Has anyone here made a point of memorizing entire games? I can't remember where I read this, but some chess trainer recommended memorizing 100 master games as a good tool for improvement. If I do memorize games, i was thinking a mix of classics and modern games that are models in openings I play. Sound good? Waste of time?

Thanks.

i've memorized thousands of masters games but not with the intention of doing that... the simple act of studying a master game permanently imprints it into my head. I can't forget a game once i've studied it.

trw wrote on 06/21/08 at 19:11:50:

i've memorized thousands of masters games but not with the intention of doing that... the simple act of studying a master game permanently imprints it into my head. I can't forget a game once i've studied it.

Wow. I think this must be a huge advantage for anyone studying to become a stronger player. And not only for games of course, but for opening and endgame theory as well. I can play over a game 3 times or more without remembering it for any length of time! In fact I would have trouble reproducing even Anderssens "Immortal" and "Evergreen" games or Morphy's "Opera game" from memory, though I have seen them countless times.  :'(

Is there a secret method? Do you study in any particular way that (deliberately or coincidentally) makes the games stick in memory?

Huh?  It isn't April 1 ...

(to clarify, I'm not referring to the OP; I know Rashid Ziatdinov has recommended such a thing, as doubtful as it sounds to me)

Okay, serious answers please. Whoever I read recommended a goal of memorizing 100 master games total. What i'm talking about is memorizing all the moves of the game as well as critical variations and at least understanding the important, but non-critical variations. Obviously also really understanding the themes of the game, the opening ideas, endgame ideas or theoretical positions, etc.

The more that I think about it, the more I like this idea. I've finally gotten to around 2000 and I'm need a break from my standard training.

Yes, this has been suggested. I remember reading about this here:

http://www.geocities.com/lifemasteraj/training.html#The

See Cool  Training Tip#18.

Personally, I have never tried it. I would be interested in hearing for persons who have actually done it.

This question and trw's post brings to mind something I just read whilst browsing in a bookshop two days ago. I was reading Dvoretsky's book on Opening Preparation and in the chapter on opening repertoire, he recommended a game as a sort of model game to play through, study and memorise. I think this is in a sense an extension of the concept of tabiyas. Dvoretsky recommends finding such similar games to memorise but only games where both sides are playing "logical" good chess. So it is not necessarily the most brilliant chess game or the game with most brilliant combination or tactic.

On reflection, this may work in the sense similar to the european medieval practice of memory cathedrals. In days when writing was scarce, memory and oral tradition were stronger. Certain persons developed practices of memorising tons of stuff. One way of doing this (and in an organised manner) was to imagine a cathedral (we are speaking after all of the medieval europe where the church and christianity was very influential!). When you need to store somethin in your memory, you must imagine your cathedral, enter it (give it different rooms), imagine a new object (in a sense similar to mnemonics), give it a connection to the memory, memorise, then exit. The memory should stay. cf an old book by Fraces Yates if anyone is fascinated.

Back to chess. I think Dvoretsky's idea is similar. If we remember certain model games, we can dredge it up when in competition, and use it guide us on playing the logical good chess. (Caveat: I am not saying the creative, intuitive and brilliant has no place in chess thinking. To paraphrase, Chess is 99% hard work and 1% inspiration!)

Apologies in advance for the long post.

PS. I need to work on my memory 'cos I have forgotten what game it was that Dvoretsky's had in mind. Perhaps someone with the book can post the game details.

I think it has more to do with understanding yourself and how you best learn. For me it is great because my biggest strength is my memory. I am a visual learner and thus playing through the game over hte course of an hour sometimes two or three (depending how the long game is!) naturally commits it to my memory. I cannot learn a game by going through it online or fast... it does not work this way for me because I am a visual learner.

My friend on the other hand learns through associative memory and he memorizes games by listening to the same song on repeat and then all he has to do to remember the games is start playing that song in his head... sort of what Gabriel was referring to.


I do think it helps to have a strong set of games in your head... I encountered a position I had never seen before yesterday in the KID and I simply recalled a similiar position played by Gelfand and was able to understand how he played this similiar position.  It helps to have these plans on access even if they aren't something you will be using yourself... it starts your mind thinking on the right path. What are the key squares?

I know many russians masters have advocated '100' games sorta in the fashion of Rashid's 300 core positions idea.

Memory and chess combines two areas of study that have fascinated me since adolescence.  (No, they aren't the only areas of study that has fascinated me since adolescence, as my wife can attest.)

