I know they all get worse past age 40 or so. But how much worse. Take somebody who was better than all but one in a million at age 25. Of those who have been in this category who continue playing chess seriously what would the new number tend to be at age 65. One in what?

(I realize that this question is complicated by the fact that modern players have improved and thus even if the old guy is just as good as ever he would slip a lot in the rankings.)

The post-40 decrease is probably not as dramatic as some may think. I believe one of the main factor is that aging chess player spend less time playing and studying chess. One master in my team is in his early 60s and has almost the same level as in his youth, because he still practices every day.

Your one in a million example corresponds to the level of a FIDE master. With enough dedication, I believe a FIDE master should stay above 2200 ELO by the time he his 65. That's about one in 500 000.

We can also take a look at two historical examples to answer your question:

Karpov. He was clearly between number 1 and number 2 in the world from 25 to 40yo. We could say that at 40 yo, in 1990, he was in the 1/(5*10⁹)
Now at 62 yo he is 167 in the world, though he doesn't play enough today and his real rank should be 200-300. That's about 1/(3*10⁷)

Korchnoi. You may prefer this example as he has kept playing chess more seriously.
At 40 he was in the top 5, which is about 1/10⁹. At 65yo, I would say he was approximately in the top 25 but someone else could give a more precise number. That's 1/(2*10⁸). At 75 yo, he was 85th in the world, which is 1/10⁸.

I used to play with a few guys at my club who had been close to 2300 at their peak and played at around 2000 strength into their 70s. I'd say a 200-300 point decline would be typical for aging players.

How about Reshevsky and Walter Browne?

Browne was top 20 at his peak and is now around 1300 among active players. Korchnoi and Smyslov were top 5 at their peak and world class well into their 60s.

I'm not sure about Reshevsky.

I think 200-300 is a good average loss estimate between from 40-70.

And the marginal loss certainly increases over time. People's level decreases much faster between 65-70 than 40-45.

Emanuel Lasker was still world champion at age 53 and is estimated to have last achieved No. 1 ranking as late as 58 years old. The following paragraph is from the wikipedia article on Lasker:

Statistical ranking systems place Lasker high among the greatest players of all time. The book Warriors of the Mind places him sixth, behind Garry Kasparov, Anatoly Karpov, Fischer, Mikhail Botvinnik and Capablanca.[116] In his 1978 book The Rating of Chessplayers, Past and Present, Arpad Elo gave retrospective ratings to players based on their performance over the best five-year span of their career. He concluded that Lasker was the joint second strongest player of those surveyed (tied with Botvinnik and behind Capablanca).[117] The most up-to-date system, Chessmetrics, is rather sensitive to the length of the periods being compared, and ranks Lasker between fifth and second strongest of all time for peak periods ranging in length from one to twenty years.[118] Its author, the statistician Jeff Sonas, concluded that only Kasparov and Karpov surpassed Lasker's long-term dominance of the game.[119] By Chessmetrics' reckoning, Lasker was the number 1 player in 292 different months—a total of over 24 years. His first No. 1 rank was in June 1890, and his last in December 1926—a span of 36½ years.[120] Chessmetrics also considers him the strongest 67-year-old in history: in December 1935, at age 67 years and 0 months, his rating was 2691 (number 7 in the world), well above second-place Viktor Korchnoi's rating at that age (2660, number 39 in the world, in March 1998).[121]

From time to time, there are players that defy generalizations about the effects of old age on playing strength. Korchnoi is another one. And there is also ex-world champion Vasily Smyslov who qualified for the candidates matches at the age of 62! After getting past Hubner and Ribli in the quarter and semi-final matches. It took no less than Kasparov to stop him in the final.

In 1988, at the age of 51, Mikhail Tal won the second official World Blitz Championship.

I will be the first to admit that I know very little about chess history. But I'm assuming that pretty much everyone that's been mentioned so far had achieved a very high rating much earlier in their lives and that they saw a certain amount of decline as they aged.

What if someone comes to chess later on in life? I saw 40 mentioned as an age where performance starts to decline, and this makes sense. But what if a 35 year old picks up the game and studies hard for many years? I would imagine that in this case the person in question would keep improving for many years, maybe even into their late 50s. Is this a fair statement?

Also, would the 60 year old who had been playing for 25 years be in a similar position to a 60 year old who had been playing for 45 years? How long does it take to reach a peak in chess? and once you reach a very high level, how hard is it to make improvements? Is there a point where two players, even though one had been playing for 20 years less than another player, would be at levels very hard to distinguish from each other?

This thread may be of some interest: When are a chess player's peak years?

Thanks gansta! Great thread.

Browne probably spent more time grinding poker than studying chess.

I think a tangent relevant subject to explore is if every motivated individual is capable of reaching a certain ranking through practice [e.g. IM strength] and the length of time necessary to reach that ranking [5 years, 20 years?].

I'm sure at 65, someone who's never played chess is capable of learning and reaching a certain level, but how much more drastic would it be had he started at 40, 20, 10, or 5 years of age?

Relating to the post, from my observation of the chess scene, I think most players retain their strength and sharpness well into their 60s.

Is that their actual decline? Or is some of the decrease caused by players getting better?

No, definitely decline in strength due to aging. Chess requires a lot of energy and older players don't have the stamina to apply maximum focus over a 3-5 hour game. They play mostly based on experience and understanding. That can work sometimes, but in positions where you have to focus and calculate variations to reach a solution, older players will struggle.

My question is about the method of ranking. If you remain the same and everyone else gets better does your number rating go down?

I'm probably nitpicking here, but I don't think there is such a thing as an ex-GM. Once the title is granted, it remains for life, no matter what a player's rating does. Just mentioning this in response to the thread title.

http://en.wikipedia.org/wiki/Elo_rat..._and_deflation

Most people agree that at top level there is ELO inflation anyway.

And yes, if you remain the same but everyone else gets better your elo should drop. Except maybe at top level because of inflation.

Last question. Take someone like Dan Harrington who I believe was about 2400 at his peak and took chess seriously but not super seriously. Now he is about 70. If he spent a year intensely studying with Kasparov would he be favored to hit a new peak?

I'm pretty sure this would be impossible. After 60yo you can't really get better at master level, no matter who you train with.

Yeah I think he was probably too good at his best to hit a new peak now. If you took an old guy who was 1600 or maybe even 2000, perhaps he'd be favoured to hit a new peak with a year's intense study.

If he really was not super serious and now was, I think he could surpass his former ability for sure and rating probably.

Once you've played a certain number of preliminary games, your rating reflects purely how often you win, lose or draw in relation to people who win, lose, or draw a certain amount. So if everyone else improved the same amount, they would all win/lose/draw with the same frequencies against each other, and your ratio of wins/losses/draws would adjust downward, sending your rating down with it.

given that FIDE didn't even manage to strip Alexandru Crisan of his title, you're probably right.

+1

But if the inflation was worth like 2-300 points rating then maybe he could hit a "slight" peak, but i do not think it is that high tho .

There us a nice exampke in Go. Fujisawa Shuko won the top professional tournament in the world at the time at the age if around 56. He held onto it for the next 5 years.

He attributed his success to not drinking during the tournament.