I'm wondering how people practise L&D, problems suitable for one's level I mean.
My friend told me to read everything in my head and never look at the answer until i get it myself.He's even against playing out the moves on board. But i feel visualise is quite important, I just cannot do everything in head now, always forget one stone somewhere or that group is actually in atari after several moves, I'm SDK.
What is the best way to do L&D then?

_________________
长考出臭棋.

Honestly? I just don't do them at all... I find them boring, dry and contrived.

That said, I do find problems in actual games looking for tesujis and the like interesting. But concocted L+D? Not so much.

Mostly what your friend said. That trains reading.

To train intuition, you can try lowering the level of difficulty and go through them fast, reviewing the ones you get wrong.

_________________
a1h1 [1d]: You just need to curse the gods and defend.
Good Go = Shape.
Associação Portuguesa de Go

Playing out the variations on a board or on a computer with stones to try and figure out an L&D problem is silly. Do you get to pause a game and ask your opponent if you could play out some variations in a corner before making your move? Course not, so it makes no sense to do so for tsumego.

The best way is to simply read everything out in your head. If you can't solve it, don't look at the solution - just go to the next problem and move on. Only look at the solution if you're 100%, and I truly mean 100%, sure you've got it. But even here there is some debate: after all, if you're truly 100% sure, why is there a need to look at the solution in the first place? Regardless, I think most people can agree that you shouldn't even consider looking at the solution unless you are, without a doubt, certain you are correct. Note I said "most" and not "all"; I have seen some high-leveled players make the suggestion of just looking at the answer when you can't figure it out, but I just can't agree with this and feel that, in the long run, it'll cripple you.

Also, by solving I don't mean just figuring out the first move to the solution sequence. You should not only be able to understand the solution line, but also why other lines fail and understand the intricacies in there as well. You'd be surprised how elegant some of the failed sequences are in problems, the traps that the tsumego composer put in there can at times be dazzling.

I advocate not putting stones down and working on being able to visualize and read moves out.

However, I disagree with the never look at the answer rule. There are simply problems that I will never be able to answer, because I simply don't know the tesuji. If I spend over 15-30 minutes on a problem and I just can't get anywhere, I'll look at the answer. I also file it under problems to do again later. I don't think that never looking at the answer is always the best idea, because finding out why you're always wrong has value too, I think. Just giving up too easily is the thing to avoid.

I think there's a need to look at the solution because sometimes people has 'blind point'. We don't try every possible move in the local area when doing L&D, our knowledge helps to choose the place that's more likely to be the correct move than others.In L&D, people will first think hane, placement, descent, jump etc...and even though one thought he has read all, but he may still miss an unexpected move.

_________________
长考出臭棋.

There are two distinct skills involved in reading.

One is being able to visualize paths through the game tree, making moves in your head and taking them back, even if a sequence involves many moves or a large number of alternatives. (This skill is not so important in a turn-based game.)

The other is guessing what the correct move is likely to be, by pattern-matching with the knowledge base stored in your brain. You can save a lot of time this way, just by starting your reading with something like "Well, I know that if anything works it'll be this placement."

The read-everything-out-until-you're-absolutely-sure method works well for training the first skill, but doesn't help a ton with the second.

Allowing yourself to look at the answer if you're stumped means that you can begin to pick up the patterns that stronger players have completely internalized. If you say, "Whoa, the solution is that placement I never would have thought of... oh, I see, it takes advantage of his shortage of liberties, okay, neat...", then you've picked up a piece of pattern-matching that hopefully you can use the next time you see a similar problem. I don't think that forcing yourself to learn all of the patterns completely through trial and error is the most efficient learning method.

i used to be stubborn and didnt look at answer until i am confident that i have to answer.
i remember spending whole weekend on one problem.
now...i dont do that any more. (maybe because i dont have time)

_________________
"The more we think we know about
The greater the unknown"
Words by neil peart, music by geddy lee and alex lifeson

If I ever do problems(I do them off and on, but consistency is what counts), I follow the first option. I think Sol.Ch covered it pretty well. Though, I will say that the other methods cannot be considered bad either. It's mostly just preference. I was always told, and I believe, that reading is very important in this game. So, I try to read out each and every variation, and try to push myself to read more. I think this helps with concentration in Go, but it has benefits in your regular life too(to learn to concentrate on one particular thing step by step and if that doesn't work, to move on rather than give up).