I agree with most of what has been written here in a general sense, and specifically the memory cathedral is indeed an excellent tool.

trw said he is a visual learner and then stated that he couldn't memorize games just by looking at them.  He has to play through the moves.  To me, this is more of a tactile and geometric learner rather than one with a photographic sort of memory.  (The term "identic" is often used instead of photographic.)

Feynman wrote about experiments he made with his friends about memory in his red book, "What do you Care What People Think?"  He discovered that people really do learn and memorize and THINK in different ways.

This isn't a surprise to me anymore, but when he wrote that, it was probably groundbreaking.  I've mentioned Eric Kandel's black book (with a picture of his face on it), In Search of Memory in other threads.

He won the Nobel prize for his work on short-term and long-term memory. He wanted to know why he remembered individual events such as leaving Austria as a young boy so clearly and yet had difficulty remembering other things that were routine. (For example, what did you have for dinner three weeks ago on Tuesday?)

His answer was to study the physical nature of memory.  Once we realise that memory is a physical activity, we can work our memory "muscles" to do amazing things.  London Taxi cab drivers are famous for developing their memories.

The techniques for memorizing are important because memory alone is useless without a way to bring it to the surface.  That's why the Memory Cathedral is such an excellent idea.  

Now, does memorizing a certain number of complete games make you a better player?  Or maybe just a certain number of positions?  One IM suggested memorizing not 100 games, but 1000 positions.  I really don't have the answer.  If I did, I'd write a nice book with friendly words like "Don't Panic" on the back of it and make lots of money.  Preferrably Euros rather than US Dollars, but I wouldn't be too picky.

Dvoretsky did mention that his memory was not that special.  Remember, he was comparing himself to other chess professionals.  I know that I don't forget games that I've played, even in simuls.  I may have trouble placing a face with a name, but once someone tells me what they played I can remember where they sat and the rest of the game.

Is that a visual sort of memory?  A tactile one?  I don't really know.  I do know that my aural memory is not as good as my visual one, but I learn best by doing.  (Don't most people?  Apparently not.  At least according to that black book by Kandel.)

I think you meant eidetic, not identic.

kylemeister wrote on 06/22/08 at 03:39:39:

I think you meant eidetic, not identic.

Kylemeister, I made so many syntax errors in that post that I thought you might have caught on.  Perhaps if I had said I have a photogenic memory?  Oh well.

Still, I appreciate your reading my notes carefully.

I thnk where having a good memory helps most is in endings, if you can remember say 200 important theoretical endings off pat then you would be carrying around your own table base

And thsi would be much more valuable then memorizing some transient opening theory- I think that was one of the ways fischer used his excellent memory.

Also to me  the detail in Endgames is harder to remember as there are less signposts in the position

There was a famous game v Taimanov in 1960? where he had memorized Averbakh's analysis of Capablanca Janowski and just drew the pawn down bishop ending really easily

trw wrote on 06/22/08 at 02:38:39:

I think it has more to do with understanding yourself and how you best learn. For me it is great because my biggest strength is my memory. I am a visual learner and thus playing through the game over hte course of an hour sometimes two or three (depending how the long game is!) naturally commits it to my memory. I cannot learn a game by going through it online or fast... it does not work this way for me because I am a visual learner.

That makes sense. I very rarely spend several hours studying a single game, but maybe I would remember them better if I did... Do you feel it is worth the time investment? Actually I do remember games I have "studied" for several hours, namely those I have played myself! But after a month or so I start forgetting even those.


GabrielGale wrote on 06/22/08 at 01:39:36:

I was reading Dvoretsky's book on Opening Preparation and in the chapter on opening repertoire, he recommended a game as a sort of model game to play through, study and memorise. I think this is in a sense an extension of the concept of tabiyas. Dvoretsky recommends finding such similar games to memorise but only games where both sides are playing "logical" good chess. So it is not necessarily the most brilliant chess game or the game with most brilliant combination or tactic.

If memory serves me well  ;) this was a game played by Dvoretsky himself in the Closed Sicilian main line. Which games apperar logical may actually differ a bit from person to person, i.e. it will be easier to remember games with position types or strategies you already have some experience with, since long-term memory works by associating with something you already know.

The ultimate examples of games that make sense to you are of course your own games, since half the moves came directly from your own chess understanding. I wonder if solitaire chess/"test your chess" is also a good method to remember games, since it simulates tournament play.

