A few years ago, Willy Hendriks (good name) released a chess book called 'Move First, Think Later' to wide controversy and critical acclaim. I wondered if you fellow lifein19x19ers had some opinions you can share about whether you feel the concepts apply to go?

For those who don't know, here is a brief summary (I don't have nearly as good a prose as Hendriks so I'm cutting some of the bits of his opening paragraphs instead):

Trainer: 'You've had the chance to have a look at the position. What's it about, what are the most important characteristics of this position? Paul, do you have an idea?'
Paul: 'Uh, yes. I would play Rc6 and if he takes it I will have Nd5'.
Trainer: 'Yes you came up with the moves right away. But let's go back to the characteristics of the position, can you say something about them?'
'Well, uh, Rc6 threatens to take on d6, ...and Nd5 is coming. What can he do then?'

Many chess books are written in the same pedantic tone the trainer is using here. They are based on the idea that you should not try out moves at random, but first take a good look at the characteristics of the position, try to make a more general plan on that basis and only then search for a concrete 'result' at the level of an actual move. This is nonsense.

No chess player thinks like this, no one has learned to play chess by thinking like this and even trainers and authors of chess books don't think like this.

(,,,Closely related to this is...)

In chess manuals you often find first a verbal piece of 'advice' of a general character, followed by a grandmaster game serving as an 'example'. The suggestion is that what the grandmaster does is not much different than following up on the advice just given. (...) And you can really believe in yourself finding this strong move in such an exemplary, rational way. So next you find yourself urging your pupils to look at the characteristics, fully expecting that the move 1 ... Be8 will follow out of them, just as it did in your case.

Humans are very talented at misguiding themselves and making conscious rationalisations for behaviour that originated on an unconscious level. ... I think this is an essential mistake that trainers easily make: confusing the ease with which you can afterwards talk about a concrete position (descriptive) with the possibility of making a set of written rules that can guide you in any position (prescriptive).

Hendriks's argument is that just as easily, effective moves point us towards the essence of positions rather than the other way round, that positional feeling is not prescribed, but built up from a diet of pattern recognition in games and exercises, and that ultimately, improvement is mostly brute force neural pattern-building, and therefore, hard work.

The basic idea really is that many of the conventional approaches to chess improvement simply are contrived and hard to implement into a clear and coherent method for OTB play.

_________________
12k: 2015.08.11; 11k: 2015.09.13; 10k: 2015.09.27; 9k: 2015.10.10; 8k: 2015.11.08; 7k: 2016.07.10 6k: 2016.07.24 5k: 2018.05.14 4k: 2018.09.03 3k: who knows?

Conclusion to be drawn from this: lots of tsumego and games.

_________________
Go Books

I began what turned out to be a longish thread on this book (viewtopic.php?f=10&t=6327&hilit=hendriks) and there have been discussions on several similar books.

As with the most honest (and I think best) recent go books, the overall message is that you have to do a lot of grunt work to build up your intuition/subconscious/System 2 or whatever you want to call it. Rationalism can trim the process but can only be the icing on the cake, and not the cake itself.

Did you have specific go books in mind?

I'm reminded of an acclaimed (chess) series by Artur Yusupov, 9 books in a structured course, where each book is divided into 24 chapters, each chapter is on a specific topic, e.g. 'Double Attack', and each chapter is in 2 parts: explaining what the theme is, by way of examples, and then a dozen problems for the student; the answers score various points, and when you get a pass mark, you move on to the next chapter. As the books progress, some topics are repeated, at a higher level of complexity.

Recently, this kind of learning, with the onus on the student, with the onus on hard work and problem-solving, and with the overall learning strategy being one of many positions, many problems (the kitchen-sink-type approach), seems very correct and 'modern'. I find it striking that some years ago, this approach was maligned under such terms as 'rote learning' and 'learning by repetition'. Now it seems that authors are going full-circle, that prescription learning and shortcuts are inadequate at building the networks the brain needs in order to implant good moves at an intuitive level in over-the-board situations.

_________________
12k: 2015.08.11; 11k: 2015.09.13; 10k: 2015.09.27; 9k: 2015.10.10; 8k: 2015.11.08; 7k: 2016.07.10 6k: 2016.07.24 5k: 2018.05.14 4k: 2018.09.03 3k: who knows?