_________________
I know nothing.

I'm with the folks who say it's okay to look at the answer if you don't get it. How long before throwing in the towel? I'm not sure, but I actually think that going too long is bad. For me, I go maybe 5-10 minutes or so before I look at the answer.

Here's why I do that: When I spend, say, a half-hour or more on a problem I can't solve, what's happening in my head is that I keep running through all the wrong sequences more than I should. So, rather than training my brain to pop out the right answer, I'm actually training it to pop out the wrong one! Once I've looked at the answer, I go back to the problem page and visualize the correct sequence a couple times to help reset my neuro-pattern-reading brain.

So, in general, when I'm doing problems, I pick a book where I'm hitting maybe 50-80% correct within my 10 minute time allotment. I might break it into section; say 10 problems. When I'm through a sections, if I've missed more than a handful, I'll repeat the section (much faster, now). I try to do that the same day, so I'm actually remembering the solution, not having to reread it. (I try to read the sequence where applicable, not just the first move.) If I can come bay the next day, and then several days later, and quickly see the right sequences, then I know it's sinking in.

As you can possibly tell, I think there's a lot of knowledge being stored subconsciously, here. And I think the way to get that knowledge into the subconscious is to keep repeating the pattern mentally. (Think of musicians -- their ability to sight read comes about this way.) Does that help my game? No -- I've been 5K for too many years! So, what I say is just my own style -- may not work for others, needless to say.

btw: I didn't answer the poll. Don't know about everyone else, but just looking at the list of choices made my brain go foggy! Couldn't bring myself to read through it!

I actually enjoy doing tsumego. I'm rarely 100% sure about my answers, and for me, seeing if I'm even on the right track is part of the fun. I've tried not looking at answers, but when I'm really stumped I get frustrated and that's just counter-productive: no fun = no tsumego. I do however have the added advantage of being able to instantly forget both problem and solution



Just out of curiosity, are there chess problems that are graded in a similar manner to the way go problems are graded? "mate in 3 moves or you have broccoli for brains" turned me off as a kid.

_________________
Patience, grasshopper.

For some reason I've always thought that most chess problems have surprisingly little to do with actually playing chess. I may be completely wrong, of course.

_________________
Offending ad removed

I think that doing problems is mostly underrated.

Problems are teachings about the basic stone interactions. What good is it to "know" or "feel" that a particular formation is safe if you lack the skill to defend or attack it?

Often, when you have been outplayed in the opening of a game but do not know where, you simply overlooked that your opponent has left a weakness somewhere which you did not exploit.

Many "peaceful" players feel that they undeservedly lose even though their opponent has played all kinds of weird invasions and juggled with three weak groups throughout the game. In the analysis, stronger players can then point out "Hey, why don't you cut? He has no defense there." or "You don't need to defend that group, it is alive!" or even "Argh, you can't let him get away with that!"

Sometimes you need to put away the velvet gloves and take out your axe. If that axe is then made out of paper, you will look a bit silly.

By the way, you should not do complicated problems. If you need more than one or two minutes to completely analyze a problem, it is not your training level yet. You can of course do it, and if you cannot solve it, you may even look up the analysis, but this should not be your main curriculum. The kind of problems that are suited for your level are those that you can solve in a short time, and you should keep solving them until you can see all variations at a glance.

_________________
A good system naturally covers all corner cases without further effort.

I used to enjoy doing tsumego, and I did a lot of them last year (about 3/4 L&D problems, the rest tesuji problems) in order to memorize them - until I got burnt out last September. I still have to find out why that happened.

About two weeks ago I have started to do tsumego again, "Graded Go Problems for Beginners" Vol 3. Comparing to one year ago (where I had roughly the same rank as now) most of the problems are quite easy, at least those where reading is involved. I got 90% right, and those that I fail at are shape/opening/joseki problems. I repeat those problems that I got wrong the previous day. I also mark some problems that I got right but that look particularly interesting to me and repeat them. When I have finished the book I will repeat all the problems that I got wrong or that I had marked.

And - hey! - it's fun again

QFT:

There may be a terminology issue here.

"Chess problem" has unfortunately come to mean problems of the "Here's a random scattering of pieces on the board that looks nothing like a real game position, find the mate in 2" variety.

The more useful "Here's a position that could happen in an actual game, find the best move" problem is usually referred to as a "puzzle", which name I don't really like. But those "puzzles" are basically exactly as useful to chess improvement as tsumego is to Go improvement.