By the way, memorizing many top-level games is a time-honored study technique in the game of Go (in which pattern recognition is perhaps even more important than in chess), used by all the top schools.  The idea as I understand it is to pick up a feel for good-looking sorts of move by osmosis.  

Just a thought...

Does "memorizing" include "understanding" as well, or is it just knowing the moves?

I make an effort to memorize games that are particularly enjoyable or relevant to my opening repertoire.  I don't know if it's helped me become a better chess player, because I'm pretty lousy, but it has helped me enjoy and appreciate the game more.


Stigma wrote on 06/21/08 at 20:02:33:

Is there a secret method? Do you study in any particular way that (deliberately or coincidentally) makes the games stick in memory?

I have not come across a secret method, but I have found it helps if I view the game like a narrative.  I can usually remember the major plot elements, and then after giving it some thought I can usually fill in the details.

For example, Morphy's Opera Box game started out as a Philidor.  3...Bg4 stands out as a positional blunder, allowing White to quickly gang up on f7 with Qf3 and Bc4.  Then Morphy swung the Q over to b3 with a double attack.  However, instead of snatching a pawn on b7, he chose to develop Nc3 (someone described this move as one of an artist rather than a butcher).  Black tried to solve matters on the Qside with ...b5, which Morphy met with a sacrifice Nxb5, followed by 0-0-0, and then the famous Q sac and mate with B+R.

Seems like you can attach a narrative explanation to almost every move in a master game, especially the classics.  Every move *should* make some sort of sense.  Unfortunately, some moves don't make sense to me because I simply don't understand them -- others don't make sense because they just don't make sense (even Morphy nods).  But these moves tend to imprint themselves even better than the standard moves.

Once you've memorized a few games, you can carry them with you and replay them at anytime, which helps to reinforce their lastingness.  The mention of the Opera Box game in this thread gave me reason to think about it again, keeping it fresh.

By the way, I recall an interesting article on Silman's website where he talks with great enthusiasm about an ancounter he had with some other player where they discussed some games that they each memorized in common.  I can't find that article now (that site seems like a mess to me), but I did run across this:

http://www.jeremysilman.com/chess_mstr_gms/master_games_1.html

Stigma wrote on 06/22/08 at 10:08:55:

I very rarely spend several hours studying a single game, but maybe I would remember them better if I did... Do you feel it is worth the time investment? Actually I do remember games I have "studied" for several hours, namely those I have played myself! But after a month or so I start forgetting even those.

Yes it has been worth it for me. I have heard some masters save anything 'below' 30 minutes is a waste of time and i've heard others say anything below an hour. I say you need to understand how you learn to decide how much time each game should really take (obviously there are other dependent factors such as length of the game..)


I think the goal is understanding but I am not arrogant enough to think I see every subtly in my studies and thus I'd say the end result is definitely more memory than sadly understanding though certainly I have learned many things from the game just not 'all.'

Fascinating subject.
Smyslov_Fan — Yes, memory and chess is very interesting and something I am having to do lots of thinking on, from an academic pedagogical perspective. This comes from trying to find ways to best teach/improve my 8 y o son's chess games. Off-topic but the ideas of Richard James (UK) at http://www.chesskids.com/art04.shtml are very interesting (ie on children, chess and understanding). And thanks for the references.

Stigma — I think you are right. I wasn't sure and therefore didn't post it but I think the game is Dvoretsky vs Timoshenko (1966?) USSR Ch. And of course memory works also with understanding and existing knowledge (it is a loop!) see below.

trw and matemax — I think memorisation is of at least two types (maybe more according to Smyslov_Fan's post and the people mentioned. 1) rote memory of very basic 1st order stuff/facts. Eg the times table or in chess, the initial moves of chess openings. ( :( I am afraid this is one area where my memory really fails me since I have trouble even remembering the Phillidor after reading about it. I have to sometimes rely on my son's memory!! :( )
2) memory by understanding. Here I think there are various ways to achieve memorisation. Some people are visual, some are eidetic, and/or some are "doers" (ie remember by physical action, eg playing in actual games).
A) I also think losing or winning games can assist in memory. Losing helps to put a "narrative" and this assists in associative memory. Similarly winnig after a protracted struggle etc.
B) Memory by doing is obviously crucial in blitz games where your mind is sort of on automatic. (This maybe akin to Michel Polanyi's tacit memory (whose ideas have abeen adopted (corrupted?) by all those knowledge management people).

