Thursday, July 24, 2008

Improve Your Chess III: Simulation, not Computerisation

Believe it or not, we play real chess using one of these:
On top of which we place some of these:
To the side of which ticks one of these:
Yet there are those instead who believe real chess is located here:

Who are they?

They are to be found chattering away on the ICC in the early hours. Analyzing graphs that show their tactical ability increasing. Arguing in some forum against century-established theory because of what a bookless Fritz 6.1 just said. Kibitzing how World Champions are patzers for having missed Crafty's anti-positional win of the exchange. Occasionally they scalp a titled player at 1+0, and into their finger notes it goes with the rest.

Some, even, when they have decided they are ready, dare to venture out to an actual, physical, real chess tournament or club, confident that their estimated grade (a conversion of their online rating) is accurate. And suddenly it's not a computer screen but a living, breathing human being sat opposite them - fists clenched as if focused on destroying them - and the eyes of spectators are watching here - there - everywhere - and at the board there's no take-backs allowed - no pre-moves - and thousand and one clocks are going tickity tock - and everything looks wrong in three dimensions - and, then, somehow, the game is lost. And there's no play again button. And no comforting suspicion that the opponent plugged in a tablebase in the endgame. No reason to suppose the opponent is secretly Fischer. No chance for immediate revenge, just the bitter taste of a definite defeat. After which, the next round and more of the same . . . Back they retreat to their computer nightly, for months, only venturing out again when they have forgotten what happened before, telling themselves instead it was bad luck anyway the previous time. And after all, that graph of their tactical ability has gone up since then . . .

Am I being cruel? I'm certainly exaggerating, I think. But people do get very much over-invested in what they think they are proving about themselves or chess using a computer, an investment that is not useful for over-the-board (OTB) play, which is the one true test of chess. Here for instance is an example of emotional over-investment in on-line chess. And anyway, I was one of these people. Well, maybe not quite. But I certainly spent too many nights on the ICC, trying to get back over 2300 . . . trying to scalp the IM who keep beating me . . . waiting for my chance to have a crack at someone in a simul . . . that kind of thing, and I have also spent a fair share of my time on sites such as Chess Tactics Server which I am no longer convinced was an optimum use of my time.

At some point I decided to more or less give that stuff up (albeit not entirely successfully) and try something else. I dusted off my board, poured the pieces out from their musty box, got out my clock out, and decided instead to use my chess time at home to, as far as possible, simulate over-the-board (otb) chess. I will give examples of what I mean by this in a moment, but first the reason.

Or rather, reasons. Firstly, I became convinced that I remember things best when they are attached to experiences that have meaningful emotional content. Otb chess is inevitably like this, for me; solving a puzzle in a book on a bus is not. This is the most important reason. Secondly, I think various forms of computer chess skew your game in some not very useful ways. An obvious example is how after hours of blitz we think far more tactically but at the expense of our positional thinking. Thirdly, I think internet chess deceives us into thinking we are better or worse than we really are. Ok, you were feeling great and beat an IM two games in a row. How do you know he wasn't drunk, or that his son hadn't logged on to his account? Or the other way around. So you dropped a hundred and fifty points over a few hours; does that mean your otb form will be terrible, that you should lose confidence, be cautious, start offering as white early draws against weaker players? Maybe you were just tired. Or maybe it really was Fischer's ghost moving the pieces that time. Who knows?

So, I started simulating otb chess at home. I did this in three main ways. Firstly, I would play practice games against computers under the following conditions. I set up the pieces and clock as if a normal game. I turned the computer screen away from me, and set the programme to speak the moves as it made them. I played only one game per session, just like in league chess. If I blundered in the first five minutes, that was it for the night. If I lost on time, I lost on time. Secondly, I set up practice positions that other people recommended as appropriate for simulating otb chess. Once again I set the clock, usually to around 15 minutes, and tried to make a decision as if the position was otb. Below are six sample positions, if you want to take my advice and try this at home. (Remember, this involves setting up the pieces, setting a clock for up to 15 minutes, and trying to make a decision about which move to play as if you were playing a real otb game.) In these first four positions it is black to move:

Whilst in these final two it is white to move:

Thirdly, I tried reading chess books in this way too. That is, not by following the games on a computer or on a travel set, but by going through the games on a proper set, playing out the annotations or visualizing each one, and so forth. Afterwards I would then put the book aside and replay the games, trying to remind myself of the annotations and trying to really "get" what was going on in the games. All of this I found really quite hard to do, especially so when using opening books, which I find extremely hard to actually learn anything from.

