Memorization in chess

There is a lot of chess improvement which amounts to memorization. Doing tactical problems, learning endings, openings, memorizing classical games. These all amount to learning multiplication by using flash cards. For instance, there are hundreds of computer training apps on the market that work in this way. Maybe it is not literally flash cards, but much of it is the same concept.

Two questions:

1. What level can this kind of "flash card" study achieve? It seems that if you were tactically very strong, knew hundreds of endgames, and knew opening theory deeply, you could reach a weak master level? Or no?

2. What are other ways of improving besides the flash card and memorization approach? In other words, once you plateau from this kind of study, what is next?

1. No. none

2. understand, practice, learn to put ideas and patterns to use in real games, build your abilities.

Pure memorization alone helps you nothing in all of your examples. Exception: You may get a good position after an opening line your opponent happens to play along with what you know. But after that, it comes down to what you can DO over the board.

The most improvement in the beginning is about tactics. But not from pure memorization of exact tactics. You will only improve if you use the tactics patterns that you are getting to know, learn to understand them and practice DOING IT until it helps you to handle your pieces better and better.

Then let's say I'm not asking about "pure memorization". Sure, if I can memorize the Phildor position and can recreate it, but cannot play it OTB then I have gained 0 ELO. But it's not hard to memorize a few dozen examples of the Phildor position. After that, I understand it, I can explain it to others, and I can play it OTB, so I have improved. You can do the same for hundreds of imortant key positions and concepts.

I suppose it would have been more accurate if I had left out the word "memorization" altogether. I am really asking more about a flash-card-based approach to learning, or put another way, a "self-quiz" approach to learning, where you learn a concept, create a flash card for yourself, then you quiz yourself periodically to reinforce the concept. This can be interactive, as it is with many computer-based chess training products. For example, an interactive self-quiz would not just show you a K+P vs K ending and let you think, "Oh yes, I know that, next..." No, it would force you to play it out against the computer several times to prove you can execute it.

The same is true of almost all tactical training. GM Ziatdinov says to do tactical problems: "...repeating this cycle until you can go through 1000 problems 'by hand' (not mind) without any mistakes. If you try this method with my tactics program and complete it, you will have the tactical ability of a Grandmaster."

When he says "by hand", he means that you recognize the pattern immediately. In other words, he says if you cannot figure out the tactic problem after a few seconds, you should look up the answer and move on. It is not important to spend 15 minutes looking at the board trying to figure it out yourself. He says it is only important to get more patterns into your brain.

So let's say I wrote a computer program to quiz me on 3000-4000 tactics problems, and I wrote a program to quiz me on 200-300 important endings (forcing me to play them all perfectly against the computer), and I wrote  a program to test that I can execute my opening repertoire perfectly. That is all very similar to a flash-card-based or self-quiz-based approach to learning. Surely it would be worth many ELO to a person who could score 100% on 3000+ tactical problems, and score 100% in 200+ important endings, and score 100% in his or her opening repertoire.

Memorization is very useful if it facilitates understanding (although I think your suggestion of the number of positions needed for master is too small). I wrote a piece about memorizing chess games a few years ago:

Fingerprints

I collect diagrams of chess positions. Once collected, I print these diagrams on cards for review. My cards of pawn endgame positions fromDvoretsky's Endgame Manual (2003) has been aiding my review of the instruction in that text, and has proven popular with some of my pupils. Sometimes in a chess lesson, I pull out the cards and fan them across the chessboard upside down. The student picks one, we set it up on the board, then he or she solves it. There are a few in that set that I do not yet play with full confidence. When I have mastered those, it will be time to create another set from Dvoretsky's book.

Meanwhile, I'm collecting middlegame positions. I have several sets of cards that I created years ago. The oldest are index cards upon which I stamped diagrams, and laboriously stamped each piece with red or blue ink on the appropriate square. When I look at these old cards, I am reminded of time I spent reviewing them between rounds at the Dave Collyer Memorial tournament the last time Gary Younker ran it. Gary died in 2001, and shortly after his death we created a foundation to honor his memory and continue his work. The 2001 Collyer was a good event for me. I started the event rated 1400 and had an even score against three B Class opponents. My run of success started late Saturday night when I discovered a practical chance in this hopeless position.

White to move


I'm down two pawns, and there's no stopping my opponent's d-pawn. In a final desperate ploy, I played 31.Rf1! Keith Brownlee had several ways to counter my threat, but instead played 31...d3?? I sacked a rook to force a draw by repetition. After the game, my opponent told me that he only examined my checkmate threats, of which there were none, but not my drawing combination. He also stated that this game was the first time he failed to win against the King's Gambit.

