Saturday, October 18, 2008

Chess for Zebras, Chapter 5 (Building your chess muscles)

A summary of Chapter 5 of Rowson's book Chess for Zebras, which is titled 'Concentrate! Concentrate? Concentrate.' In this Chapter, he picks up on a point he made in Chapter 1, essentially expanding it into a general training method.

Recall that in Chapter 1, Rowson argued that the way to improve is not to gain more cognitive understanding or knowledge of the game of chess (e.g., knowledge that you should get your rooks onto open files), but to develop skills at chess. In that chapter, one of his suggestions for building chess skill was to take complex problems and spend about 20 minutes solving them. The present chapter unpacks this suggestion, both justifying it and describing how to do it.

In this chapter, he opens by reprising a theme from Chapter 1, saying, "I believe that the surest way to improve our results is to practice solving complex problems at home, and to do this as often as possible, in a timed and disciplined manner. This helps to make concentrating more habitual and therefore makes it easier for us to focus at critical points in our own games." Rowson feels very strongly about this technique, saying "One exercise completed with full application for 20 minutes is easily worth an hour of playing through games."

Rowson's suggestion is simple. First, find a good position. This can be done many ways, but the key is that it is a position you find somewhat confusing, that you feel you don't understand. It shouldn't be a simple tactics puzzle, for instance. You can find positions online, in master games, or puzzle books, but ideally the positions should come from your own games. Save them on your computer for later use (I save tough positions when annotating my own games, squirreling them away for later Rowsonalysis).

Second, get to work! Set 20 minutes on your timer and hack away. The focus should be on analyzing the position, working out potential solutions to the problems it poses, thinking through any sharp lines to quiescence (and one move more!). You want to read the position rather than write your biases and preconceptions into the position with counterproductive narratives.

Rowson's method has many positive consequences. For one, it makes you less likely to fall into the trap of being a 'lazy detective.' You see a couple of clues, they suggest one suspect, and instead of pursuing the evidence and alternative possibilities, you just go along with the story you wove from limited evidence. Rowson's method doesn't leave much room for this. He says, "It gets you away from all the storytelling stuff and makes you more focused on working things out over the board. It also makes you more attuned to your opponent's possibilities, because in the process of training you realize that to be accurate you must constantly anticipate your opponent's ideas."

When you are really concentrating on a position, you are totally immersed, focused on exploring the game tree, testing out ideas and plans, hacking away at the position. When you are in this focused state, it is what the psychologists would describe as 'Flow.' In the state of Flow you are immersed in the task, not thinking about irrelevant details (I'm hungry), or even about thinking itself. You are thinking about chess, the concrete position in front of you and the possibilities therein. Similarly, when reading you are not thinking about reading, you are just reading.

Rowson also discusses the importance of setting up conditions before a game so you are more likely to experience Flow. If you are hungry, tired, or distracted by general emotional worries, it is much harder.

Rowsonalysis isn't particularly unique. Something similar was suggested by Kotov as I summarized here, and sometimes goes by the name Stoyko Exercises. I did it before my tournament last week for four days, and it clearly helped a lot, especially when I was mentally fresh in my first game. One consensus many people seem to be reaching, though, is that blindfold chess is not a great way to work on chess concentration. You don't play blindfold in tournaments. Practice how you play!

That finishes the summary of Part I of Rowson's book. So far, I am quite impressed. The first five chapters form a helpful coherent story that I can summarize as follows:
To improve at chess you don't need more book knowledge. Such knowledge often interferes with your performance by blinding you to the position in front of you. What you need is skill, and one way to improve your chess skills is to analyze complicated positions with intense focus for approximately 20 minutes.


Anonymous Anonymous said...

Intresting, intresting, intresting!

So analyse your own games in this way (stop at complex situations and set them aside for later thinking about that position for 20 minutes) will help you improve and at the same time work to find the truth in the game you have played. It may cost you a few days instead of a few hours but after awhile your analytical 20 minutes will become better and better and it will even become handy when analysis in the game your are playing.

That's it. I gonna order the book now. Cant postpone it any longer.
Thank you again for a wonderfull, good, summary. On to part two!

Btw, i wonder how long you are doing this Rowson suggestion and if you are already feeling that you analyse better because you are more focussed at the task at hand?

10/19/2008 06:17:00 PM  
Blogger BlunderProne said...

Its difficult to find a balance.

Long ago, I would study openings in hopes to improve my game. Then endgames, then the circles, now I am doing master games ( old timey).

I collect my games I play and review them with fritz etc. But I need to be more disciplined in my efforts to study MY OWN games. I need to put the time in and do the exercise of Rowanalysis.

But here's the rub, not one method is the correct course. Without opening knowledge, you are at some disadvantage on the first few moves. Without Tactics, you miss the one move wonders and more. Without the master game studies you can't see it all come together. Finally, without your own self analysis of your own slow games, you can't see where you go wrong in your processing.

Finding a way to combine all these methods in a healthy dose of each is the trick. Sure, at some point 90% tractics is the best bang for the buck, but if your fuzzy on positional ideas and don't know where your game is going then Studying master games is a good remedy.

Add to it the balance of life activities... then it gets mroe complicated. Yet, I still sign up for more events ( like next weekend, a one day event, the greater boston open).

I sure hope to have some time to go over games, tactics, my wall of shame, and now review of my most critical mistakes in my own games in positions where the outcome is unclear.

10/19/2008 06:27:00 PM  
Blogger Blue Devil Knight said...

Chesstiger: Thanks for the generous comment. I haven't done this since before the tournament last week. I've been really tired, and focused most on analyzing my games (but for each game I did save a tough position to analyze later--the one drawback is that I analyzed the whole game, so already know the best move in the position--I need to save them long enough to forget my analysis. I think in the future I should do the analysis until I reach such a position, and then do the Rowsonalysis then, and then finish the rest of my analysis).