Before I summarize today's post, I would like to just reiterate who my intended audience is with these posts: adult players who've hit an impasse. For the most part we are different from children who are growing up in a world where computers are just a straightforward part of chess without any complications. For such lucky children, their computer resources don't seem new and different and such strange temptations, but just a part of chess culture (as say New In Chess does for me.) This means they don't fall into the various traps that coming late to the world of computerized chess risks, as I did.

So, my conclusion. We learn best about chess over-the-board, and over-the-board is the real test of any player. However, various forms of computer chess are so artificial they frequently teach us nothing and merely deceive us, or at the best drastically skew any learning that does take place. Therefore, we should only use a computer in so far as it helps us simulate otb chess, and simulating otb chess at home is a key principle with which to engage in any training activity. Simulation, in other words, not computerisation.

PS. This post is part of a series started here. Number IV is scheduled for Monday next week, and I intend to follow-up with four or five posts on Mondays after that one before closing the series out.


Naisortep said...

Interesting post. What program do you use to announce the moves? Does Fritz have this feature? I assume you enter your moves in algebraic without looking at the screen?

Tom Chivers said...

Fritz has a "speak move" feature so I use this. I haven't found a way to type in my move (which would be ideal.) Instead I make the move on the board on the screen - but with the evaluation window closed and so forth, so there are no cues or clues as to what the computer thinks. I also make the move finally - no changing my mind once the clock is pressed, even if seeing the position again on the computer screen makes me think differently etc.

ejh said...

Fritz has a "speak move" feature so I use this.

Christ. Does it have a "going outside for a fag" function too?

ejh said...

I'm strongly of the view that online blitz does basically nothing to improve your OTB chess and may even hurt it. I know that strong players spend a lot of time on ICC, but I think that's a question of keeping in tactical trim - much in the way that a professional darts or snooker player must practice every day or lose their edge. I don't think it improves one's chess at all and it risks developing all sorts of bad habits - playing for tricks, losing without a fight because you can always play another game, dicky openings - that will hurt you in proper chess.

As it happens, I gave a lecture at the club a few years ago about my experiences in correspondence chess, which is essentially the opposite of online blitz - but I came to very much the same conclusion. It probably wouldn't help your OTB game at all. Which is OK, that's not why you're doing it, but it's worth saying all the same.

Improvement is really about work - well-organised and well-directed work rather than working till you drop - and about thinking about what you're doing. It's also about not forgetting what you've learned, which in a variety of ways is a great deal more difficult than one might think.

Anonymous said...

"I set up practice positions that other people recommended as appropriate for simulating otb chess."

Rowson recommends pretty much the same thing in his "Chess for Zebras". The trouble is, my positional understanding is not good enough to select such positions by myself.

Do you have any pointers to good online (oh, the irony!) resources with more positions like this?


Tom Chivers said...

Hi grrrr. The ones I gave are from the forum - we had a small group of people providing them for one another. Unfortunately I don't have a reliable source that I can point you toward for more, and the chessworld group hasn't continuted.

Why not try setting up a group of contacts via email interested in this, who would then swap positions for simulation amongst each one another, and after each simulation offer each other coaching? (Ie, the person volunteering the position would have read up on it, or analyzed it comprehensively, that kind of thing, and the person responsible for volunteering the position would rotate amongst members of the group.) I would be interested in that if someone else had the time to organize it . . .

I read the Rowson books around the time I started to do this - I no doubt picked up the idea from them. In fact my post on Monday will be about reading the Rowson books. It's a shame the Rowson books don't have a companion "Resource Book" for this kind of thing.

Tom Chivers said...

Christ. Does it have a "going outside for a fag" function too?

That as well as kicking you under the table and dropping smelly food on the board...

