On the other hand, some of those early achievers then fizzle down into the crowd of strong but not elite GMs. Take Illya Nyzhnyk, for example. He was all the rage at some point, and now, at 28, he is not even in the top-100 list. (In fact, I wonder if he ever was.) I think the list of incredibly promising players that become just-one-more grandmaster is long.

https://uscf1-nyc1.aodhosting.com/C...

Nowadays, no-one would give much notice to a thirteen year-old USCF master.

** Fusilli's Theory on the Declining Promise of Chess Prodigies **

Chess prodigies of the past were more likely to leave their marks in chess history as (adult) elite players of their time than chess prodigies in the 21st century.

Reasons for this:

a) In the past, remarkable talent at a very young age meant remarkable natural skill (and intuitive understanding of strategy), since the number of top trainers was limited and computers did not exist.

b) Furthermore, the world of professional chess players was small and selective, reducing the ability of medium-level talented players to get exposure, compete, and reach their full potential. (Just think, for example, of top-level South American players with limited exposure to top competition in Europe throughout the 20th century.) This would have given early starters a permanent advantage. Today, there is a lot more competition because chess has become both more global and more accessible.

c) In the 21st century, remarkable talent at a very young age is a combination of natural skill (as in the past), and learning of an intensity and quality far greater than ever before, thanks to the density of superior coaching and the ubiquitousness of chess engines.

Point (c) means not only that talented players reach their peak earlier, but that many can ascend very rapidly not just because of their natural skill, but also by force of memorization and superior tactical learning (computers are surely superior coaches at this).

This means that some players can become prodigies, and early grandmasters, even if their natural skill is not at elite level. But in adulthood, it is that natural skill that makes the difference between the elite players and the other GMs. (FWIW, I think the ultimate advantage of elite GMs over other GMs is that they have an edge on their understanding of position and strategy. But this is VERY debatable.)

Of course, I am introducing concepts that need definitions. What is "natural skill"? How do we measure all this? Do we have good data on prodigies of the past who failed to thrive, or do we only know the success stories (which would bias any empirical evidence in favor of my theory)?

I do believe that we all can agree on the following:

a) Many chess prodigies of recent times have fizzled. Poster child (pun!) for this: Illya Nyzhnyk

b) Chess has gone global and the number of players and tournaments is larger than in the past by an order of magnitude. That is obvious, but what matters is that chess competition is more open and accessible than in the past (do we agree on that?) which means prodigies have a harder time climbing the ladder and staying on top (if they get there).

<I doubt this last part.>

Do you agree with everything else?

<I would argue that it's now easier to maintain high standards for longer, assuming, of course, one is still prepared to put in the work.>

I guess we need to gather systematic data and check.

Starting with the youngest GM ever, this was Sergey Karjakin in 2002, whose record was surpassed by Abhimanyu Mishra in 2021.

By October 2005, Karjakin was 44th/46th in the world, while Mishra today is 152nd/155th.

The Wikipedia entry on chess prodigies has a list of players who reached GM before age 14: https://en.wikipedia.org/wiki/Chess....

Of course, one can select the players you listed and conclude that prodigies have long-lasting batteries. But then I see names on that list that I either never heard of or I have forgotten because they fizzled.

Kids surely learn faster than kids in the 20th century by a long, long shot. It took 33 years for Fischer's 1958 youngest-GM ever record to be surpassed by Judith Polgar in 1991, but we have gotten used to that record having to be updated much more frequently.

My thesis is that those who showed huge talent early on in the past were more likely to remain top world class players in adulthood, but to get data from the pre-computer generations, age at which players reached GM will not be useful. We might need to go by something like age at reaching a rating threshold, adjusted by rating inflation.

By implication.

<I guess we need to gather systematic data and check.>