Life In 19x19

One of the most contested topics among Baduk/weiqi/Igo players are the subjects of:

1: How to improve through set puzzles relating to Igo, with defined and usually clear objectives. Known most commonly among westerners by the japanese word tsumego. How do you use them properly?
2: Kifus of professional games. How do you use them? Which games should you look at?

After quietly looking through many discussions, and pondering/sitting very calmly
while levitating through intense concentration/shamelessly jumping up and down in excitement over what I like to think is my great idea, I've come to this conclusion: I know nothing, I cannot come to a conclusion yet; I still have a lot to learn. It is a more complex issue than many try to conveniantly paint it as. But I hope that very statement allows me to change things from "I know nothing" to 'I know nothing (much!)' and the little I do know, I believe to be important.

And I'm so confident that the "method" I'm about to mention is closer to the right track that...

"...Within 5 years there would be no-one in L19 community who are impossible for me to win against in an even match..."

...Mind you, I'd have touch on a few other issues that are surprisingly related to studying.

And if you were wondering, no, I am not sitting with my fingers crossed hoping a pro doesn't make an account. On the contrary, I'm ready to rise above the bar, no mattet how high that bar may be.

ps: Next post arriving shortly.

psps: A more accurate desription of my goal is to secure a win against everyone on this forum at least every 20 games.

Elom wrote:One of the most contested topics among Baduk/weiqi/Igo players are the subjects of:

1: How to improve through set puzzles relating to Igo, with defined and usually clear objectives. Known most commonly among westerners by the japanese word tsumego. How do you use them properly?
2: Kifus of professional games. How do you use them? Which games should you look at?

After quietly looking through many discussions, and pondering/sitting very calmly
while levitating through intense concentration/shamelessly jumping up and down in excitement over what I like to think is my great idea, I've come to this conclusion: I know nothing, I cannot come to a conclusion yet; I still have a lot to learn. It is a more complex issue than many try to conveniantly paint it as. But I hope that very statement allows me to change things from "I know nothing" to 'I know nothing (much!)' and the little I do know, I believe to be important.

And I'm so confident that the "method" I'm about to mention is closer to the right track that...

"...Within 5 years there would be no-one in L19 community who are impossible for me to win against in an even match..."

...Mind you, I'd have touch on a few other issues that are surprisingly related to studying.

And if you were wondering, no, I am not sitting with my fingers crossed hoping a pro doesn't make an account. On the contrary, I'm ready to rise above the bar, no mattet how high that bar may be.

ps: Next post arriving shortly.

psps: A more accurate desription of my goal is to secure a win against everyone on this forum at least every 20 games.

You sound like Charles Atlas xD. Btw, for a 1/20 win rate you'd need to be something like... 4d EGF? (since we have a couple EGF 7d here) ballpark number, maybe 3d is enough

One of the fustrating aspects I frequently find on study discussions is that tsumego is treated as one single activity, causing confusion over weather to look at the answers. A similar problem arises over weather it's useful to study/replay professional-level games, stemming from the same misunderstanding. There is reviewing pro games and there is reviewing pro games. Tsumego, then there is also tsumego. In fact, replaying pro games is closer to doing tsumego than replaying pro games. Solving tsumego is closer to reviwing pro games than solving tsumego.

If you have just about become totally confused, consider this: I have mentioned it before, but I believe that there are 5 levels of tsumego relative to any player:

1_easy
2_unchallenging
3_medium
4_challenging
5_hard

And 5 levels of tattention to detail when replaying pro games:

1_skimming
2_light
3_medium
4_deep
5_intense

What you may consider a "hard" or "easy" tsumego is up to you. For me, i treat easy probblems as ones that I would get wrong only once every 20-30 problems, if I have a one minute time limit. Likewise, "hard" means I only find the solution once every 20-30 times, with one minute to spare on each one. Medium means 50/50.

As for professional games, scrolling through as fast as you possibly can while "getting the gist" of the game is 1, reading out every possiblity, trying to gain a complete understanding of, excruciating on, each and every move, 5.

