How to read

[Bill Spight]

"Vital point of the position" / "Shape point" -- as I understand these terms -- is much more than only board points where you can place a stone. There will be a SmartGo Book coming soon that covers this topic thoroughly.

Glad to hear it. There is some good material on shape in English, but we can certainly use more.

[Bill Spight]

First of all thanks to everyone for their answers, they were very helpful!

Not to disparage reading, but I am a big fan of seeing.

When I try to solve a tsumego, I don't really see the stones on the board. It is hard to say, I kind of try to save the information, like: this group now has 3 liberties, or after a throw in that is no longer an eye. Maybe that kind of technique isn't very practical?

It is practical. Analysis is good, and often a timesaver.

Should I try to just visualize, but I guess that would be even harder for me...

You may find the "Rapid Calculation" thread interesting, especially the posts on meaning and visualization.

[belikewater]

The less skilled solvers looked at points on which to play, but the more skilled solvers looked at potential eye points.

This is an interesting statement. Would you be able to flesh it out a bit more as to what the difference looks like in an example? I thought most players looked for eye points but I may be misunderstanding what that means.

[Bill Spight]

The less skilled solvers looked at points on which to play, but the more skilled solvers looked at potential eye points.

This is an interesting statement. Would you be able to flesh it out a bit more as to what the difference looks like in an example? I thought most players looked for eye points but I may be misunderstanding what that means.

My guess is that the less skilled solvers were looking where stones would be played in variations. Like, "I play here, he plays there, I play here, he plays there, here, there, here, there, . . ."

[Gotraskhalana]

I am also a big fan of shape. However, shape points are not the same as eye points. There is some research on eye movements of players solving tsumego. The less skilled solvers looked at points on which to play, but the more skilled solvers looked at potential eye points.

Thanks for this interesting bit of information which harmonises well with my views on learning and teaching.

I just want to remark that this does not necessarily mean that it is a good teaching goal to make everyone immediately look at the eye points. (See the research on how gifted children read leading to the conclusion that everyone has to be made to read whole words first.)

[Bill Spight]

I just want to remark that this does not necessarily mean that it is a good teaching goal to make everyone immediately look at the eye points. (See the research on how gifted children read leading to the conclusion that everyone has to be made to read whole words first.)

Oh, so that's where that idea came from. Thanks.

One point that I repeatedly make is that how talented children got to pro strength at go may not be the best way for amateur adults to improve.

[Charles Matthews]

I am also a big fan of shape. However, shape points are not the same as eye points. There is some research on eye movements of players solving tsumego. The less skilled solvers looked at points on which to play, but the more skilled solvers looked at potential eye points.

Oddly, taking the offered tangent, the number of genuine "shape" problems is a tiny fraction of the number of life-and-death problems. I don't recall seeing a discussion of why.

One reason would be that life-and-death is traditional "homework" to set for aspiring players is that it is easy to set. And for a pro at least, easy to assess.

Some other obvious points are that the answers are not negotiable (well, for a given rule set). And following on from that, it is usually a hell of a lot easier to see that status is "ko", compared to reading the ko to the end in a real game position.

Further, life-and-death positions are nearly always on sub-boards. The intricacies of context are pushed to one side. The but-outside-wall-has-weaknesses argument is usually not allowed.

So it looks like there is a spectrum of sorts. At one end you have life-and-death where the answer comes out in black-and-white as a status from a relatively short list. At the other, "shape" problems cannot even tell you that shape A is one point better than shape B in a given overall context, unless you are dealing with endgame technique.

Obviously, to get back to Bill, life-and-death problems that hinge on recognition of "half an eye" do have special value. Amateur low dans would probably recognise about a dozen patterns?

I think we could do with consideration of the "technique gamut" as a whole. I'm really a sceptic about tsumego drill, without some balancing study.

