Two current members of the forum, jts and Li Kao, have recommended that I do Tsumego in addition to playing games in order for me to get better. Currently, I hate doing problems in general, since I find them excruciatingly boring to plow through. Yes, I'm one of those that prefers games to problems.

Here's my dilemma: I hate problems, but I have to concede the fact that jts and Li Kao are in fact correct here. I'm going to have to do problems in order to get better. The repetition and the exposure to different situations are in fact very beneficial. I don't know how to make solving tsumego fun. Therein lies the dilemma. How do you make Tsumego fun, so that it doesn't have to be a chore every time that I solve them?

_________________
My plan to become an SDK is here.

start with the easy ones

I don't think I recommended that you do tsumego, exactly. I said I would have recommended it if you hadn't said you weren't enjoying it.

You didn't quite answer two of my questions:
(1) Which tsumego sets is it that you feel are boring because they're too easy?
(2) How long was it taking you to do the Cho Elementary problems you found unpleasant?
(3) Are there any sorts of sequences in Go that make you go "omfg this is so cool"?

You certainly implied that I should be doing Tsumego.
And, I might add that your point has been taken.



1. I didn't necessarily say I felt that they were too easy. Maybe it's because I feel like I'm right in the middle so everything is either too easy or too hard. This could also be because my reading and visualization is sporadic at best.

2. The problems being too easy or too hard may or may not be the issue here. Maybe, I have to build up to doing long sessions of tsumego?

3. Anything involving a Ko fight, snapbacks are hard to see and they are pretty cool.

_________________
My plan to become an SDK is here.

It helps to find a set of problems which fit just at your level - they make you think a bit, but they don't take longer than a minute or two to solve. Personally, I didn't start doing tsumego until I was 8k, and I only did it infrequently then because I didn't enjoy doing it until I reached 1d. I think it's an acquired taste that you enjoy more once you start noticing it paying off in your own games.

Ultimately, if you want to get better, you must improve your reading. Tsumego is a great way to do that. However, there are other ways, such as playing lots of games, especially against someone stronger. Seeing stronger players play certain patterns can internalize those shapes into your own game.

As a little demonstration, here is a basic shape which can be cut:

Click Here To Show Diagram Code

[go]$$B Cuttable
$$ ------------------
$$ . . . . . . . . .
$$ . . . . X . . . .
$$ . X X . . . X X .
$$ . . O O O O O . .
$$ . . . . . . . . .
$$ . . . . . . . . .
$$ . . . . . . . . .
$$ . . . . . . . . .
$$ . . . . . . . . .[/go]

There are three ways that I can think of to discover that this can be cut: by finding the solution during a game, by solving it in tsumego form, or by having someone else cut you up in this shape. The problem with finding it yourself during a game is that it is easy to think that there is nothing all that suspicious about this shape, and not look for a cut. Or, you may look for a cut, but because you don't know there is one you may give up trying to find it. But when you have it presented in problem form, you know that there is a solution so you can persist until you find it. There's also the possibility of learning this shape by having a stronger player cut you apart when you make this shape, which is actually how I personally discovered this cutting tesuji.

Answer to tsumego:

_________________
We don't know who we are; we don't know where we are.
Each of us woke up one moment and here we were in the darkness.
We're nameless things with no memory; no knowledge of what went before,
No understanding of what is now, no knowledge of what will be.

1. Why don't you stick with too easy for a while? For example, the beginner exercises on Sensei's library: how long would it take you to do 10 of them, thinking through every variation for black and for white?

2. I was asking how long you took to help *you* decide whether they're too long. The modal tsumego should take you a minute to solve, with a few of them taking up to five minutes.

3. Do you mean the ko threats, or setting up the ko? gochild2009.appspot.com has 40 introductory snapback problems and and 163 introductory ko problems, so if you like that kind of tesuji, dig in.

And I just saw DE's post, so I should reiterate that I agree with him: I wasn't being coy when I said "tsumego would help, but you really just need to spend more time reading." If you like playing more than problems, play. But focus when you play, if you can.

_________________
http://tchan001.wordpress.com
A blog on Asian go books, go sightings, and interesting tidbits
Go is such a beautiful game.

I used to hate tsumego. I still don't 'like' to do them, as in, I don't go out and buy tsumego problems with a great 'Glee, more tsumego!' attitude.

