17 year old Firouzja notches up a 2877 performance to add to his other achievements over the past year…
Places him 17th in the live ratings
The fact that there are pre-teen grandmasters and autistic savant masters suggests that chess mastery is significantly a matter of aptitude rather than learning. A meta-analysis of 19 studies (full text not available) on chess skill and intelligence found that chess skill is, not surprisingly, linked to general intelligence, but that this link is stronger in those who are younger or of weaker skill. The better people are at chess, the less the differences between them can be explained by intelligence. The lead study author offered this analogy:
By way of comparison, Burgoyne said, consider the world’s best basketball players. Although there is essentially no correlation between height and points scored at that level, that doesn’t mean height isn’t important in basketball.
The fact that intelligence correlates with chess ability thus does not suggest that intelligent individuals can become GMs via intensive study, any more than the association of height with basketball ability suggests that any tall person can become LeBron James if they try hard enough. The existence of very young GMs strongly suggests that aptitude plays a major role.
This is speculation, but I would guess that one of the necessary skills is visualization. People differ more strongly than most realise in visual ability; at the extreme end of the inability to visualize is aphantasia. As far as I’m aware this is unrelated to general intelligence. I don’t have aphantasia, but I have unusually poor visualization ability. The ability to play a chess game blindfold is well beyond me. I can’t even play at beginner level, I just lose track of what the board looks like. As far as I’m aware, all masters can play blindfold games at a high standard and many can play multiple games simultaneously. I’m just never going to be able to do that. It’s not a matter of training, my brain just doesn’t work that way. It seems obvious that this hampers my ability to calculate accurately at chess. This is probably also why I much prefer blitz, where intuition and pattern recognition are more important than ability to accurately evaluate positions 10 moves down the track.
Edit: It’s possible that a large number of intelligent people could become GMs if they studied very intensively as children; the Polgar experiment suggests this could be true, but I don’t think that has any bearing on whether intelligent people can become GMs later in life. It might be similar to learning new languages, which is easy as a child and very difficult as a monolingual adult.
4 Likes
I wonder how much he could accomplish just going all in on opening prep. You don’t need much of a special talent for that, and while he’s not going to beat any super GMs with it, maybe it’s good enough to get to 2300-2400.
I doubt opening prep would get you anywhere. You don’t get better than like a pawn advantage out of opening prep and a 2000-level player is still going to lose to an IM with pawn odds. If you listen to a guy like Nakamura, he was playing like the Dutch Defense and Grand Prix Attack as primary opening choices up to IM level. Detailed opening prep only became relevant at master level.
I just encountered some interesting neuroscience research about chess ability.
When thinking about a move, grandmasters use the frontal and parietal cortices (areas associated with long-term memory and higher-level processing) much more than amateurs, who use the medial temporal lobe. This difference is very pronounced:
The thing is, Ognjen Amidzic, who did this research, believes that this ratio is set at quite an early age and doesn’t change after that. In fact, he’s developed a test for children which predicts whether they are capable of attaining mastery or not.
1 Like
I’m not sure about that but I w/e I could be wrong. Obviously +1 positions range from being very easy to play to next to impossible play, but a 2000 player should be able to beat an IM in various exchange lines with relatively few pieces left if they somehow get a clear pawn+ advantage.
I think you’re conflating “a clear pawn+ advantage” with +1 positions here. A simple, winning pawn-up endgame is not +1 in an engine, it’s like +4. A typical IM is rated like 2400 to 2450, which based on how ELO works suggests that they should score like 95% against a 2000 player. That a 2000 level player will need a decisive advantage to win against a 2450 is sort of true by definition.
Everyone you play against at master or near-master level is going to know the lines they play well enough to keep you from getting a decisive advantage out of the opening. If you play an IM and open e4 and they play the Caro-Kann, you can have ChessBase in front of you showing you all the latest Caro theory and there is still no way you’re going to get anything like a decisive advantage. At best you’ll get like +0.6.
No, I’m not talking about converting +4 endgames. You mentioned a pawn advantage out of opening prep and that a 2000 level player is losing to an IM with pawn odds. As I say, I could be wrong, but that feels like too big a gap. There are +1 positions that are very intricate and there are +1 positions that are relatively straightforward, and I think a 2000 player should be able to convert the relatively straightforward ones against an IM, if they are able to get there. I think if I practiced I would consistently beat a 2200 CPU with pawn odds (400+ ~my rating).
Let’s say the IM wants to show off and plays the King’s Gambit against the 2000 player, you don’t think you the 2000 player gets a decisive advantage if they had stockfish playing the opening? I’m pretty sure all King’s Gambit lines get to -1 pretty fast.