[longshanks]

A strong player gave me some advice to do tsumego as quickly as possible without worrying about reading. i.e. guess until correct using software which makes you play out the entire line. The idea is to learn the vital points by repetition. Although this approach may work in tandem with brute force iterative reading, I question its use by itself.

I definitely need to do some tsumego. I don't think that's in doubt. The only question is how much and how often for it to be productive?

[tapir]

Hello. I am a KGS 2-3k. Many people keep telling me that the fastest way to get stronger would be doing lots of tsumegos.
Some say you should do very hard problems, and just try to read as far as possible. Others say you should do easy problems and solve them quickly.
I've been trying both, but I don't really see any improvement in my game. Maybe I'm doing it wrong? When I try to visualize the stones after a few moves I forget how the position was and have to start reading all over again. How do you read? Thanks for your answers!

What helped me was taking a step from reading as "reading / silently talking myself through the moves" to only visualising stones on the board, i.e. from processing text to processing images.

Reading might not even be the weakest part of your game. Without reading everything falls apart, but if you get better at reading out ill-judged sequences, you won't see the quick and huge improvement in your game results some people report. Don't forget playing games over all the studying.

[Pio2001]

Hi,
I don't like doing tsumego, but I noticed that the ones that made me really improve were problems grouped by theme : throw-in problems, hane on the first line problems, shortage of liberty problems, descend problems...
It looked to me as a kind of cheating at the beginning, because the answer is in the title : the right move in throw-in problems is to throw-in etc.

I don't know if it improved my reading, but it definitely improved my local fighting skills. Now when I look at a position, I naturally ask myself "isn't there a possible trow-in here ?", or "what if I descend instead of hane ?", while, several months ago, I wouldn't have imagined that such moves were possible.

[Bill Spight]

A strong player gave me some advice to do tsumego as quickly as possible without worrying about reading. i.e. guess until correct using software which makes you play out the entire line. The idea is to learn the vital points by repetition. Although this approach may work in tandem with brute force iterative reading, I question its use by itself.

I think that that approach is in general good for DDKs, who need to build up their go "vocabulary". I also think that if you use that approach, again, you should focus on problems where you can get the first move right about half the time. One nice thing about software that steps you through the solution is that, even if the problem is too hard, somewhere along the line you are likely to hit on a subproblem that is the right level for you.

[bombur]

maybe its all the psychology classes i took in college, but I feel part of the reason people do not see improvement after putting in time solving tsumego is in your perception. what I mean by that is often tsumego is only focused on a very small area of the board. There is no time limit, no opponent, no consequence for inaccuracy, and you already have the awareness that their is a "correct" path to be found. It is easy to let your mind work freely in this scenario.

What has helped me is to program myself to associate every problem as part of a game I am playing. I visualize that I am the color to move, that I have an opponent waiting for my response, and make sure I find a solution in a fairly short amount of time. Even though only a section for the board is pictured, I try to let my mind believe that it is apart of a full sized board.

This might all be nonsense, but I have found that by mimicking the elements and environment that occurs in games, and forcing the brain to calculate in the same frame of mind translates to improvement in real time games

[WiseMouse]

Hi,
I don't like doing tsumego, but I noticed that the ones that made me really improve were problems grouped by theme : throw-in problems, hane on the first line problems, shortage of liberty problems, descend problems...
It looked to me as a kind of cheating at the beginning, because the answer is in the title : the right move in throw-in problems is to throw-in etc.

I don't know if it improved my reading, but it definitely improved my local fighting skills. Now when I look at a position, I naturally ask myself "isn't there a possible trow-in here ?", or "what if I descend instead of hane ?", while, several months ago, I wouldn't have imagined that such moves were possible.

I have a korean problem book that sorts the problem by technique.(Will edit in title later) It's just amazing to me that it can say "Hey stupid! Use a ladder!" and I still can't read some of them to completion.

Title is 바둑 사활: 1000 제. 1st volume of 2 if i recall correctly. Essentially 1000 L&D problems sorted by techniques&tesujis.