I've expounded upon this dislike a couple of times, but, let me express the why.

A) Tsumego are not, save in specific corner cases or common shapes, often real world applicable. I particularly dislike ladder problems that contrive to snake, by way of first line stones, a ladder around the whole board, and back to its own head. They're pretty, and cute, and easy to read, but they are not, you know, useful, as such. The exceptions are corner tsumego arising from common joseki and noseki variants. These are very very useful, but only at a certain level.

But wait! I know what you're thinking! Mista Samurai, You can't say Tsumego aren't useful, they teach you Life and Death, which is an important part of mid and late game play, and you yourself win lots of games based on killing huge groups, and you learned to do that through tsumego reading exercises!

To which I'll answer:
Tsumego themselves present you with a very oversimplified situation. The surrounding stones are in place, the net is drawn, the race is on, your goal and drive are clear. In a game, you must harass and chase and harry and split until you finally find something surrounded and helpless, ready for that killing blow. The capability to create a life and death problem on the board is much much more crucial than the capability to actually kill the group, especially if you can, keeping in mind the whole board, harass and chase and create profit while your opponent struggles just to live. In as many games as I've killed huge groups, I've failed to kill anything, and crushed my opponent by attacking them into heavy, large unscoring groups, while building influence and territory by attacking. I've also lost games while making huge kills. Tsumego improve your reading. They do not win you games, and they are not useful for whole board thought. In fact, very often, after a stretch of intense tsumego activity, I will lose several games while trying to remember this fact.

B) Tsumego are easy to learn bad lessons from.
It's easy to learn general lessons which Do Not Apply by doing a lot of tsumego. I've seen other stronger players remark on this, where the movement of a single stone changes the content of a problem. This is especially true in a game, where the addition or loss of a single stone can drasticly change the content of the problem.


This isn't to say, I don't do tsumego, or that I recomend you not do tsumego. I've spent months doing nothing but tsumego, and viewed that as a very profitable area of study.

But, tsumego are not a pancea for strength.
They are a very specific tool to increase your capability to read in very specific situations.

Reading is the key though.

I don't read nearly enough (as anyone who's watched me play can tell you) so I did tsumego until a lot of simple reading in corners and L&D I can do at a glance. I don't need to spend a lot (any) time looking at dusk eagle's problem to tell you that white doesn't need more than 1 particular stone of that line to make the cut work, and that with that many stones in white's line, black cannot connect with one move, and that the better answer for black is 2 at 3 so that the corner is safe.. and I learned that stuff from doing tsumego until I can see all of it at a glance.

But, I still dislike tsumego, because people tend to declare that doing tsumego will make you stronger like it's some sort of magical salve, when in fact, tsumego are not useful unless you know how to make use of the reading techniques they're teaching.

Still, to make them more palatable, I suggest..
A) Do craploads of them. I mean, put smartgo on your iphone, or find a way to put a load on your android, and do them every time you have half a second. on break at work? Tsumego. Bored at school? Tsumego. waiting for your opponent to move? Tsumego. They don't have to be hard, but if they are, who cares?
B) Don't care about getting them right. This is a terrible thing to do, but it's how I got through them. Just pick an answer at random, and learn why it's not right. Don't look at the right answer, don't spoil the problem, just pick something, read it out, and try to find the fail. (or, let smartgo show you the fail). Learning the failures will teach you as much as learning the correct lines. After one answer, if it's wrong, mark the problem to be done again, and move on to the next. When you've gone through all your current set, go back and redo all the ones you got wrong. Leave out whatever you got right. Do this until you've got no more remaining, and start over. When you can solve each of the problems, the first time, without any errors, get a tougher set of problems. really internalize the problems. Know why all the answers don't work, and why the right answer does. The goal is not to study right answers. The goal is to study your wrong ideas, to see why you were wrong.
C) Do craploads of them. I mean, all the time. You should be doing one now, because I'm rambling and boring.

This works for tsumego that take 10 minutes to solve/read out. It works for tsumego you can solve at a glance. This takes all the stress out of them, as you don't care about getting them right. It won't make them 'fun' but it does give me at least, a certain satisfaction, when I solve the problem right the first time.


Speaking of. I need to find a good smartgo problem program for the android. Mr Keirulf? No chance? Ok. I'll just have to keep hunting.