Before we step any further, what is it that we want train when we solve Igo puzzles? "Well," you may say, "reading of course". If I where respond by asking, "what part of reading?", you would look at me in a funny way, then, while Igo is obviously too dynamic for me to dream of being able to "classify all the aspects of Go playing" here are some of the thing that come from the top of my head:

READING:
Depth
Width
Accuracy
Speed

DIRECTION OF PLAY:
Valuation of area
Recognitiom of weak groups
Joseki
Understanding of thickness and strength, etc...
Much, much more I can't remember at the moment

Anyway, the above is just an example classifying a fraction of Go. But the point is, that when you conbine different aspects into one, you get a Go player.

There is tsumego, then there is tsumego. Pro ganes and pro games. Hmmm...

Arrrgh! I can't believe I forgot about Mr Dinerctien
3p! I apologise about that! Hmm, all I can say is watch out, I'm coming!

(*sweat* man breakfast is strong *sweat*)

Anyway, thinking about tsumego and pro games, I could only come to this logical conclusion:

level 1 before you go to bed and when you wake up
level 3 in the day
level 5 luxury special


to be continued...

ps: @RBerenguel, hmm, I'm thinking that after being "whooped" in the panda-prelims, I'd do some of my dynamic-tsumego and by next year I could come and claim my GLORY... okay, maybe not, all I want to do us produce a good kifu (never done so in my life)

Elom wrote:Arrrgh! I can't believe I forgot about Mr Dinerctien
3p! I apologise about that! Hmm, all I can say is watch out, I'm coming!

(*sweat* man breakfast is strong *sweat*)

Anyway, thinking about tsumego and pro games, I could only come to this logical conclusion:

level 1 before you go to bed and when you wake up
level 3 in the day
level 5 luxury special


to be continued...

ps: @RBerenguel, hmm, I'm thinking that after being "whooped" in the panda-prelims, I'd do some of my dynamic-tsumego and by next year I could come and claim my GLORY... okay, maybe not, all I want to do us produce a good kifu (never done so in my life)

My own training regime for tsumego used to be (actually, rebooted it yesterday, so "it uses to be") doing mostly categories 1 or 2 from your list. Add some 3-4 or even 5 from time to time, but the meat comes from 1-2 (Lee Chang'Ho's tesuji and tsumego vols 1,2, Cho Elementary.) I sprinkle some 2-3-4 among the Anki studies with problems/situations from the Dictionary of basic tesuji, but these work more as "learn the technique" than as pure reading problems.

Okay, *stretch* I'm finally back with more of an explaination. Ironically, a discussion has started on L19 discussing pro games, subcon v con, but I wonder how much we can know for sure considering how little we know about the brain.

But sub "v" con, if you want to put it that way, plays an important role in learning anything. Amcient cultures have been fascinated by the left/right back/front equilibruim without even knowin, from ancient Egypt to ancient China.

Let's say we had two ahtletes traing to compete in the olympics. If I say, "they're both runners" and you saw one train by constanly running 100 metres multiple times, while the other keeps running 5-10 K every day, am I really telling the whole story?

Elom wrote:Okay, *stretch* I'm finally back with more of an explaination. Ironically, a discussion has started on L19 discussing pro games, subcon v con, but I wonder how much we can know for sure considering how little we know about the brain.

But sub "v" con, if you want to put it that way, plays an important role in learning anything. Amcient cultures have been fascinated by the left/right back/front equilibruim without even knowin, from ancient Egypt to ancient China.

Let's say we had two ahtletes traing to compete in the olympics. If I say, "they're both runners" and you saw one train by constanly running 100 metres multiple times, while the other keeps running 5-10 K every day, am I really telling the whole story?

Nope, long distance runners do sprint sessions from time to time (but most sprinters don't do many resistance sessions, though)

Heh, well of course, any serious athelete in any discipline almost always uses a combination of different types of training, you're correct but I guess my explanation was slightly off, as you have pointed out (thanks ).

