Life In 19x19

zafuri95 wrote:Here's some weekly updates on my progresses:-

1. I've been practicing sometime across the week with OGS's Spectral 4k bot.
2. I occasionally won some games but generally I got punished by the bot.
3. I find that I gave my opponent influence too easily. Even though I get the territory I wanted, due to the massive influences I don't get to fight or reduce them at all. It looks like I need to buckle up and learn how to balance influence and territory.
4. I've been doing tesuji drills and life & death on daily basis.
5. My "Shape Up!" reads doesn't help me too often, I don't come across many situations where I can put them to use. But I keep reading it since I committed and tried to create a collection of shapes that I will encounter most often to refer. Hope it helps.
6. The book that helped me the most was Elementary Series: Attack and Defense. I took a peep at "All About Influence" to fix my worry and anxiety to handle thickness / influence. I may include that book into my study in future.

Conclusion:-
I'm trusting the daily drill of tesuji to help my games, and I wish I calculate more often in games. The time pressure puts me to play on "feeling" too often and I think that will develop really bad habits, I'm trying to shake this off. I'm also learning more on practical theories that I will encounter more often so I can put them to use. Concluding this week's performance, I'm still at the same rank at OGS 10K and no changes despite all the wins and losses. I hope I'm able to see my OGS rank progresses in coming weeks. I've also started an IGS account at 10k and have not played any game yet as OGS is my main server now. I'll continue this weekly reflection post on Sundays for coming weeks.

Weaknesses encountered:-
1. Influence handling
2. Direction of play
3. Josekis (Not corner josekis, they are josekis of reduction, invasion etc.)
4. Lacking fighting spirit (Wanted to settle down too often and choose sub-par moves and put less pressure to opponent, not confident to handle cuts)

I've also posted my sgf for reviews at "Game Analysis" session. AI reviews don't help because I don't understand the "whys". Please point out what I can do better to improve

Studying techniques, L&D, concepts, openings ... is definitely going to improve your game but only in the long run. If you are sensitive to short term (rank) progress, you'll have to consciously integrate aspects of Go which matter most to the outcome. At this level, these are often aspects of gamesmanship (not giving up, managing time, think of alternative moves) and very mundane things like keeping track of liberties of a group (to avoid stupid losses).

Basic techniques will get you a long way. See https://senseis.xmp.net/?BasicInstinct

Knotwilg wrote:

zafuri95 wrote:Here's some weekly updates on my progresses:-

1. I've been practicing sometime across the week with OGS's Spectral 4k bot.
2. I occasionally won some games but generally I got punished by the bot.
3. I find that I gave my opponent influence too easily. Even though I get the territory I wanted, due to the massive influences I don't get to fight or reduce them at all. It looks like I need to buckle up and learn how to balance influence and territory.
4. I've been doing tesuji drills and life & death on daily basis.
5. My "Shape Up!" reads doesn't help me too often, I don't come across many situations where I can put them to use. But I keep reading it since I committed and tried to create a collection of shapes that I will encounter most often to refer. Hope it helps.
6. The book that helped me the most was Elementary Series: Attack and Defense. I took a peep at "All About Influence" to fix my worry and anxiety to handle thickness / influence. I may include that book into my study in future.

Conclusion:-
I'm trusting the daily drill of tesuji to help my games, and I wish I calculate more often in games. The time pressure puts me to play on "feeling" too often and I think that will develop really bad habits, I'm trying to shake this off. I'm also learning more on practical theories that I will encounter more often so I can put them to use. Concluding this week's performance, I'm still at the same rank at OGS 10K and no changes despite all the wins and losses. I hope I'm able to see my OGS rank progresses in coming weeks. I've also started an IGS account at 10k and have not played any game yet as OGS is my main server now. I'll continue this weekly reflection post on Sundays for coming weeks.

Weaknesses encountered:-
1. Influence handling
2. Direction of play
3. Josekis (Not corner josekis, they are josekis of reduction, invasion etc.)
4. Lacking fighting spirit (Wanted to settle down too often and choose sub-par moves and put less pressure to opponent, not confident to handle cuts)

I've also posted my sgf for reviews at "Game Analysis" session. AI reviews don't help because I don't understand the "whys". Please point out what I can do better to improve

Studying techniques, L&D, concepts, openings ... is definitely going to improve your game but only in the long run. If you are sensitive to short term (rank) progress, you'll have to consciously integrate aspects of Go which matter most to the outcome. At this level, these are often aspects of gamesmanship (not giving up, managing time, think of alternative moves) and very mundane things like keeping track of liberties of a group (to avoid stupid losses).