Here are a couple of good sites that feed you chess puzzles and track your results so they can adjust the difficulty: Chess Tempo, Chess Tactics Server.

But there're only 3 of those, and I already memorized the solutions!

My chess career was so short and happened so long (almost 30 years) ago that I don't actually remember much of it, but I'm pretty sure I never came across those puzzles, which, as you describe them, sound useful indeed. I only remember the other kind, which I was never good at, and which I thought had not much to do with chess.

_________________
Offending ad removed

Being a Chess player myself, you are, in my view, half right and half wrong. Most tactical and positional problems have a lot to do with actual chess playing. Sometimes they teach an essential concept, show a nice little tactical gem, or show how to win or draw from a problem. Doing tactical problems is indeed a way to improve one's chess elo rating.

Now, for the part of this in which you are correct. Some people "compose" problems for fun and they have artificial board positions or the type of puzzle isn't supposed to be best play, i.e. helpmates or selfmates. These type of puzzles really don't help improve one's rating in the chess world as they're supposed to be for fun.

_________________
My plan to become an SDK is here.

Hmmm...This may explain why I've stopped improving at chess. Actually, in all seriousness it's about half the reason. The other half the reason is that I've gotten lazy in doing tactical puzzles. I still enjoy the game, don't get me wrong, but my improvement has been very slow, if at all.

I'm going to go out on a limb here and say that looking at the answer is not always bad, and does not always cripple you in the long run. Why? Simple...Let's say that you're doing a training regimen for several weeks with 1200, or so, problems in order to improve. As you start out you do all the problems in 12 days, then you repeat the process in 8 days, then 6, then 4, then 3, then 2, then finally you're going to do all the problems in a single day. After 10 minutes or so, you should look at the answer study it, then move on. The first few times you hit the "problem" problems you're training your reading ability, and the last few times you're training your pattern recognition ability. Does looking at the answer hurt in this case? Probably not, and in all actuality it probably helps.

Though, I think I know the situation in which Araban is talking about. The situation is giving up on the problem, and saying you are "not good enough" to solve the problem. In this situation, one does indeed cripple one's ability.

In essence, just because you can't solve a problem right now, it DOES NOT mean you will never solve the problem. It just means that you might have to stretch a little before you CAN solve it. All, I think, Araban is saying is just that. You might have to stretch, sometimes a little and sometimes a lot more, before you can solve a problem. There is nothing wrong with this fact. In fact, I have found that the more I struggle to learn something and the more frustrated I get, the more I learn and grow in whatever I've struggled in.

Personally, as I embark on my journey in improvement in Go, I'm going to try to follow this little piece of advice. Looking at solutions, at this stage would only hurt me as I'm only 18-19 kyu.



Hmmm...To look or not to look at the solutions? That is the question. In practice, I agree, with GoCat that going too long on a problem is a bad idea. Also, repetition is the key to learning, and when you look at the solution to something, you should repeat that problem several times in the following days to make sure you are learning, and not just skating by. I think that looking at the solution, too early, has harmed my chess ability as well as the non-repetition once I looked at the solution.

How long is too long to spend on a problem? I don't know, but all I can say is that it's different for each person.

All in all, I think that both Araban and GoCat have good ideas. When I actually do tsumego, I'm going to somewhat blend their ideas. Though I think that Araban is right, with the idea that we all need to stretch our abilities. I'm going to start by never looking at the answers, because to improve I need to stretch, and later when pattern recognition becomes key, I'll start doing problems repetitively, maybe 3 out of 10 max, I'll look at the answer and focus on those.

Also, the point of Tsumego, IMO, is to learn something. As with all puzzles, if you're not learning anything you're not stretching enough since they're too easy or you're giving up too easily, again IMHO.

_________________
My plan to become an SDK is here.

Lately I've been doing problems out of Making Good Shape (in what little time I've had for problems). Most are "How should black play to make good shape for his stones?" Well, I don't know what shape looks like, so I can imagine all sorts of different lines of play, but even on the easier ones, I won't be certain that I'm right. I know what life or death is, but shape is such a nebulous concept...

I don't think it's really possible to never look at the answers. Sure, you should do your best to be sure of the answer to a point, but certain tricks are just too hard to learn without having actually seen them at least once. It's almost like never letting stronger players point out better moves when you review your games. The Socratic method only goes so far.

_________________
KGS 4 kyu - Game Archive - Keyboard Otaku