Maybe what I was trying to say was that the aspects of running, weather long distance or short, are identical- in theory. But what actually happens is that some elements of running are more or less pronounced or important depending on the distance, to the point where some elements nearly dissapear and are negligable

While this is less extreme in Go, the following points remain true:

For obvious reasons, you need to need to read deep. If you don't know why deep reading is important, let's just say you have a lot to learn . However, breadth is also very important. Looking at many different possiblities, like in the openening, the broad reader can quickly make positonal judgements because reading won't get in the way. Think of it like this: let's say someone is tasked with the challenge of remembering a number sequence-- 1234123512361237. Easy, right? This applies to extremely pure versions of deep reading, like one-way-streets or ladders. There is a pattern. A rule. Less pure and more close to most reading situations is 1734193513361137. Now there is an element if randomness. Here's a sequence half the length-- 07563581. It's more tricky! what just happened there!

Well, it's obvious.

The top pros read deep. The top pros read wide. The top pros read accurately. The top pros read quickly.

When people reccomend doing lots of problems really quickly, it's not to train depth, but speed, accuracy and decisiveness. Espeially useful in a byo-yomi situation where you do not have the luxury to look at every possible move *and* spend as much time as you like while doing so. But what if you're in the title match for the Kisei, and ended up in a really bad position? I'm sorry, but no matter if had all day to think-- I mean literally-- not being able to read deep enough is not being able to read deep enough.

Elom wrote:Heh, well of course, any serious athelete in any discipline almost always uses a combination of different types of training, you're correct but I guess my explanation was slightly off, as you have pointed out (thanks ).

Maybe what I was trying to say was that the aspects of running, weather long distance or short, are identical- in theory. But what actually happens is that some elements of running are more or less pronounced or important depending on the distance, to the point where some elements nearly dissapear and are negligable

While this is less extreme in Go, the following points remain true:

For obvious reasons, you need to need to read deep. If you don't know why deep reading is important, let's just say you have a lot to learn . However, breadth is also very important. Looking at many different possiblities, like in the openening, the broad reader can quickly make pisitonal judge because reading won't get in the way. Think of it like this: let's say someone is tasked with the challenge of remembering a number sequence-- 1234123512361237. Easy, right? This applies to extremely pure versions of deep reading, like one-way-streets or ladders. There is a pattern. A rule. Less pure and more close to most reading situations is 1734193513361137. Now there is an element if randomness. Here's a sequence half the length-- 07563581. It's more tricky! what just happened there!

Well, it's obvious.

The top pros read deep. The top pros read wide. The top pros read accurately. The top pros read quickly.

When people reccomend doing lots of problems really quickly, it's not to train depth, but speed, accuracy and decisiveness. Espeially useful in a byo-yomi situation where you do not have the luxury to look at every possible *and* spend as much time as you like while doing so. But what if you're in the title match for the Kisei, and ended up in a really bad position? I'm sorry, but no matter if had all day to think-- I mean literally-- not being able to read deep enough is not being able to read deep enough.

What I've found is similar... but different. And kind of akin to sprinters/marathoners. A go game is essentially a really long session of very very easy reading exercises, intersped with complicated/very complicated/impossible problems. A relatively good training (in the sense the subjectively, I've felt my gameplay much, much sharper than with any other kind of training) is long sessions of very (or relatively very) easy tsumego. The other aspect is being constant (which also happened in these sessions because they spanned several days.) Hence my current (and former, before I stopped in August, you can read about it in my study log) is doing just (this has been changing and increasing during time, but essentially) Cho elementary L&D problems, day in and day out (using Anki to spread them out in a sane way.) Basically, each day going for a run, no matter if it is long (if I have time I pick any of the many tsumego books I have, or use NGA tsumego assignments), short (just the Anki session.) Of course, this needs tweaking (adding more problems, increasingly more difficult, fighting the utter boredom of doing many "easy" problems for the first days) and the feeling of "boring chore." All are difficult, but are also part of training focus, so not bad per se.

RBerenguel wrote:

Elom wrote:Heh, well of course, any serious athelete in any discipline almost always uses a combination of different types of training, you're correct but I guess my explanation was slightly off, as you have pointed out (thanks ).