Mark Leonard said...

If I ponder one of the test positions for 15 minutes and select a move, how do I subsequently verify my analysis? Are there solutions (best moves)? It seems self-defeating to check with a computer...

Tom Chivers said...

Hi weakspeaker. You should be able to locate the positions on the forum. Or I could probably locate the "solutions" (for want of a better word) for you.

It's interesting I've had responses from two people already interested in doing this. I really think if someone set up an email group for this kind of thing it would take off . . . If someone else was willing to run it, I would certainly advertise the fact via the blog.

Glenn Wilson said...

What? You use a clock and still call it "real chess?" "Real chess" requires well developed Sitzfleisch.

I find that all forms of chess that I have played seriously have helped me at all other forms of chess. Correspondence, OTB slow, OTB blitz, online blitz etc all contribute to my understanding of the game and skill at the game in different ways.

But I do agree that if your interest is in optimizing your performance under certain conditions (e.g. "League Play") it makes sense to simulate similar conditions as part of your training.

At the same time, for players that frequently miss simple tactics, I believe that various forms of training at simple tactics is the biggest payoff in time invested.

Tom Chivers said...

I thought about mentioning books on simple tactics on the post, saying there are two kinds one of which is preferable to the other. One uses words to describe the game position given, so involves the reader and thus comes closer to the ideal of simulation. The other just gives positions blind and so is less good, closer in kind to websites like chess tactics server and so forth.

Anonymous said...

d5 (OK I've seen this one before)


Tom Chivers said...

Analysis of all the positions can be found here. With the first and third positions there is some ambiguity as to the right move. However, PG, you only found one of the intended moves (1.d5 in the final position.)

Did you set up a board and pieces and clock all that? The response to this thread so far has made me think I might have been wrong about chess players always refusing advice!

Anonymous said...

0/5! No I didn't use board and pieces, just analysis in my head and intuition. And you wonder why my grade keeps going down?

Truth be told, improving my chess is not my top priority at the moment. Never has been actually.

If I'd set up the board I'd have got more correct but enjoyed it less. This is why I'm not an IM.


Anonymous said...

I recognise position 2 - it's in
'Imagination in Chess' by Paata Gaprindashvili (published in the UK by Batsford). I'd recommend this book as a source of positions for analysis - it has 750 of them (and comes with answers).


Tom Chivers said...

Btw, in the first white-to-move diagram lurks a particularly masterly, unusual and instructive continuation available to white. (When I did this, I got the right first move, but not intend the correct follow-up.)

As such I think it's worth posting the whole game here - for the solution go to move 25:

1.Nf3 Nf6 2.g3 d5 3.Bg2 Bf5 4.O-O c6 5.d3 e6 6.Nbd2 Na6 7.a3 Be7 8.b4 O-O 9.Bb2 h6 10.Re1 Nd7 11.e4 Bh7 12.c4 dxc4 13.Nxc4 c5 14.b5 Nc7 15.a4 Bf6 16.d4 cxd4 17.Bxd4 Bxd4 18.Qxd4 b6 19.Red1 Nc5 20.Qe3 Qe7 21.Nfe5 Rfd8 22.Nc6 Rxd1 23.Rxd1 Qf8 24.Qf4 Ne8 25.a5 bxa5 26.b6 axb6 27.Nxb6 g5 28.Qe5 Ra6 29.Qb8 Bg6 30.Rd8 f6 31.Bf1 Rxb6 32.Qxb6 Nxe4 33.Qxa5 N4d6 34.Qa7 Bf7 35.Rb8 e5 36.Qd7 Kg7 37.Ne7 Qh8 38.Bd3 h5 39.Nf5+ Nxf5 40.Bxf5 Kf8 41.Rb7 Qg7 42.Qe7+ Kg8 43.Be6, 1-0

In 'Complete Chess Strategy' vol 3 ('Play on the Wings') Pachmann I'm told writes of the move in the diagram: "Black's pieces are now thoroughly disorganised allowing White to break through on the Q-side with a neat series of tactics which illustrates convincingly the effectiveness of his own pieces. The black rook is to be first victim." Reshevsky was white btw.