On Sunday morning I beat a B Class player in a game that summoned more tactical courage from me than was my custom. Flash cards contributed to my confidence. Within the next year, I bought some software that facilitated creating professional looking printable diagrams, and my index card collection went into storage. I collected dozens of positions from Lazlo Polgar's Chess in 5334 Positions (1994) and several databases. I printed these positions on cards with a diagram on one side and the best moves on the other.

My initial non-provisional USCF rating was in the low 1400s, but before it was published I played in an event that pushed it up to 1495. That was in 1996, but in 2000 I was back down to 1400. My success in the 2001 Collyer rocketed me up to 1450, and in 2002 I climbed over 1500. I faltered briefly in 2004, dropping to 1487 before rising to 1600 in 2005. I made it over 1700 for the second time in 2008, and kept climbing over 1800 in 2009. If I am to cross over 1900 in 2010, my training must step up a notch.


Ziyatdinov's Method

Rashid Ziyatdinov advocates learning entire games thoroughly. In GM-RAM: Essential Grandmaster Knowledge (2000), he lays out a plan for improvement based on 300 key positions. Half of these are endgame positions--most are pawn endgames and rook endgames--and the others stem from classic games. His fifty-nine games from which the middlegame positions arise span less than a century from a few 1851 victories of Adolph Anderssen to Mikhail Botvinnik's 1936 defeat of Saviely Tartakower.

I find myself drawn to certain aspects of Ziyatdinov's method. My cards from Dvoretsky's text lack the answers on the back, for example. I'm also working on memorizing games, including those in Ziyatdinov's fifty-nine. His most compelling idea is the notion that key diagrams function as fingerprints of whole games. Most collections of diagrams highlight tactical motifs. There are certainly quite a few tactical shots in Ziyatdinov's collection. But memorizing, studying, and knowing thoroughly a limited set of games--the plans that led to what happened over the board, and what might have happened--goes beyond tactical patterns. The 120 middlegame positions in GM-RAM "are like the fingerprint of the games--from this fingerprint, the associated game can be identified" (77).


Karpov's Best Games

Although I share with Ziyatdinov the conviction that nineteenth and early twentieth century games merit our attention, I am unwilling to limit my study to these old games. I may end up with more than the legendary 300 positions as I pursue Ziyatdinov's regimen (he expects the reader to supply nearly four dozen of the 300). As I am going through the best one hundred games of Anatoly Karpov that were published in Chess Informant (see "Coincidence?"), I am collecting diagrams. These diagrams are fingerprints for games worth knowing as thoroughly as Anderssen's "Evergreen Game".

Some of the positions from Karpov's games feature tactical shots. In this position from 1973, Karpov's tactical shot provoked Spassky's resignation.

White to move


The following year, in the ninth game of the World Championship Candidate's Match, another tactical shot by Karpov provoked another resignation by Spassky.

White to move


Then, in 1977 at Las Palmas, A. Martin Gonzalez perceived the futility of further resistance when Karpov's move threatened a clever mating net.

White to move


Such tactical shots are the bread and butter of chess training. But, it seems to me that if I can comprehend the thought processes that went into finding the move that Karpov played against Vlastimil Hort from this position in 1971, it might become part of the knowledge that can elevate me to expert class.

White to move

Hort played on for another eleven moves as Karpov increased the pressure. This diagram is the fingerprint of the earliest of Chess Informant's list of Karpov's 100 best. It is a positional masterpiece, Karpov's signature. As I collect these diagrams, I aim to learn the games from which they stem. 

That sounds much better already :-)

But it still seems to me that this would waste far too much time on the flash cart part, and fail to use much easier and more relevant ways of learning.

For a chess player, the best way to memorize things is direct chess experience.

(Anand himself used to say that in order to memorize all the loads of opening preparation he was doing for his world championship matches, he used to play training games --- with one of his seconds --- and the lines in which he lost one of these training games, he would automatically remember best.)

Say you want to remember the basic Philidor way of drawing a rook endgame (Philidor 1 according to Dvoretsky).

Then one inefficient way can be to memorize numerous examples for that exactly.

One very efficient way to learn and absorb it can be to understand the information that you just need to place your king in front of the pawn, put your rook on the 6th rank, and immediately when he moves his pawn to the sixth rank, your rook shifts to the first (or second) rank drawing by giving checks from behind.

