Inspite of having such a large number of possible positions and games why do high level players bother about some specific game from some million years ago ? It may be helpful in understanding techniques and tactics but memorizing them seems very unusual. In fact memorizing games puts the person with more calculating ability at a disadvantage in time constrained games. Shouldn't the goal be to develop the chess aptitude rather than memorizing.

  • 2
    You'll be interested to know that when top chess players are briefly shown a random position from an actual game, they are often able to recall the position with stunning accuracy. That's because the position tells a story - what each player is trying to do and the tactical paths open to them. However, when you show grandmasters a completely scrambled board - that is, one where all the pieces are placed randomly - they really struggle to recall what went where. The pieces no longer tell a story. – Ingolifs Jul 17 '20 at 1:22
  • @ingolfis yes, that is probably because of their training. – shashank shekhar singh Jul 17 '20 at 3:51

Memorization is mostly a side effect, not the end goal. Top players can spend hours or days analyzing a single game to try to understand all of the instructive ideas. Memorization naturally flows from that.

  • 19
    One way to understand this if you've ever worked really hard writing a research paper, you will end up having its content memorized, even though you did not try to do this. – StaleMate Jul 15 '20 at 2:40
  • @StaleMate irrelevant analogy. Research paper you memorize your work, not every word and when you wrote it , where you wrote it and all that. – shashank shekhar singh Jul 15 '20 at 5:34
  • 11
    @shashankshekharsingh Well the exact text of a research paper is just more information than, say, a hundred chess moves. Also, there is more logic connecting the moves than the words. I would expect that while studying a paper one would probably memorise the ideas in a paper perfectly and in order since they logically flow from one another, but the exact words are not as tightly related (there are many different ways to phrase the same ideas). Chess moves are similar in the sense that one flows from the other mostly (both by continuing a plan or responding to threats). – 11684 Jul 15 '20 at 12:04
  • 1
    @shashankshekharsingh, No. When I say memorize, I mean memorize. Word for word. Maybe I am different, but when I have gone over it countless times, making it sound exactly right, I end up being able to basically write it out from memory, word for word. – StaleMate Jul 15 '20 at 18:41
  • 1
    @jpaugh I assure you, the choice of the next move is more restricted than the choice of the next word in a sentence. In a text there is always the options of inserting a subsentence, word order is mostly flexible and there are very few situations where you can’t have your pick of a number of synonyms. In chess, you can’t just insert a manoeuvre while executing a tactic, move order is only occasionally flexible and the only true “synonyms” I can think of are two moves that are both mate in 1. – 11684 Jul 18 '20 at 19:47

It's not as if top players sit down to memorize an old game. They don't spend time only memorizing, memorization is not a goal.

But they study lots of games. Old games, new games, famous games, their own games. Constantly.

And they have amazing memory so they remember these games, as a side effect of studying them, move by move.

Of course there are also players who don't quite study constantly, or who don't have amazing memory. But they tend not to become top players. Memory is very important in chess at the very highest level.


"Seems very unusual": source? Every time I've heard a good player (from club-competent to world-class) talking about his games, they were remarkably capable of precisely recalling positions. So I would say that

  • it is "very usual";

  • it stems from thinking long and hard about those positions, either during games or while studying;

  • it seems like a good trait, to the extent that I assume that a particularly good memory is a requirement for a good player.

  • "unnatural" would be a preferable word. there is nothing wrong in memorizing games , but if you compound it with the fact that so 'good' many games are possible, why do they care about some individual game, not only that they frequently cite it in conversations , the game , the place where it was played , date and time etc et. I find that bizarre is what i want to say – shashank shekhar singh Jul 15 '20 at 5:20
  • 7
    @shashank: if you find this behaviour weird and counter-productive, then I recommend you avoid baseball fans and (especially) coaches! Not that they typically have every pitch of past games memorized, but the interest and ability to retain information about games, basically comes in their case just from enjoying the sport they follow/play. – Steve Jessop Jul 15 '20 at 12:27
  • @steve i am from india, i don’t know a thing about baseball. But nice to know that. Thanx – shashank shekhar singh Jul 15 '20 at 13:45
  • 5
    @shashankshekharsingh: you are free to find that bizarre. Other people disagree, they are sort of obsessed by chess. Somehow people obsessed by chess have a greater chance of becoming a world top player than people who don't care that much. – RemcoGerlich Jul 15 '20 at 14:00
  • @RemcoGerlich of course people disagree, its a free world. Obsession doesn't mean you can be good at chess. I am sure there must be someone (maybe a gm) who has more positions and games memorized than carlsen but his calculating ability is a different thing altogether. And at the highest level ,IMO, this ability to find positions and lines is more required than memory. at some point in the game , the memorised variations have to be parted with. – shashank shekhar singh Jul 15 '20 at 17:14

When studying an old game, you can take however long you want to analyse if its the best move. The conclusion that this is the best move came from "calculating ability" and not from "memory".

There is not need to "calculate" this again when in a match. It's faster and easier to play from memory.

Kinda the same as when learning the times tables as a kid, you first learn to understand how they work. Then you memorize them for quick and easy use.


Memorization of entire games from a million years ago was never used as a method to train young players; or at least not those players who went on to become grandmasters. (There are other things that do end up taught repeatedly and usually mainly memorized, like crucial opening tactics and endgame strategies.)

It is easier, for a strong player, to memorize somebody else's game just after seeing it played out once than you might think possible, because chess has taught them the logic connecting the moves together. In fact they were likely able to predict many of the moves even before they learned which move was actually made in a given position.

A strong player will be good at remembering chess-like positions as well as sequences of strong (natural, logical) moves, and they will be correspondingly weak, compared to weak players, at remembering "positions" created by randomly positioning pieces on the chessboard, or poor, nonsensical, randomly selected moves. It is because every position objectively determines the strongest continuation or a few strong options, but there's often fifty more weak moves, including many that no player would normally consider in the given position.

  • Exactly. The only thing they'll have to memorize is A) the starting position, and B) any moves that they couldn't predict --- and our brains are very good at memorizing "surprises," anyway. – jpaugh Jul 17 '20 at 18:11

If there is a long complex line it is easier than trying to figure out the right move while the clock is ticking.

  • that's why said it is kind of unfair. When nakmura played a position just from memory in one of his streams , chat said - " don't cheat" of course jokingly but i didn't find that funny at all – shashank shekhar singh Jul 15 '20 at 5:22
  • 5
    Is there a reason you are using two accounts? – TheSimpliFire Jul 15 '20 at 13:54
  • @TheSimpliFire Unless I missed something that you saw, it's possible for two unrelated people to have the same display name. It's not unique like "account names" on other websites. – jpaugh Jul 17 '20 at 18:14

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.