That was my point in the Caro example, like yes having ChessBase in front of me with King’s Gambit theory is probably a pretty big advantage against an unprepared IM who is forced to play the opening, but you don’t get to play at master level against opponents who are unprepared in their opening. Against an IM who is for some reason expert in the King’s Gambit, I would expect to get crushed at it even with ChessBase in front of me. I would obtain a somewhat better position in an open game and then get crushed.
Edit: I guess to clarify, if you took the world’s foremost King’s Gambit theorist and said “with best play from Black, what is the best position White can obtain at the limits of your knowledge” and then put a 2000 level Black player against a random IM as White in that position, my money is going on the IM.
1 Like
I also just kind of disagree with this, maybe provide an example? Like I agree there are easier and harder +1 positions to play. What I disagree with is that there are any +1 positions that are decisive because I think that’s a contradiction in terms. If an engine plays against itself at a +1 position at blitz time controls it should only win something like 70% of the time; that’s the definition of a +1 position, whereas if you limit an engine to 450 ELO weaker than its strongest rating, it will beat itself like 95% of the time from the starting position. It therefore seems true by definition that 450 ELO is a stronger advantage than a +1 position. If a +1 position provides greater advantage than 450 ELO, then it isn’t a +1 position.
1 Like
Grunch, but what do you nature vs nuture chess bros think of Polgar and his whole “If I had a son, he’d be world champion” or whatever. And didn’t all of his daughters become at least women’s GMs.
EDIT: I now see Chris touched on it briefly.
If IMs in practice are too good at openings such that they will rarely if ever misplay openings of their choice to yield a greater than pawn advantage to an opening book, then I underestimated IMs, and studying opening theory won’t do much to help a 2000 player beat them. I’ve played a handful of 2200 blitz players and I’m reasonably confident that if I studied nothing but opening theory for 3 months I’d close a 300 ELO deficit by half. Ofc, 2200 blitz is not a FIDE rated IM, and blitz is barely chess anyway, but this where my take is coming from.
Incidentally, the King’s Gambit may have been a poor opening to try to make my point, take a look at this mess playing the most frequent move on lichess’ opening explorer:
A 2000 player is a huge dog to make a draw here against an IM, although the eval is -1.
To get you an example of a +1ish game that’s relatively straightforward to play I’d have to look through some of my games, sadly I don’t keep notes and searching through chess com is a pain in the balls.
This was from a 1750ish vs 1750ish game a bit ago. I understand that an IM is not getting to a position like this as black, but a 2000 player should beat an IM from here as played (+1.7 white at the free tier chess com analysis thing), but I could be wrong.
There might be some correlation between being among the best poker players in the world (in general, not actually saying Timex ever was one of the best. not interested or have an opinion on that argument) and having innate abilities for chess.
Either way, Bill Perkins snap accepted the bet and Timex eventually backed out as it makes no sense to dedicate his entire life even if he thinks he has a tiny edge over the 1:10 line.
Yeah like 1.7 > 1, but even that position where Black’s shortcomings are obvious, I doubt I score above 50% playing like a 2200.
1 Like
Yeah visual memory is part of it but the actual visualization part is taxing for me. Like I’m pretty sure at the point I lose track of the position in a blindfold game, I would have no trouble reciting all the moves that got us to that position. I just can’t see clearly enough what that looks like on the board.
I wonder if that has implications for memory as well - like that neuroscience researcher said it appeared that what the GMs were doing was sort of building up a library of positional motifs in long-term memory. I frequently see masters or even enthusiastic amateurs who are like “oh yea this is just like the game Korchnoi - Karpov, Linares 1995” or something and I’m like, what the fuck. I can have watched a game the day before and have only a dim idea of how it went. It’s known that people with aphantasia have difficulties with autobiographical memory, which is the kind of memory where I say like “what did you do last Saturday night”, which is a problem for people who can’t get the answer via visual recall.
I played a league game against an FM (Peter Sowray) once, a ridiculous mismatch, and decided my only hope of not being humiliated was finding an obscure but still playable line that he wouldn’t be too familiar with.
He opened d4, I took him into the rare Grunfeld line and he took me apart and afterwards told me that line hadn’t been used much since a couple of games between Symslov and someone else in the 1950s lol.
3 Likes
I recently played a blindfold game against a beginner. It was way more intense than any other game I’ve played. I did manage a win, even though I blundered a bishop because I simply forgot it was there.
I just started playing so here is a big facepalm move based on the engine EV thing. I somehow make a move that swung me from +54 to even, seems hard to do!
The Ke5 on move 51 was about the same EV. I had a rook and king left while they had a knight and king, I was trying to chase down the king and ended up drawing
