating with some "piece-meal" statistical arguments. One thing they can't show (nor even Kramnik) is a World Class performance.
Kramnik now-a-days daily watches the new movie "KingKong" to practice the chest-beating : "I am a World Champ and performance does not matter" :))
I think, I surely hurt a minority's (Kramnik fans) feelings, but guys his performance is the weak link in this discussion.
-Amit
All Chess ratings are relative - there is no unit of measurement. If we divided all ratings by 10 they would still be valid relative ratings, with GM's being about 260-280 on the scale.
What evidence is there for ratings inflation? It is difficult to maintain 'standard' ratings over time since the pool of players whose relative strength is being measured is constantly changing.
What is needed is a 'standard' chess player whose strength does not change over time. Perhaps a computer program could do this but it would need to be running on the same hardware (or at least the same speed processor).
Personally, I think that the increasing numbers of players with 2700+ ratings are genuinely as good as the few players in the past that had 2700+ ratings. Everyone is aware that chess knoweldge has advanced. Today's players are truly 'standing on the shoulders of giants'.
Sadly, they have not included Russian superfinal for Jan Rating list, which would have seen Kramnik falling to No. 7, Anyways, there will always be a bunch of Kramnik fans who will continue to justify his mediocre performance and falling rating with some "piece-meal" statistical arguments. One thing they can't show (nor even Kramnik) is a World Class performance.
Kramnik now-a-days daily watches the new movie "KingKong" to practice the chest-beating : "I am a World Champ and performance does not matter" :))
I think, I surely hurt a minority's (Kramnik fans) feelings, but guys his performance is the weak link in this discussion.
-Amit
Sadly, they have not included Russian superfinal for Jan Rating list, which would have seen Kramnik falling to No. 7, Anyways, there will always be a bunch of Kramnik fans who will continue to justify his mediocre performance and falling rating with some "piece-meal" statistical arguments. One thing they can't show (nor even Kramnik) is a World Class performance.
Kramnik now-a-days daily watches the new movie "KingKong" to practice the chest-beating : "I am a World Champ and performance does not matter" :))
I think, I surely hurt a minority's (Kramnik fans) feelings, but guys his performance is the weak link in this discussion.
-Amit
The ELO rating is a relative scale.
Imagine that a spaceship arrived from the planet Mars with the sole purpose to improve the chess skills of the earthlings. They would go in orbit around our planet one night and send out mysterious infra-violet ELO-beams, and the next morning when we all woke up, our chess-plying skill would be improved by exactly 100 ELO points.
Would our ratings rise then? No, they wouldn't, because we could only play and win ratings points against each other. So, if everybody's skills improves (or deteriorates) simultaneously, there is no way a rating system can take account for that.
Consequently, when ELO ratings actually DO raise over the years, this has got to be due to flaws in the ELO system itself. The general level of play may or may not be higher today than it was in the past (I think it is) - but there is no way we would know that by looking at the ratings.
I'd like to see the ratings be more dynamic. In tennis, rankings are only based on results in the past year. I wish their was a similar thing in chess. Why should a players rating years ago have any relevance to their rating now?
Okay fine, people latched onto one minor comment (absolute vs. relative ELO) without addressing the major one, namely, that today's players are simply better. What is all of this talk about rating inflation? "We have more players rated over 2700 nowadays." We sure do. Next question? "The average rating of the top 20 players is higher than it used to be." Yep. For good reason, they're stronger.
Sure there is an overall increase in general knowledge, particularly in the openings. But there is no way there are suddenly 25 players today stronger than than the top five players (other than Kasparov and Karpov) just fifteen years ago. Nor do I think they are all stronger than Korchnoi was in 1978.
Plus, the Elo system has been in effect long enough to draw some conclusions from individual careers. You can have a higher rating today with worse results. In 2003 Korchnoi's rating was 2628, just thirty points lower than when he was the clear #2 in the world and facing Karpov for the world championship. As remarkable as he is, his play was not nearly the same level in 2003 as in 1978. His ranking reflected this - he was probably in the top 70 at 2630, but not the rating. Elo is practically worthless for historical comparison.
Quote:
""The average rating of the top 20 players is higher than it used to be." Yep. For good reason, they're stronger."
Sorry, but no.
If everybody got stronger, the average rating would rise. If everybody got weaker, then the average rating would also rise.
So, we look at the ratings and see that they're rising. What does this prove? Nothing, except that the ELO system is inflated. That is exactly the point of the "absolute vs. relative ELO" discussion.