Maybe what I was trying to say was that the aspects of running, weather long distance or short, are identical- in theory. But what actually happens is that some elements of running are more or less pronounced or important depending on the distance, to the point where some elements nearly dissapear and are negligable

While this is less extreme in Go, the following points remain true:

For obvious reasons, you need to need to read deep. If you don't know why deep reading is important, let's just say you have a lot to learn . However, breadth is also very important. Looking at many different possiblities, like in the openening, the broad reader can quickly make pisitonal judge because reading won't get in the way. Think of it like this: let's say someone is tasked with the challenge of remembering a number sequence-- 1234123512361237. Easy, right? This applies to extremely pure versions of deep reading, like one-way-streets or ladders. There is a pattern. A rule. Less pure and more close to most reading situations is 1734193513361137. Now there is an element if randomness. Here's a sequence half the length-- 07563581. It's more tricky! what just happened there!

Well, it's obvious.

The top pros read deep. The top pros read wide. The top pros read accurately. The top pros read quickly.

When people reccomend doing lots of problems really quickly, it's not to train depth, but speed, accuracy and decisiveness. Espeially useful in a byo-yomi situation where you do not have the luxury to look at every possible *and* spend as much time as you like while doing so. But what if you're in the title match for the Kisei, and ended up in a really bad position? I'm sorry, but no matter if had all day to think-- I mean literally-- not being able to read deep enough is not being able to read deep enough.

What I've found is similar... but different. And kind of akin to sprinters/marathoners. A go game is essentially a really long session of very very easy reading exercises, intersped with complicated/very complicated/impossible problems. A relatively good training (in the sense the subjectively, I've felt my gameplay much, much sharper than with any other kind of training) is long sessions of very (or relatively very) easy tsumego. The other aspect is being constant (which also happened in these sessions because they spanned several days.) Hence my current (and former, before I stopped in August, you can read about it in my study log) is doing just (this has been changing and increasing during time, but essentially) Cho elementary L&D problems, day in and day out (using Anki to spread them out in a sane way.) Basically, each day going for a run, no matter if it is long (if I have time I pick any of the many tsumego books I have, or use NGA tsumego assignments), short (just the Anki session.) Of course, this needs tweaking (adding more problems, increasingly more difficult, fighting the utter boredom of doing many "easy" problems for the first days) and the feeling of "boring chore." All are difficult, but are also part of training focus, so not bad per se.

What is Anki? I have printed out a small booklet of Cho elementary L&D.. I also have it on my phone in pdf form.

RBerenguel wrote:http://ankisrs.net

I only found intermediate problems.

Where did you find elementary problems?

(I am talking about the finished decks)

Krama wrote:

RBerenguel wrote:http://ankisrs.net

I only found intermediate problems.

Where did you find elementary problems?

(I am talking about the finished decks)

I created the intermediate one from the SGFs. I created the elementary from images, so legally I can't really post them. I could re-generate them from SGFs, but would take me some time to remember how the script worked and also a little more to find the SGFs (I know they are somewhere in my drives, but where?)

RBerenguel wrote:I created the intermediate one from the SGFs. I created the elementary from images, so legally I can't really post them. I could re-generate them from SGFs, but would take me some time to remember how the script worked and also a little more to find the SGFs (I know they are somewhere in my drives, but where?)

What are the legal differences here?

RBerenguel wrote:

Krama wrote:

RBerenguel wrote:http://ankisrs.net

I only found intermediate problems.

Where did you find elementary problems?

(I am talking about the finished decks)

I created the intermediate one from the SGFs. I created the elementary from images, so legally I can't really post them. I could re-generate them from SGFs, but would take me some time to remember how the script worked and also a little more to find the SGFs (I know they are somewhere in my drives, but where?)

I read this and another thread where you discuss Anki. After noticing that it was about $25, I realized that it was overkill since all I would use it for is Go. Then I noticed EasyGo for iOS, for about $12, has built-in repetition training, includes some problems, and can work with any folder of SGF problems. I haven't purchased it yet, but as I already have so many SGF problem sets, I probably will soon. Of course, I would love to hear from anyone who has used EasyGo! EDIT: https://itunes.apple.com/us/app/easygo/id492566615?mt=8