"The best way to study go is to play go."

you don't like tsumego? don't do it. if you have a chance to play games instead, it is just good in my opinion. especially considering you declare yourself as a 12k KGS, it is not yet neccessary to improve by doing tsumego. if you want to train your reading, just try faster time settings / smaller boards to put more emphasis on tactics and you are fine

i don't know you, maybe you are stuck at your rank and need something to unstuck you or you have very weak reading or whatever and tsumego is the right choice, but i got to my 1d KGS only by playing, plus reading a few books and once going through 1001 Life & death problems (a very minor part of time i spent with the game). i don't say it is the only right way, just a possible alternative

_________________
Spilling gasoline feels good.

I might be wrong, but probably not.

Playing go is your hobby, not your job that your life depends on. I'm not at all convinced of the benefits of doing something that you "hate" and apparently strongly resent. Just stick to playing games (and reviewing them) for now until in the future you may discover your taste for tsumego. Or look at pro games, if that interests you. But right now, from the sound of your post, I think forcing yourself to do tsumego in spite of your strong distaste for them, is more likely to drive you away from the game than to benefit you to a degree where it's worth the boredom and the "suffering".

You're not playing Go for a living. You're playing Go for fun. So do what's fun for you.

Samurai, you seem to be implying that reading is all about pattern recognition. While pattern recognition is extremely important, tsumego are probably not the most efficient way to acquire it (though certainly not considerably worse than playing or reviewing games either).

But reading is not just that, most of all it is visualising concrete variations. Other than actual reading during games, tsumego are certainly the best way to practice this skill of thorough evaluation. There is no reason not to continuously practice this skill, because in the end, there is no limit to how fast, far, and precisely you can calculate variations and the better you are at it, the better you will do in your games. While go is not chess, where calculation is (almost) everything after the book moves, if Cho Chikun says that it makes for 80% of the middle game, that is probably not too far off the mark.

To this end, the way you suggest to do tsumego (fast and without trying to be correct) seems very counter-productive to me and only geared towards pattern recognition. Sure you will gain something from memorising the patterns, but as you describe very well in your post, the value is not that big (then again, it is probably a good idea for common tesuji).

I would strongly suggest to do exactly the opposite. Don't worry too much about knowing the correct move, but instead adopt a somewhat distanced, methodical approach to it and go through as many variations as you feel capable of. Whether you end up completely solving the problem or not, you will gain something from the exercise.

Of course you can practice your reading skills just as well during actual games, but the benefits of tsumego are that you can always do them, whether you are on the loo, on the train, or about to fall asleep. You also get instant feedback, whereas in a game it is easy to get complacent and not even try, because just placing stones based on intuition and pattern recognition can seem good enough.

Pattern recognition IMO is just a side-effect of doing a lot of everything (playing, reviewing, tsumego), and the more you enjoy something, the more you will want to do it.

As for enjoying it, with both chess and go problems I noticed that I started enjoying them a lot more once my visualisation skills improved. There is a lot of satisfaction to be gained from looking deep into a position, and successfully discovering its nuances (while on the other hand, rote memorisation of patterns seems extremely dull to me). As such I am sure you will start enjoying tsumego eventually, but as others have said, just spend your time with whatever you enjoy the most right now. The important thing is to practice reading, not specifically to do it with tsumego.

I just focus on doing them quickly. Most of the tsumego I do is with GoGrinder on the phone. Just try to figure out the critical points and go through the problems as quickly as you can. You'll get some wrong and then go through the set again. Since I generally do it in places I only have 5-10 min at a time, it doesn't get boring.

Just to reiterate, because you've been getting different advice, my suggestion is to choose problems that are sufficiently easy that you can exhaustively look at all of white's possible responses, and still be done in a minute or two. Although John F. and David O. ganged up on me for saying this a few months ago, I think you're much better off choosing easy problems, going through the possible solutions exhaustively, and not looking at the answers, than if you choose slightly harder problems, click away randomly, look at the solution, and move on to the next problem.

The point of tsumego, imo, isn't really to train your eye to intuit the right first move. There's a sense in which they do (that's why I said that if you had done tsumego, you wouldn't need 13 minutes to see that M19 snapback), but there's a broader sense in which the move in tsumego is always going to be the wrong move. Hane on the first line: generally, a miserable move. In tsumego, an amazing move. Playing the 1-2 point: so bad, that it has become code for "please resign". In tsumego, and amazing move. Self-atari: makes people scream expletives at their computer. In tsumego, we call it 石の下 and that makes it okay. Making dumplings: painful. Playing a stone in enemy territory where it's sure to get captured: a painful mistake. Making dumplings in enemy territory where they're sure to get captured: well, that's 中手! Duh.