BP: And how! There are so many things to do to improve at chess, and even if I didn't have a life outside of chess I wouldn't be able to do it all.

Based on just doing it four days (with a real board), I am convinced this method works. In the next chapter, he has even more useful information about the nature of the analysis, literally the kinds of images that are in people's heads as they improve. It is very cool, as it ties up some of the loose ends I had in my mind. Really, I should have summarized 5 and 6 together, but I didn't realize at the time how closely they meshed with one another. Hence, I'll try to get the CHapter 6 summary out fairly soon.

10/19/2008 06:50:00 PM  
Blogger Blue Devil Knight said...

BP: one more point. I know of coaches who swear by working through master games up through the main periods (from romantic, to the modern "scientific" play, to hypermodern, to whatever it is we have now, probably 'eclectic').

The key is, if it feels like torture, it is not a good method for you right now. :)

10/19/2008 06:51:00 PM  
Blogger BlunderProne said...

I agree. It has to be fun. I enjoy "listening" the chess classics right now.

It's a huge lesson to watch the evolution of the game. I am getting to the point where I can udnerstand how a new opening developed over time.

Prior to embarking on this journey of a historic study of master games, I was hacking my way at understanding positions and openings. Taking "Every Man" series of openings for religion and not really understanding where I was going. Now, I feel like I have a little more of a clue ...and its fun.

After HAstings 1895, I am heading to New York 1924 and then where I really wanted to be was Zurich 1953. I got that manual a couple years ago and Bronstein was excellent but it all went over my head. London 1851 was a nice start because the games were relatively easy to comprehend. Hastings takes a more positional feel. Now that I am almost closing this event, I feel prepared for the next class in Hypermodern History 301.

BTW. I am converting my posts to a chess base file complete with links to the annotated games in the data base along with the narration and pictures etc. I am thinking of "packaging" these up in CDs and Pimp them at events. You think I could pull that off? If so, what would you be willing to pay for something self contained?

10/19/2008 08:22:00 PM  
Blogger Blue Devil Knight said...

BP: That's a great idea! Depending on cost, I think it would sell, at tournaments and even more online.

10/19/2008 10:13:00 PM  
Blogger Unknown said...

You have a nice blog here! I offering you a link exchange...
My blog:


10/28/2008 10:26:00 AM  
Blogger likesforests said...

The tough question:

Should we spend 20 minutes/day on Stoyko exercises or is that time better spent another way?

Coaches and the medical profession seem at odds on this issue; there is the argument that new patterns have a bigger effect on depth of calculation than sheer speed, and that we already get practice with calculation during games and while studying new material.

11/02/2008 03:48:00 AM  
Blogger Blue Devil Knight said...

LF: Rowson seems to think it is one of the best ways to improve, but of course that is open to debate! He speaks fairly highly of the Circles, too, for people with tactical weaknesses.

I think it is probably best to use examples either from your own games, or from other games in your opening, to maximize the likelihood it will pay off (similar tactical and positional themes will recur).

Of course, the way in which the analysis is done is important. He didn't say too much about that here (other than, let the position dictate your thinking during the analysis, rather than vice-versa).

11/02/2008 01:02:00 PM  
Blogger likesforests said...

"LF: Rowson seems to think it is one of the best ways to improve, but of course that is open to debate."

I wouldn't debate that conclusion. I would debate at what *level* setting up rich middlegame positions and analyzing them for 20-120 min is an optimal or even very good way to improve.

My hypothesis is that maybe it's best for a player who already knows many tactical, strategic, opening, and endgame patterns by heart--which reviewers of the book have alternately pegged as 1800, (2) 2100, or 2300 elo.

You're of course aware of Training in Chess: A Scientific Approach which argues that reducing the breadth (branching) of our search through pattern acquisition has much more effect than increasing the speed of our searches. And that more advanced players often tend to calculate fewer variations than less advanced players.

"I think it is probably best to use examples either from your own games, or from other games in your opening, to maximize the likelihood it will pay off (similar tactical and positional themes will recur)."

Excellent idea. So, as I study the positions in My System or play solitaire chess with a game that illustrates key opening / endgame / middlegame themes I'm acquiring new patterns and simultaneously getting calculation / visualization practice.

That sounds much more useful for a sub-2000 player than finding rich middlegame positions and spending 20-120 minutes unwinding them. I guess the morale then for us sub-2000s is to try to always have *some* tough chess material where the positions take at least 5 minutes (as in real games) to figure out and we have good explanations to turn to when we're down trying it ourselves.

11/02/2008 02:38:00 PM  
Blogger Blue Devil Knight said...

LF: I think his whole approach would work best for someone that has a lot of knowledge (e.g., put rooks on open files) and has basic tactics in real games down pat. For Rowson what is needed isn't more book knowledge or narrative explanations of a position, but the ability to work things out over the board by thinking about concrete variations.

However, I think you are right that this may not be the best way to train for someone below a certain level. Certainly it's a good way to practice thought process away from a rated game, and most games at lower level are lost either by basic tactical errors or sloppy thinking (usually both). His method will work for the latter, and I am frankly not as sure about the former. The circles are pretty good with the former, but as I work through Chessimo I am struck by how unnatural many of the tactics are, how they seem almost like compositions (even though I think none of them are actual compositions). The best way to learn tactics would be to automate the acquisition of tactical errors you make from your actual games and hack at them circles-style.

This could be done fairly quickly by just playing a whole bunch of blitz games, where there will surely be missed tactics galore, and save the positions when the evaluation function changes.

Of course, who has the time for that?

11/02/2008 02:53:00 PM  

Post a Comment

<< Home