Tsuumego question: Segoe's Tesuji Dictionary

[RBerenguel]

Recently I bought Segoe's Tesuji Dictionary, since it is quite praised in Sensei's Library and everywhere else. In case you don't know, problems are divided by subject, and then divided in "streams", A (hardest), B and C (easiest). Solutions are located at the end of each chapter. Currently I can solve C level problems, or at least approximate the solution (I've found several cases when I played 1-2-3 when only 3 was needed, in reference to this Sensei's page: http://senseis.xmp.net/?OneTwoThree)

I can figure out part or even some complete lines of B problems, but I can't tackle A class problems yet. And I don't know how to focus on the book:

* Should I only do C level problems?
* Only B and C, checking the answers to learn the new tricks?
* Try all, and learn from the solutions?

[Kirby]

My personal preference is to ignore the level and do the problem until I am 99% sure I have the right answer. If I don't get to that level of confidence, I either keep trying or skip the problem.

99% is a figure I made up to mean, "until I feel I am certain my answer is correct."

[oren]

I vote "Try All, Learn from Solutions".

Learn the ideas behind them and then go through the books again later.

[SoDesuNe]

I vote "Try All, Learn from Solutions".

Learn the ideas behind them and then go through the books again later.

Yep, that's my appraoch with "Dictionary of Basic Tesuji", as well ^^

[Cassandra]

The solutions are based on similar ideas. The higher the level, the more complicated, or longer, the solution sequence is.

In my opinion, a good method is trying to solve all C-problems of a chapter, before having a look into the answer section of this chapter. And then to continue with the C-problems of the next chapter.

However, it is acceptable to have a look at the answer after each problem's trial, if this pleases you more.

After you have finished "C", you might go through the C-problems for a second time, or start with the B-problems.

[RBerenguel]

Thanks everyone for your answers. I'll probably mix SoDesuNe's, oren and Cassandra's suggestions. I'll start by doing the C level problems of 3-4 chapters, look at the solutions (David's arguments for looking at the solution here are too strong) and then start again with C (flipping the book around). Once I've done 3-4 C chapters 4 times, I'll go for the first B problems, even if I'm close to solving one or two already (some are somehow easier than others)

[Cassandra]

(David's arguments for looking at the solution here are too strong)

Indeed, it does not make any sense, never to look at the solutions, if you have no Sensei nearby. There might be the danger of not recognizing that you were on the wrong road for a long time.

(some are somehow easier than others)

The levels given in books must not be taken absolute.

Often a problem gets a higher level, not because the given solution is so very difficult, but because there are several variations, which are not so easy to refute.

[Bill Spight]

There is actually some research on such questions. I am not really conversant with it, but here are a couple of basic points.

First, the 50% rule. Generally learning is most effective when you can complete around one half of the tasks. On that basis, it sounds like the C problems are too easy and the A problems are too hard, right now.

Second, overlearning. When people talk about going over problems until you can solve them very quickly they are talking about overlearning. A music teacher recommended that when you work on a passage that gives you difficulty, don't stop when you can play it right, keep going until it is something that comes easy. He made an analogy with how the body heals a broken bone so that the bone is stronger where it was broken. Overlearning! See http://senseis.xmp.net/?Overlearning.

BTW, I have doubts about going over problems until you can solve them at a glance. At some point the brain takes a shortcut via recognition. "Oh, yeah, that problem. Black here, White here, blah, blah, blah." The problem is that recognition may rely upon only a few of the stones in the problem. The result in a real game may be, "Black here, White here, oops!"

One possibility when you see the main lines of play at a glance is to consider more lines. See http://senseis.xmp.net/?GoProblemsTheFudgeFactor
Another idea, which I thought of recently, might be to close your eyes and mentally reconstruct the problem, up to the essential stones. If you can do that, then that should avoid the recognition problem I mentioned.

Another thing, which goes somewhat against the 50% rule, is that when you have a course of study that you are going to complete within a few months (like a semester ), the initial order that you go through the material does not much matter, as long as there is overlearning. That may not matter with the Segoe Tesuji Dictionary, because you can leave the A problem for later. But something like Life and Death, which you plan to master, you might as well not worry about difficulty.

[TKP]

Great advice Bill! From a practical perspective I've found that 'overlearning' is quite effective with go problems. I'd just go through the problems in order and then start again from the beginning. See how that goes for you.

[eyecatcher]

This discussion here was inspiring me to get a copy as well. I decided to, from now on, first take care of all "C" problems ... with looking at the solution only if i think i got it right... and maybe if i feel a bit more advanced (maybe a few month?) starting with the "B" problems later.

By the way i posted a review at my study journal/Go Blog with some pictures and expanation about how to see if it is whites or blacks turn (it is very easy ) http://www.truthandgo.com/?p=598