According to a recent NY times article...

today's 40th ranked player equals 1970s Karpov.

seems reasonable

Does not seem reasonable to me. If Karpov of the 1970's were playing now (and not 'today's Karpov'), I say he'd be in the top 5 (at least). I hate trying to compare players of different eras - to me this does not make sense and can't really provide any useful information (despite all of Jeff Sonas' valiant attempts). Today's players have learned from Karpov, and are thus operating with a superior knowledge base, so skill levels can not be compared fairly. Furthermore, there is rating inflation. The sort of comparison used above might state that the 100th best player today would defeat Paul Morphy - but again, if Morphy were alive today I say, like Karpov, he'd be in the top 5. Games changes, knowledge base changes, conditions change (hello Rybka). One more example - is it fair to say that the 40th ranked tennis player today is better than, say, Rod Laver? To me, this type of comparison is equally unreasonable.

(meow)

Rating inflation isn't really relevant, the article in question compared a player's average move to those of several skilled bots, and was primarily about detecting cheating. (Either that or there've been two similar articles recently.) The point about the greater knowledge base is a fair one though, and I think that point is actually the reason the article brought it up, not that a large crop of geniuses picked up the game who are far superior to anyone who came before.

Innate talent among the top players isn't getting better at that rate. It may bot be getting better at all, depending in part on what proportion of the innately gifted learn and make a career out of chess.

But the tools for training are so much better now that I don't find this result surprising. If they said he'd not be in the top 100, I also wouldn't have been surprised, though if they said he'd be 10th, I wouldn't have been, either.

Early 1970s or late 1970s? By 1978 he should be nearly even with 1972 Fischer, and even in absolute terms, ranking that as 40th is ridiculous.

"He has also discovered that the way people play has evolved. According to his analysis, the player now ranked No. 40 in the world plays as well as Anatoly Karpov did in the 1970s, when he was world champion and was described as a machine."

Here's the article in question, http://www.nytimes.com/2012/03/20/science/a-computer-program-to-detect-possible-cheating-in-chess.html?_r=2&hpw=&pagewanted=all

Playing well doesn't just mean playing what a computer would consider the best move. It means playing a good move that will put you in a position you know how to handle. So, with computers playing such a large part in training now, we can expect that Regan's measure of how well people are playing show more even rapid improvement than probably actually exists.

But whether that's true or not, there is very rapid improvement. 10 or twenty years ago (I forget exactly when) most of the professional players older than their early 30s retired from major tournaments all within a few years of each other. They just weren't used to the new training regimens and knew they wouldn't be able to keep up.

A similar issue occurs in historical baseball simulation, which is what I'm working on professionally. Compared to his peers, Babe Ruth was astonishingly good. But if you could take him mid-career and use time travel to bring him to the current day while giving giving him only a couple of weeks to learn how the game is played now and no change in his physical training, he would struggle.

The other reason why this sounds strange is the difference in motivation. Maybe most of the people here were not alive in the 70s and 80s, but chess was an amazing sensation, especially in the soviet block during the cold war. World championship games opened the news in the whole western world. Chess players were celebrities. The average person on the street who watched news or read newspapers could tell who the chess world champion was and his opponent. Before Deep Blue chess players were regarded as the most intelligent humans on earth. And after Deep Blue they are considered geeks (except by us geeks, of course ;-).

Maybe the game has been so much influenced by computers that a computer measuring human strength is not such a great idea. But I don't really know. It may also be true.

If Karpov of the 70s was alive today, and had access to all the tools and resources that he didn't have in the 70s, he'd be better than he was in the 70s. So if a raw 70s Karpov would be around 40th, it wouldn't surprise me one bit if Karpov++ would be at least 50 ratings points higher, which would plonk him around 10th.

We mustn't forget to train up all the other greats too, though, to be fair to them all.

I agree MisterCat completly, you can use his arguments on lots of other sports.

Besides sports, that line of reasoning cuts ice in other domaines of our society like education level, ...

Ratings are not a measurement of talent but actually of playing level. Chess has evolved ergo the rating "inflation". That been said, although I'm willing to accept computer evaluations as a guidance they are far from an indication of perfect play.