And how does that approach differ from that of say, 1800 in chess or 1600 in go?



Well, sure. Both go and chess have advanced greatly since those days, and not from a kitchen sink type approach. What has happened recently is that the general principles and strategies developed in chess and in go (and in Japanese go in particular) since those times have been shown to be not as certain as was believed. That fact does not say anything in particular about how to improve at either game.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

Is it important to learn go concepts? I think so.

Click Here To Show Diagram Code

[go]$$B Snapback
$$ . . . . . . .
$$ . . . . . . .
$$ . . . X X . .
$$ . . X W W X .
$$ . . O 1 . X .
$$ . . . O X . .
$$ . . . . . . .
$$ . . . . . . .[/go]

captures the stones by a snapback. It is called a snapback because . . .

Click Here To Show Diagram Code

[go]$$B Snapback (II)
$$ . . . . . . .
$$ . . . . . . .
$$ . . . X X . .
$$ . . X O O X .
$$ . . O B 2 X .
$$ . . . O X . .
$$ . . . . . . .
$$ . . . . . . .[/go]

If captures . . .

Click Here To Show Diagram Code

[go]$$B Snapback (III)
$$ . . . . . . .
$$ . . . . . . .
$$ . . . X X . .
$$ . . X O O X .
$$ . . O 3 O X .
$$ . . . O X . .
$$ . . . . . . .
$$ . . . . . . .[/go]

can take back. The full snapback sequence is not played, as a rule. is kept in reserve as a ko threat.

Now, most beginners learn the snapback early in their go careers, whether they pick it up or figure it out or it is explained to them. This particular shape and sequence is not the only snapback. It may fairly be called a prototypical snapback, but it is not the only one.

Click Here To Show Diagram Code

[go]$$B Snapback threat
$$ . . . . . |
$$ . . . X X |
$$ . . X X O |
$$ . . X O O |
$$ . O O . . |
$$ . . . O 1 |
$$ . O O X . |
$$ . . X X . |
$$ . . . . . |[/go]

threatens a snapback.

Having seen the first snapback, does the beginner "recognize" this position as a snapback threat? Maybe so, maybe not. Do we think that a beginner needs to see a number of examples of the snapback in order to get the idea? Do we think that it is detrimental for the beginner to analyze the first snapback in order to understand the concept? Do we think that it is beneficial? Do we think that it is harmful to the beginner to explain what a snapback is? If we have explained the snapback concept to the beginner, do we think that it is easier for them to recognize the threatened snapback when they see the position? (I do. )

Edit:

Click Here To Show Diagram Code

[go]$$B Double snapback
$$ | . . . . . . .
$$ | X X X X X X .
$$ | O X O O O X .
$$ | O O X . O X .
$$ | . 1 O O O X .
$$ ---------------[/go]

is a double snapback. I don't remember how or when I learned the snapback, but I was about 5 kyu when I learned the double snapback. I learned it from one example, by analysis. I thought that it was wonderful.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

If you are reading this note, that thread is a must read.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

I think that there are parallels between learning go or chess and learning a language. As I recall, the number of patterns learned by a chess pro are of the same order of magnitude as the vocabularies of native speakers of a language. We pick up most of our vocabulary as children, and children go or chess players also pick up most of the patterns that they use in playing the game.

But for most of us to master our language just picking up words is not enough. One of the most powerful ways of building vocabulary is to analyze words into their meaningful constituents. Most of us do this as teenagers, as part of our schooling. But that does not mean that when we speak we slow down to construct our words morpheme by morpheme. That is not the purpose of that analysis.

Here is where I suspect that Hendriks constructs a straw man. He states:



IIUC, not even Hendriks recommends trying out random plays, but do the books he refers to really recommend analysis, planning, and then trying to find a play on every move? As I recall, Znosko-Borovsky recommended that procedure for chess, but only once or twice a game. Oswald Jacoby recommended such a procedure for bridge, but only at trick 1, and maybe a second time later in the hand. Honinbo Jowa recommended analyzing a go position only three times during a game. IMO, amateur go players are relatively weak in analysis and not all that great at planning. Books that help them analyze and plan could help them to become stronger players, without claiming that one should go to all that trouble on every move.