After you get that information, it is best if you get to try it against someone who desperately wishes to win such a rook endgame against you and despairs against how easily you hold the draw with your method.

Then you  are very likely to be able to use that info about the Philidor draw, not only tomorrow, but for a very long time, probably forever, because you just can do it now, and you are very unlikely to ever "unlearn" how to do it.

Sure gathering a lot of information (ideas, patterns, motives, ...) can be very helpful and essential to your chess learning. I would say that you will tend to most efficiently absorb  if you do not get sidetracked by making flash cards, but rather use actual games and practice to burn in the info.

Extreme example: You may try to learn by heart many examples on how to mate with knight and bishop, and still fail under pressure. But if you get a hang on practising it a little by having understood some essential parts (I will drive the king to the border, likely he will flee towards the wrong corner, then I will go for Kf6, Nf7, Bf5 while his king is on f8 or g8, then my knight will jump along the W shape of f7, e5, d7, c5, b7 while the bishop does the job of controlling the required light squares and the king will march over f6-e6-d6-c6-b6, and the mate will be with his king on a8 and my knight on a6 and my bishop on the a8-h1 diagonal)

and start making a sport of it of telling your chess friends that you will able to mate them with bishop and knight in blitz with only a minute left on your chess clock (half a minute if you are a fast blitzer ;-) ))

then you will KNOW that you will be able to DO it by tested experience :-)
 

The point is that most often, real playing experience is usually much more efficent than memory cards for absorbing and mastering chess skills.

I'd dare say that doing thousands of "easy" (for you) tactics that you can all solve in the end (or any method that can possibly result in a 100% score) will become quite a bit boring with time, while not being the most efficient way of advancing. After practising and managing the more basic combinations, it may be much better to advance to new, more challenging tactical situations, where you will not score 100% easily, but where you will keep being challenged towards searching more for ideas and calculating deeper down the lines, trying different motives that you know, and different new ideas that come to your mind based on the skill and feeling you have developed by now. Being challenged by something new that you still need to figure out is much closer to the real game experience (and much more beneficial for improving your actual practical playing strength) you are going to have in real practice than being able to easily and safely reproduce what you have already learned in a "flash card" way.

I am working my way through memorising 10,000 combinations. It is partly an experiment. I want to see if it has an effect on my chess rating.

Perhaps it will, perhaps it won't. But the only way to find out is do it and see what happens.

Hi, Steven. What were your ratings like 3 months ago? Have you noticed a big improvement?

--------------------------------------------------------

I have also just started something similiar, I think this subject depends upon what kind of learner you are. Some people are better suited to studying by memorization (SRS, of course -- https://en.wikipedia.org/wiki/Spaced_repetition).

I am finding that memorizing hundreds of tactics and my favorite openings thoroughly, together with trying to improve my strategical thinking by analyzing my games and others, is really helping me "see" deeper while playing. Especially in OTB Blitz. 

I think that if the suit fits, wear it. Study chess in a way that makes you happy, otherwise it will become a laborous activity. 

Interesting comments about memorization in chess!

Chess is memory and only memory! Botvinnik and Fischer, among others, stated this many times. Kasparov played entire games in a few minutes because he had memorized openings all the way to the endgame. Magnus stated he has memorized 10,000 games and he proved it several times in the past. Alehkine also memorized thousands of games, as have all grandmasters.

Interesting comments about memorization in chess!

Chess is memory and only memory! Botvinnik and Fischer, among others, stated this many times. Kasparov played entire games in a few minutes because he had memorized openings all the way to the endgame. Magnus stated he has memorized 10,000 games and he proved it several times in the past. Alehkine also memorized thousands of games, as have all grandmasters.

1. I think the flashcard study by itself is not enough to reach any level in chess - I mean you're building a pattern bank, but you still need to learn how to use it in real games, which involves calculation, evaluation and various decision making skills. However, I believe it would still help a lot to make progress (especially for adult players). I see the pattern bank as a toolbox - the larger, the better - but you must still practice how to use them.

2. Practice (ie. playing OTB long games) and analyzing with stronger players.

Hi everyone!

I'm looking for a book which contains chess combinations (diagrams) stored in a box.

There are two kind of diagrams in the box:

Grey coloured - easy

yellow - hard.

I forgot the name and the author.

If somebody could help me, please send me a message to my count: illessandor

Thank you, good luck