Maybe I'm going overboard, but my point is that randomly zipping through the answers to tsumego and hoping to absorb patterns isn't the real advantage of tsumego, because cultivating an instinct, through pattern-recognition, to play on the first line, on the 1-2 point, to self-atari, to make dumplings, to play in enemy territory, is actually an ambivalent thing, if not a bad thing. The real advantage of tsumego is that they help you learn to sit down, close your eyes, and read. Two variations of three moves, then three variations of six moves, then four variations of ten moves...

Again, and I can't stress this enough, 'looking at the answer' is not in my suggested method. Try, fail, don't worry about it. That's my method. Learn why you failed. Read out why you failed. Don't spend time trying to figure out the right move if you don't enjoy spending time reading.

I don't do pattern recognition. As I said before, it's almost always wrong. One of my strengths is, however, a near photographic memory, which allows me to quickly check shapes against known shapes. But I don't do that, generally, except when evaluating possible places for first moves.

I'm not trying to absorb the pattern, I'm trying to 'see' all the moves at a glance. Like, when I look at Dusk Eagle's problem, I don't recognize the pattern, but I do 'see' the first move, followed by black's attempts to stay connected. Then I see that connection is futile. Then I see that black can't connect with any one move, and shouldn't try. None of that has to do with the pattern. It all has to do with how much of the solution I visualize at a glance.

Again, my method is not a 'good' method for doing tsumego. As you can see, everyone thinks I'm suggesting something I'm not (pattern recognition) or thinks I'm doing it wrong (by not reading out every possible variation). However, it does make tsumego less like a chore where I have to sit down and read everything out, and more like an exploration.

Reading is a huge part of go. Learning to do that reading quickly is another huge part of go. Also, Dusk Eagle's problem is another example of one of the flaws of tsumego. The following is hidden because it deals with the final answer for the problem.



In short, No, my way is not the most complete way to study tsumego. Nor would I describe it as (edit: A good way to do) tsumego. But, as someone who has hated tsumego in the past, it's the way that I've found to make them bearable.

Samurai, but what you describe is still pattern recognition. It may be confusing in Go because of the term "pattern", but it does not just apply to patterns of stones, i.e. shape (though arguably everything in Go is shape).

So when you say that you see the moves instantly, that is because you recognise the pattern (or sequence) of moves to cut this shape. Of course you are not supposed to actually play out that sequence, that goes without saying. It is also not enough to see that this shape can be cut. Seeing this sequence only helps to answer the local question of "can this be cut", which makes reading out the (local or global) situation a lot easier and faster. E.g. if this shape would appear in a variation 20 moves down the line, most players would probably struggle to read that it can be cut, without recognising the pattern. But to get to this point, you still need to be able to analyse a tree 20 moves deep and correctly visualise the result.

Exploring a problem by trying out different moves, without reading out everything, will help you learn new patterns, but it will not teach you how to correctly analyse a situation. Since you call that process "a chore where I have to sit down and read everything out", I guess you don't enjoy reading very much, so it is understandable that you would rather to Tsumego in a way that is more enjoyable to you and still helpful.

Because you approach Tsumego this way, it is also understandable that you don't consider them all that useful. E.g. the contrived ladder problems seem useless to you, because they don't teach you any patterns that are applicable to real games. But that is missing the point of those problems, which is to practice seeing a large number of moves ahead. Ladder problems are excellent for this because they don't branch very much (sometimes not at all), so you can focus entirely on practicing the skill of playing out the moves in your head, and to visualise the final outcome. It may be a chore, but an immensely useful one.

Start with the easy ones. If you still hate life and death problems, do tesuji problems instead. If you hate them too, read ladders instead. Short ladders, long ladders, ladders running into every kind of position, twisted ladders, ladders going around two corners, ladders going back to where they came from...

I knew you wouldn't be able to resist.

I feel as if "and if you don't like go, play it anyway!" would be even more Helelish.

_________________
Occupy Babel!

3...2...1...moderation?

_________________
Occupy Babel!