Hendriks makes a good point that seeing plays can help analyze a position, but, as I have pointed out in a previous note, plays and sequences of play are part of concepts in go (and, I claim, in chess as well). There is no strict dichotomy there.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

Side remark: oh, I didn't know Jacoby was a bridge player. He is a well known name in backgammon circles, too (IIRC he has a book, and there's a Jacoby rule for money play.)

_________________
Geek of all trades, master of none: the motto for my blog mostlymaths.net

At first I thought the title of this thread was some sort of sarcastic twist. In Go I'd say you should think first and move later. There's nothing worse than placing a stone, then later on looking at the position and realizing you're made a wrong turn AND also spotting the move that you should have played instead.

I get that this thread is about more than this, but truly, alot of new players might be lead astray if they think the thread title is advice to follow.

I can see it applying somewhat to the early fuseki, but not to the middle-game or endgame. It kinda follows along with the advice that if you've got 45 minutes to spend on a game then use 5-10 min on the opening and spend the majority of your time for reading out the middle-game fights and the endgame. This is because most amateur games aren't decided in the opening, but in the middle-game, so it makes sense to spend alot of time there in order to better read out fights and find good moves instead of mediocre moves.

Jacoby was quite phenomenal.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

Just posted on ChessVibes where there is a link to a long interview with ex-World Championship challenger Boris Gelfand:

Kerans asked Gelfand about the technique of calculating variations, mentions Alexander Kotov's Think Like A Grandmaster, a book that is now somewhat controversial because e.g. Valery Beim and Vladimir Kramnik have stated that they don't think in “branches of trees” at all.


Gelfand concurred and added: “My search for moves is also chaotic."

As a very poor bridge player, I always messed up my transfer bids.

I was unaware of how many different games Jacoby excelled at (until doing some web searches a few moments ago) ... I only knew his bridge reputation ...

Wasn't there some research done years ago that indicated that in terms of the calculation of variations, chess pros and experienced amateurs were about equal. The main difference was that the pros explored different variations.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

Visualize whirled peas.

Everything with love. Stay safe.

Yup, I've read this "recently" (last year or so.) It either appears (as a reference) in Coyle's The Talent Code or Greene's Mastery.

_________________
Geek of all trades, master of none: the motto for my blog mostlymaths.net

Adding to that, pros skipped variations that were "clearly" (for them, well stoked in chunks already) useless, whereas amateurs had to analyse them. So, in terms of width both (may) explore the same, but the pro does it much faster (better pruning AND probably better speed of forecasting moves) and accurately (practice, practice, practice.)

_________________
Geek of all trades, master of none: the motto for my blog mostlymaths.net

This is interesting and - while disconcerting for those among us who advocate the analytical approach - surprisingly little surprising. I've been writing down my thought processes in my journal and I've acknowledged that even those times when I play a decent thinking game, I don't analyze at every move. That goes to tell me something: it's one thing to cure automatic or random play, but analyzing every move is just not natural and, by the (chess) pros not even desirable.

This leaves me with the question: even if the pros have a chaotic move search, should we necessarily have one as well, while in our learning stages? Or should we force ourselves into a more analytical mode first, to forget about the metrics later and allow the acquired knowledge to be approached by a faster regime in our brain?

This is interesting and - while disconcerting for those among us who advocate the analytical approach - surprisingly little surprising. I've been writing down my thought processes in my journal and I've acknowledged that even those times when I play a decent thinking game, I don't analyze at every move. That goes to tell me something: it's one thing to cure automatic or random play, but analyzing every move is just not natural and, by the (chess) pros not even desirable.

This leaves me with the question: even if the pros have a chaotic move search, should we necessarily have one as well, while in our learning stages? Or should we force ourselves into a more analytical mode first, to forget about the metrics later and allow the acquired knowledge to be approached by a faster regime in our brain?

Well the statement is that pros know how to weed out useless plays because of their experience. That seems to presuppose that analytical thinking as a beginner is necessary, so as to gain that experience that can be later relied upon.

I do think there comes a point, though, where analysis turns into overanalysis. Analysis is good if a player is trying to learn or apply a proverb ("Should I hane at the head of two stones in this circumstance?"), or read out a life and death scenario, but beyond that I think it starts to become counterproductive. A player has to learn to gain confidence in their experience and intuition somehow.