This doesn't mean that they didn't get stronger. It just means that you can't tell by looking at the ELO. You'll have to look at the games.
It is actually possible that ratings are rising without inflation beeing the correct term. With a static system, the time it has been in effect, seen up aganinst the players average lifetime (withdrawn average age for starting to play) will enter the calculation. With players such as Korchnoi, there is actually a possibility that the ratings calculations have not yet reached an optimal amount of data, but will stabilize when the playing population evens out in regards to "births" and deaths.
One "goal" of the ELO system is to maintain the average rating at some fixed number (I'm not clear what this number is or should be). There has been an increase in this number over time, regardless of cause. Whether the players play better chess or whether there is a weakness in Elo's equation...it doesn't matter..it's not what Mr. Elo intended for a relative system. Conclusion: ELO ratings have inflated.
For Goodness Sake. Somebody call Jeff Sonas; let's hear a bit of sense from someone who knows. Go look at his site, and this will all become clear.
http://www.chessmetrics.com/
Elo inflation :
Some part of the elo inflation are linked to the increased number of chess players.
Some other part of the inflation is due to the progressive reduction of the fide minimum rating, who has been progressively reduced from 2200 to 2000, 1800, 1600 and now 1400.
Just try to figure out how it worked when the bareer was set up to 1800 for instance : a lot of players rated between 1500 and 1800 just made two good performances and entered the rating, thus inflating the rating.
My own experience : 15 years ago I was playing a lot, I had a 2000 french rating. I was playing 50 games per year. Now, I'm rated 2188 fide, I never play, and am 100% certain that I'm much weaker than I was 15 years ago. Most of the time I meet 1900 to 2200 fide players, they are authentic patzers, like me. 15 years ago, a 2200 player was rare and strong. Today, there are a lot of 2200 players who play patzer chess.
The inflation that makes that 20 players today have 2700 ratings is much less significant than the collapse of chess understanding at 2200.
Another idea about inflation and titled players : when I started chess, 20 years ago, GM was really a huge achievement, and it was clear that you needed a sound chess understanding and some genious to reach that level.
Today I have the feeling that anybody can reach that 2500 mark. I don't have the feeling that 20 or 30 years ago you could become GM just being a cheap tactician. Today I personnaly know at least 3 GMs who are cheap tacticians.
15 years ago, with my 2000 rating I could do nothing against titled players. Today, against the last 10 titled players I've been playing against otb, I've made 5,5 points. I'm clearly not stronger today than I was 15 years ago (I played 50 games per year, and I'm playing less than 5 per year now). So I think that the inflation has showed much more effect from 2000 to 2400 than above 2600.
"Today I have the feeling that anybody can reach that 2500 mark."
That's a nice thing to say, it makes a patzer feel good. But looking at some of my games, I have to tell you that it's lacking in realism.
I don't know whether he literally meant "anyone" but yeah, you do get the feeling GM's today are of much much lower caliber than, say, in the 1970's.
Are you guys factoring in how your own maturation in understanding both in chess and in life influences your perception of 'inflation'? I would say that today's super-tournaments, Linares, Corus, for example, are richer in many ways than those of 40-50 years ago. Also, today's super-GM's have to deal with shorter time controls and other stressful factors such as a very loud public opinion, not to mention the heightened strength of lower-rated opponents, all of which combine to produce so many upsets and other erratic outcomes. I would say the top players fully deserve every rating point they have.
Ratings are always relative in my opinion. Ratings will keep on inflating. Check the records of previous years. Just 3 years ago, there were hardly 10 players above 2700. Now there are 19. These players play against each other and have more opportunity to increase their rating. So in 3 years time we'll have 5 players over 2800 and probably 10 players in 5 years time.
I disagree with some that said that the increase in rating reflects that players are stronger. Strength is relative too to the period of when chess is played. You cannot compare Fischer's rating of 2785 in 1972 with Anand's 2792 and say Anand was stronger than Fischer...it just doesn't work out like that.
Alphonse -- I agree that now is probably a more difficult time than ever to be at the chess top, if only because of the amount of preparation required. And today's top players deserve full credit for remaining at the top, given the cutthroat competition they have to face. It's just that that has nothing to do with the rating system. It's not a question of whether they've earned their ratings or not; it's a statistical artifact that permeates the entire system, as can be seen in the sheer numbers of GM's and 2500+'s compared to 20 years ago.