There is of course the Susan Polgar's "thesis" that she has memorised at least 10,000(?) positions if not more which assists her in her games. (This is where the topic segues into my interest in children and chess. Do children below 11-12 learn by brute memory and mimicry with only a very rudimentary logical analytical thinking? Of course children who are "gifted" are a special case or more like on an extreme end of a spectrum where they may have been "blessed" with advanced logical analytical thinking ability a la Carlsen et al.)

Lastly (in another long post :)): Anand (according to commentators!) recently failed to remember corredtly a transposition and lost a game to Ivanchuk(?). Re remembering old master games, I think there was also a recent example where the GMs were replaying a game from the 1960s without realising it.

Having said that, I noticed that the recent "junior" GMs/WGMs are not really studying master games from the 40s-60s. An example is Hou Yi-fan — check out the interview on Chesslife (USCF website) in a report by IGM Rogers.

I am jealous. I hardly can memorize phonenumbers and birthdates and always forget where I have left my keys/glasses unless they are on permanent/fixed (what is the correct word here?) place. Twenty years ago I had no problems reproducing my own games, but now I forget them within a few minutes after their conclusion.
At the other hand I am pretty good at memorizing algorhythms. That's very handy when teaching maths and physics. But 100 chessgames? No way.

GabrielGale wrote on 06/22/08 at 23:22:23:

Lastly (in another long post :)): Anand (according to commentators!) recently failed to remember corredtly a transposition and lost a game to Ivanchuk(?). Re remembering old master games, I think there was also a recent example where the GMs were replaying a game from the 1960s without realising it.

http://www.chessgames.com/perl/chessgame?gid=1493939&kpage=1#reply20
Ivanchuk vs Anand.

Karpov did the same thing in
LarryC vs Karpov
http://www.chessgames.com/perl/chessgame?gid=1069116

Top GM to Top GM...

"How many games have you got in your database?"
"Thirty thousand"
"Thirty thousand? I've got over 5 million in mine, how can you work with thirty thousand?"
"Err, well that's all I can remember"

JEH,

I like that!

I was looking through a combinations book today and recognised several classic positions.  I don't remember all the moves to Euwe-Keres, but I do remember ...c6!!

There's another thing about memory that isn't discussed much among chess players.  Sometimes it takes quite a bit of work to remember the right way to play something.  For instance, I had the opportunity to play as Korchnoi did against Tatai in an exchange French today, but forgot about that game (and one I had won previously using the same system) until just after I had made the move.  

Memory is such a complex beast.

Eight years ago I played Kramnik two days after he won the world championship from Kasparov (it was a stunt for the paper I work for). We played quickly without a clock and he wouldn't let me record the moves. I won a pawn in the late opening/early middlegame but got outplayed, especially in the endgame, and lost. He then started to write the moves down for me (without using a board) but got lost in the middlegame although he was able to reconstruct a position later in the game and write the concluding moves.
My point is that here was the world champion who couldn't remember the moves of a game he had completed just seconds before.
I can't prove this, but I suspect the section he couldn't remember was down to it becoming random through me playing some very illogical moves.

Hello:

This is all very interesting. I thought one recommendation was to go through 1000s of games and learn by osmosis? But this thread is talking about spending 30+ minutes per game. Then you cannot go through 1000s of games.

I am interested in giving this a try though as I am cannot really remember games well.

trw, can you tell us how you spend the 30+ mins on a game (set up on board? Solitaire style? Lots of calculations of variations)??????

Thanks for any tips,

Gerry

trw wrote on 06/22/08 at 17:16:03:

Stigma wrote on 06/22/08 at 10:08:55: I very rarely spend several hours studying a single game, but maybe I would remember them better if I did... Do you feel it is worth the time investment? Actually I do remember games I have "studied" for several hours, namely those I have played myself! But after a month or so I start forgetting even those. Yes it has been worth it for me. I have heard some masters save anything 'below' 30 minutes is a waste of time and i've heard others say anything below an hour. I say you need to understand how you learn to decide how much time each game should really take (obviously there are other dependent factors such as length of the game..) I think the goal is understanding but I am not arrogant enough to think I see every subtly in my studies and thus I'd say the end result is definitely more memory than sadly understanding though certainly I have learned many things from the game just not 'all.'