Phaedrus said...

Hi Tom,

Much food for thought here, and over all a brilliant series of posts. Please allow me to suggest a very interesting source for positions for this type of exercise. It is the book "The best move" written by Hort and Jansa.

The book has 230 positions all taken from the games of these two grandmasters. Some tactical, some positional, but not as clear cut as most positions in exercise books.

I am looking forward to the follow up on these posts. Keep up the good work.

Tom Chivers said...

Thanks Phaedrus for your nice comment and the book recommendation. It's out-of-print atm here and as such rather out of my price range unfortunately...

Anonymous said...

Recently spotted on playing 1 minute games..against Tom Chivers haha

Tom Chivers said...

Do as I say, not as I do!

Bungerting Baloner said...

Another book is GM-RAM with many interesting positions.

I do not believe it at all harmful, and in fact extremely helpful, to do computer analysis POST doing the OTB exercise. Set up the position, set your clock, do your 20 or 30 minutes, and then go to the computer for in-depth review of your chosen move and ideas.

I also find Chess Tempo ( helpful as an alternative to CTS, which emphasizes speed far too heavily. CT has a "standard mode" where you take as long as you wish. The positions tend to be more difficult than at CTS. So overall there is a better correlation to slow-time-control OTB chess. Of course, it doesn't have the advantage of physical board and pieces, but I suppose you could instead use it as a source of interesting positions instead.

For opening study, which I too don't generally spend all that much time on given my low playing level, I find there is no comparison for the physical act of moving the pieces around, again and again, to really reinforce the moves and patterns. I start each variation from scratch, moving the pieces in the same patterns over and over. It makes a difference, I believe.

Anonymous said...

hello everyone, I am TheSalesman(FM) on the ICC. The training positions you have there if I recall right are positions from a book called Imagination in Chess by Paata Gaprindashvilli ( not sure if I spelled his name right). This book was highly recommend by Jacob Aagaard.

There is a book I want to purchase called "The Critical Moment" by Dorfman, which deals with "critical positions". They have rated it low, but these people... they don't appreciate these works.

As far as the suggestions made in the article, I agree with a lot of what is being said but it is important not to go overboard. for instance, if you spent 100 hours on the CTS I think the mental processes that you experience in decision making would be more beneficial than say, the amount of time you spend flipping the pages on a book, in addition to setting up the pieces on the board. I know there are more efficient ways of using the time, but we cannot put the computer away as a whole just because we were not born in the information age. I share your opinion with OTB simulation, but I think that no less than 1/4 of your training should be done with the computer. You are simply not using your tools to their full potential.

Thanks for the article, and I hope this was insightful.

Tom Chivers said...

Thanks Salesman for your advice and for the book recommendation. No, the positions I gave were not sourced from the book you mention, although someone else in the comments above mentionned position 2 is included there nonetheless. I suppose there is also a chance they are *all* included there, which would be some coincidence!

Tom Chivers said...

Btw, Dennis Monokroussos's link to this post on his blog *The Chess Mind* here also generated some interesting comments.

Anonymous said...

"Yet there are those instead who believe real chess is located here:

[Picture of computer with animal]

Who are they?"

People who own that sort of animal.
I have two of them and they climb up on the table and knock over the pieces every time I try to use a real board and pieces.
The can't knock over the pieces on "chessboard.exe".

ejh said...

But they can jump on the keyboard and have essentially the same effect.

Anonymous said...

IMHO it's very difficult for an adult chess player who has been playing seriously* for a number of years, and has seen his playing level plateau out for some time, to improve.

* by seriously I mean he's already played hundreds of rated games and has put a lot of time and effort into studying openings.

Sure, he might be able to improve his results by working harder on his endings, and making more effort to study openings that give him positions that suit his style (messy, positional, etc.), so that his rating may increase a little, but he won't be playing better chess.

Ali Ocken

ejh said...

I think that's quite right, and although I've added 100+ Elo points over the last few years I don't think I'm a better player than I was.

However, we can turn that on its head and say "it may be possible to achieve better results without actually having to become a better player"! Now isn't that a tempting offer?