"So in 3 years time we'll have 5 players over 2800 and probably 10 players in 5 years time."
I'll bet that won't happen. I bet the next 2800 player will happen in 2011.
"You cannot compare Fischer's rating of 2785 in 1972 with Anand's 2792 and say Anand was stronger than Fischer..."
Who's saying that Fischer's max rating capability was 2785? I'm certain he could have broken the 2800 barrier had he played instead of going into seclusion.
"Today I have the feeling that anybody can reach the 2500 mark."
Thanks for the laughs, i'll let you know when I break the 1500 mark...
There is no rating inflation, the average player is simply much stronger today than they were back then. What has increased the strength of these players? The answer is the computer. Opening, tactics, positional understanding, ending knowledge, all of these are stronger now thanks to computers.
I disagree. No way are all today's 2650+ GMs stronger than Tal. Actually, I would categorically say Tal was stronger than almost all of them, and then some.
This is an issue that comes up in the Olympics all the time. Remember the "4 minute mile" barrier? Once considered unbreakable, today world class athletes see it as a career challenge, not a wall.
So are they better athletes? Obviously the measuring devices (the clock and the length of a mile) haven't changed.
The answer seems to be that today's training methods produce superior results for athletes who are great, but not necessarily the top 5.
What this means is that if a dominant runner from the 50s were trained with today's methods, he would still be dominant--because he'd be much faster than he himself was under old training methods and equipment. (Just the improvement in shoes alone has been estimated to grant a 2 second boost.)
I suspect something similar would apply to chess. Tal starting out today would rise to a higher rating than the Tal of the 50s. But that doesn't necessarily mean the ratings mark itself has changed.
Obviously not "anyone" can become a 2500 player. But just as more people can run faster than a 4 minute mile today, it makes sense that more people can make the 2500 mark.
Below, a very interesting article by a coach who has helped 5 athletes break the 4 minute mile mark. He talks about a specific program to do so, the challenges in getting even world class athletes to follow the program, and differences of opinion on the best way to train.
http://www.pponline.co.uk/encyc/0058.htm
I think the same kind of themes run through many of the "how to become a Grandmaster" articles.
Interesting.
Duif
duif,
"Tal starting out today would rise to a higher rating than the Tal of the 50s." Sure, especially if he had a new kidney, but I meant that the absolute rating is no method of comparison. Distance and time are much more absolute measures than ELO. The strength indicated by 2650 in Tal's time was higher than that indicated by 2650 now is what I meant.
Actually, Jeff Sonas published what I consider to be the definitive word on relative strengths. Dont have the url though.
I can't understand why Mig staes that Corus will be the real test for Aronian status. Who says that a well know tournament is a "real test". I think that every tournament is a "real test" no matter if the players names are knows or if they are more obscure Gms. What's happened the last time Kasparov played outside "The Real testers"?. Completely outplayed by some obscure Rublevsky and humiliated by an unknow Huzman!
What are the "Real Test"?
I think mig stated that Cortus will be the real test because although Aronian has played very well and recently won the FIDE world cup, he has played few games against the Super-GMs this year (superGMs being the top 5-10 players). He played Svidler and Ivanchuk with impressive results. But can he consistently place well in a tournament with players such as Anand, Topalov, Kramnik, and Leko among others?
Dionyseus says, "I bet the next 2800 player will happen in 2011." But Harold Camping of Family Radio says 2011 is the year of the coming of the Messiah. Dionyseus, are you on to something that we should all know about?
Well its clear for me that dionyseus is a member of a satanic cult that foresee the future. Vade Retro Dionyseus!
To clarify: we can see that in any field where humans are dedicated to improving their performance they will, over time, do so as a group, so that the very best achieve new heights. Even on completely objective physical tasks like the 4 minute mile or the pole vault.
The interesting thing is that the result of the average person doesn't necessarily improve, because in order to get that extra level of performance, extraordinary levels of dedication, training, and financial investment may be required. The every day person isn't ready to invest that. But for the very best, additional improvements can be found.
So the top runners get faster. The top scrabble players score more points. The top pole vaulters jump higher. The top race car drivers drive faster. And the top chessplayers have higher ratings.
It would be strange indeed if chess hadn't shown an improvement among its top 1/10th of a percdent, as a group. To the human mind, records are made to be broken, and collectively we usually find a way to do so.
Again, the interesting thing in other competitive fields is that we don't always see improvement in the middle of the pack. The energy goes into making the best even better. It's not a conscious thing. It's is a natural response to competition.