Matemax wrote on 06/22/08 at 15:03:51:

Just a thought... Does "memorizing" include "understanding" as well, or is it just knowing the moves?

In the Go model, it is specifically just knowing the moves.  Understanding is supposed to come subconsciously.

Hello chess people,

Im not sure how this can help but Ive put together a free site that helps with mnemonics and memory. Let me know by using the contact sheet if it does - Id be very interested and pleased- pelase spread the word if it could be useful- the things you guys talk of reminds me of learning pi to x decimal places which is easily possible using mnemonic techniques (see video on site) ... JogLab.com

Just reading through this thread has made me realize how many potentially fruitful training techniques I have yet to try, especially focused on memory. As a player who has been stuck on the same level for 5 years now, this gives me hope. I've also woken up to the idea that my optimal learning style may be more linked to kinetics and "action" than I have assumed. The chess books that have given me the most are those I have sat down with and played through all the moves on a physical board, while trying to understand. But this preference may actually be a liability in the computer age...

@trw
The technique you mention of using music to create a kind of situation-specific learning is very interesting, and I have yet to try it out properly. Thanks for reminding me of it!

@dsanchez
The idea of creating a narrative out of every game appeals to me, I will certainly try it. I think one of my favorite opening books of all time, Pirc Alert, did this to some extent (with a lot of odd, creative names for various strategies that made them stick in memory).

@GabrielGale
Lots of interesting points. I think the rote learning you mention suits itself to the kind of flashcard/repetition system that is developed in the program "Chess Position Trainer". I have been meaning to make use of this both for opening and endgame theory for a while, but actually preparing the files with exactly the information I want to memorize is such a drag...

@Jo
I have some books on mnemonics and have often wondered about how they could be useful in chess. But mnenomics give an artificial structure to a mass of information that appears overwhelming and/or unstructured (like decimal places of pi), while I'm not sure chess fits this description.

Chess moves, plans and games ARE already meaningful and structured by virtue of the link to both visualization/calculation and narrative. I find it hard to escape the conclusion that the most effective memory system for chess must make extensive use of this already available structure. Then what place is left for mnemonics? I'm not ruling it out here, in fact I secretely hope mnemonics can still be of some use!

If it's any help, I haven't memorised any master games! On the other hand, I like to think that I would understand them! ;)

GMTonyKosten wrote on 06/24/08 at 11:12:26:

If it's any help, I haven't memorised any master games! On the other hand, I like to think that I would understand them! ;)

Hopefully you have memorized some of your own (master) games  :o

Matemax wrote on 06/24/08 at 13:10:57:

GMTonyKosten wrote on 06/24/08 at 11:12:26: If it's any help, I haven't memorised any master games! On the other hand, I like to think that I would understand them! ;) Hopefully you have memorized some of your own (master) games  :o

I am very sure Tony can reproduce his spectecualr game against Speelman  ;D
http://www.chessgames.com/perl/chessgame?gid=1276741

J/k of course ;)

Very interesting discussion here. Unfortunately it seems to be primarily theoretical as nobody seems to have much experience with intentionally memorizing a number of master games as a training tool. So, I'll gladly be the guinea pig and see how this works. BTW, AJ Goldby is the author of the site where I originally read this, so thank you to whoever pointed that out earlier. I've started by (re-) memorizing the opera house game. I know, only 17 moves. Too easy, right? This was actually a pretty ideal game as it demonstrates the problem with 3. ... Bg4 in the Phillidor and also highlights the balance between time, development, quality of position and material. I've read through a couple different annotations of the game so I feel like I understand it reasonably well.

So here's my plan. And I'm open and eager for any suggestions you all might have. I'm going to choose 25 games total to memorize and I'd really like to pick games that are models for specific openings, demonstrating typical plans or middle game motifs for each opening. I'm not too picky and play definitely doesn't have to be perfect for both sides. So here are the lines that I'd love to hear any recommendations for model games:

The White Side of:
1. CK advance or fantasy variation, especially the fantasy
2. French classical with 4. e5, f4, Nf3
3. Alekhine exchange variation (developing queenside pieces first, Be3, Nc3, Rc1)
4. Qxd5 scandinavian (Nf3, d4, c4 plan for white)

The Black Side of:
1. Sicilian closed, alapin with Nf6, grand prix attack
2. KID c6 and a6 vs saemisch, Na6 classical, averbakh

Anybody have any games that they love and especially that are educational?