How to get better at Go?

[tekesta]

http://turtlemechy.wordpress.com/2013/02/25/how-to-get-extremely-good-at-go/

I agree with some of the points the author made in the first third of the blog post. Especially that brute-forcing Tsumegos is wrong and that openings are far overrated compared to the endgame. After that I feel it gets murky quite fast.

How strong is the author and how far did he follow his own advice?

The author of the article posts mostly mathematics-related material, but apparently no rank or Go server is posted, so I'm lost as to how to ascertain his/her playing strength. The info in the bottom ½ of the article does appear to be nonsense, but the author does make it clear that if anyone wants to acquire the playing strength of a pro, an exceptional amount of time and effort will have to be invested. Which is why most pros began learning Go when they were children or adolescents; it's much easier to learn complex things at that age, especially in a fun and challenging atmosphere.

I have been applying for myself the advice presented in the article and my Go abilities seem to have strengthened some. I replay game records on a regular basis and do life & death problems. What I need to do more of, though, is play actual games.

[Bill Spight]

In Go, there are tsumego, which is the name for puzzles depicting Go board positions, whether local or global (whole-board).

Small point. Most whole board problems are not tsumego.

[Bill Spight]

http://turtlemechy.wordpress.com/2013/0 ... ood-at-go/

It's bullshit.

Not that there may not be something useful there.

But it's bullshit.

[skydyr]

http://turtlemechy.wordpress.com/2013/02/25/how-to-get-extremely-good-at-go/

It's bullshit.

Not that there may not be something useful there.

But it's bullshit.

I got the impression that the author is basically accepting as an axiom that the closer you are to the end of the game, the more important or fundamental it is, and extrapolating from there.

[topazg]

http://turtlemechy.wordpress.com/2013/02/25/how-to-get-extremely-good-at-go/

It's bullshit.

Not that there may not be something useful there.

But it's bullshit.

Not only do I complete agree, but you made me actually laugh out loud, so you get a like and a comment


My advice is to accompany any tsumego and go problems with playing, and more playing, and more playing. If something goes disastrously wrong, review a little to see you could have tried instead to see if you could have achieved a better result, but don't go overboard. Just pick one or two bad local results in a game to play around with, then go play another game.

[Kanin]

There is always this clear and concise guide to studying go at Go Game World: http://www.gogameworld.com/2013/03/30/h ... go-series/

Good luck!

[Boidhre]

http://turtlemechy.wordpress.com/2013/02/25/how-to-get-extremely-good-at-go/

It's bullshit.

Not that there may not be something useful there.

But it's bullshit.

I got the impression that the author is basically accepting as an axiom that the closer you are to the end of the game, the more important or fundamental it is, and extrapolating from there.

I'm more thinking that what he's saying makes some sense once you get strong enough that your games are close pretty regularly (as opposed to ddk where close is a 20 point margin ). Endgame first is pretty dubious advice for a pure beginner. Though any kind of problem would probably be beneficial alongside games early on.

[tekesta]

There is always this clear and concise guide to studying go at Go Game World: http://www.gogameworld.com/2013/03/30/h ... go-series/

Good luck!

The article on this website is more concise and detailed than the Turtlemechy one in explaning study régimes. I've read it before and it reminds me a lot of what I mentioned earlier. As well, a couple of the teachers on GoGameWorld are Chinese pros, so I can go to the bank on their advice

[tekesta]

Small point. Most whole board problems are not tsumego.

Hmm. I always thought tsumego meant any Go problem. What requirements must be met for a go problem to be considered tsumego?

I do not know the term for whole-board problems in Japanese. Hmm... is it 全盤題 (zenbandai)?

Large problem collections are a recent phenomenon, aren't they? Or did large problem collections exist before Cho Chikun's Encyclopedia of Life & Death? I imagine Go Seigen, Kitani Minoru, and their contemporaries studied weiqi in very much the same ways as their predecessors. They would mostly replay game records and play actual games, with some problems to strengthen analytical ability.

[Bill Spight]

Small point. Most whole board problems are not tsumego.

Hmm. I always thought tsumego meant any Go problem. What requirements must be met for a go problem to be considered tsumego?

See http://senseis.xmp.net/?Tsumego

[karaklis]

See http://senseis.xmp.net/?Tsumego

I was just surfing around at Sensei's and found the following approach to study tsumego:

http://senseis.xmp.net/?BeginnersGuideToGoProblems

1) Quick and Fast (just guess the first move)
2) Slow and Sure (read out until you found the solution)

Which one should be tried first when working on a certain tsumego collection? (I've never used the first approach, so maybe that's the reason why I've never become strong..)

[John Fairbairn]

Which one should be tried first when working on a certain tsumego collection? (I've never used the first approach, so maybe that's the reason why I've never become strong..)

Both styles have pro advocates, but I sense that the fast method is garnering more and more support, and has always been used more. However (actually a big HOWEVER), it is not about "guessing" the first move. It is about recognising it, and the method involves no less work than the slow method. The main idea of the method is to do the same problem set repeatedly (and many times) so that the right move just leaps out at you. The first few times you do a problem you are supposed to give it some thought, of course, but if your x-ray eyes can't see the skeleton fairly quickly, you are advised to turn to the solution and get out the autopsy knife. In this process, like Dr Mallard, you make sure the "victim" talks to you and you pay particular attention to the different defences that can be tried. You will know you have mastered a problem when you can look at it after some time away from it and see not only the first move but also the skeleton.

[Mike Novack]

You need more than the 1st move.

Even if the problem were presented without knowledge that there is a solution (problem presented in the form "what is the status?") very often the first move is obvious. In many situations, there is clearly a point that "if the opponent takes this point the matter is decided". But while that first move might be a necessary condition all by itself maybe not a sufficient condition.

However, being able to quickly spot the first move is an important skill all by itself and in over the board play guarantees that "the problem remains a problem" (the status remains undecided for the moment).

[gowan]

This might be a good thread in which to recommend "three move reading". Kato Masao 9p published several books on life-and-death, tesuji, etc., in which he advocated three move reading. This works as follows: you choose a move, then think about what your opponent would/could do in response, then what your next move would be in response to that. Repeat this with a different choice of your first move until you find something you think is good. Some pros seem to think that if you can consistently do three move reading it will take you to dan level amateur strength. I think that if you begin with three move reading and keep it up before long it will become automatic and your reading will progress to greater depth.

[SoDesuNe]

This might be a good thread in which to recommend "three move reading". Kato Masao 9p published several books on life-and-death, tesuji, etc., in which he advocated three move reading. This works as follows: you choose a move, then think about what your opponent would/could do in response, then what your next move would be in response to that. Repeat this with a different choice of your first move until you find something you think is good. Some pros seem to think that if you can consistently do three move reading it will take you to dan level amateur strength. I think that if you begin with three move reading and keep it up before long it will become automatic and your reading will progress to greater depth.

My problem for the moment is that I instictivly stop after (around) three moves. I have to actively remind myself that I should read deeper.
Just yesterday I neglected to play a simple Tesuji because after three moves I thought my opponent could thwart my plan, although it was completely fine after five moves. In another game I didn't kill a corner because I thought I had to defend a cut (otherwise a tail from my group would die and my opponent could live either way). After eight moves (straight forward ones) it was clear that I could connect this tail and he just could have caught two totally unrelated stones.

Maybe I solve too many easy problems right now, though even there I get so many wrong on goproblems.com right now. I will observe it further =D