Basic techniques will get you a long way. See https://senseis.xmp.net/?BasicInstinct

I've wondered what the qualitative difference might be between solving 'easy' tsumego (tsumego aimed at a few ranks below yourself) and 'hard' tsumego (tsumego aimed at a few ranks above yourself). My impression is that instinctual training from easy tsumego is long-term improvement plan while gamesmanship is more so about mental fortitude and perhaps overriding bad instincts most kyu players may have? In any case, it seems, according to the reasoning I'm using here, is that the former works best for long-term improvement and the latter for short-term elevations in gamesmanship. Would you say that's true in your experience? I thought of focusing on long term strength from short games in the week and short-term strength from long games on the weekends too.

Elom wrote:I've wondered what the qualitative difference might be between solving 'easy' tsumego (tsumego aimed at a few ranks below yourself) and 'hard' tsumego (tsumego aimed at a few ranks above yourself). My impression is that instinctual training from easy tsumego is long-term improvement plan while gamesmanship is more so about mental fortitude and perhaps overriding bad instincts most kyu players may have?

Knotwilg's gamesmanship qualities aside, IMHO most people who want to improve do themselves a disservice by solving easy tsumego. I used to be a bridge expert. In large tournaments with hundreds of pairs, my typical result was about 5th place. One of my partners told me that of all her partners, I was the quickest to work out what was going on. (Bridge is a game of hidden information.) Maybe so, but I did not develop that skill by solving easy problems quickly. I took my to time to try to solve and thoroughly understand difficult problems. It was that understanding that enabled me to figure things out quickly at the table.

How much time you want to work on a problem is up to you. My feeling for kyu players is anywhere from 30 seconds to 2 minutes. But the problems should be hard enough that you get only around half of the right in that time. OC, you should review those you miss and overlearn them. And it doesn't hurt to review easy problems from time to time. Like one afternoon a year. Solving easy problems is a long term plan for non-improvement. You have to challenge yourself.

How much time you want to work on a problem is up to you. My feeling for kyu players is anywhere from 30 seconds to 2 minutes. But the problems should be hard enough that you get only around half of the right in that time. OC, you should review those you miss and overlearn them. And it doesn't hurt to review easy problems from time to time. Like one afternoon a year. Solving easy problems is a long term plan for non-improvement. You have to challenge yourself.

Yes and in fact I find my progression fastest when I was solving tsumegos slightly harder than what I can cope with. It was exhilarating getting them right. That aside, my local Go club player handed me few questions coming from the Kiseido's book "GO Graded Problems for Dan players: 4 kyu - 1 Dan" and I could actually solve them within the suggested time frame.

My current frustration comes from the disparity between my ability to solve tsumegos and my actual game judgment, evaluation and decision making. In small boards, I played my way to 1 Dan in FoxGo, but that doesn't translate to 19x19 games. But I've learned from a Chinese article that this is actually normal and I should not equate tsumego solving ability to real gameplay ability.

So I concluded that I need to work on the bad habits and the correct mentality to play.

I did a lot of easy tsumegos for a long time and the main downfall I experienced is becoming complacent.

Easily seeing the answers gets you in a feedback loop that there is no need to check the lines to the end or really look for the strongest move (yours and your opponent's).

I ended up with playing "vital points", either not checking beforehand if they would work or just lazily reading out some (wishful) lines.

So, yeah: You have to be challenged to grow, it's a muscle thing : )

PS: Don't think too much of tsumego book titles. The regular Graded Go Problems series is way harder than the proclaimed target audience. The Graded Go Problems For Dan Player series is way below the average (KGS) dan-player. There is an essay in the book Treasure Chest Enigma, where the author hints that especially newspapers in Japan gave easier problems but wrote something like "30 seconds for 1-dan". It gives a cozy feeling to the audience, nothing more.

On the subject of easy tsumego, I've concidentally just started translating a short video by Japanese pro Yanagisawa Satoshi 5p (柳澤理志) on exactly this subject. It's part of a 4 part series of videos, 5 to 10 mins each, focused on how to use tsumego to improve. I intend to translate all of them and will share them on the forum if there's interest, but for now you may want to know that the first (and longest) video is about why you should focus on easy tsumego.

The reasoning is that more complicated tsumego are built on top of the simpler ones. This is an entirely literal statement. The positions found at the beginning of beginner tsumego can also be found 1 or 2 moves in to the solutions of a more advanced ones, and those advanced problems will again be found 1 or 2 moves in to the solutions of yet more advanced tsumego. It is by building this foundation step by step that one learns to solve more advanced problems.

He also stresses that it's important to repeat problems and always read out the answer thoroughly. This applies even when you remember the answer from last time you solved it. I'd also like to mention Kageyama's "Lessons in the Fundamentals of Go", which is a book available in English I think many people here have read*. The very first thing he has people do is read out ladders. Of course we all know the solution to a ladder, but that's not the point. It's about reading practice and fully internalising a common pattern.

So my opinion is that although SoDesuNe's experience of improving more from hard problems because they forced him to read is likely very common, the important difference was not the problems themselves, but actually that he was reading in one instance and not the other. In other words I think people are focusing on the wrong variable.
*I'm still near the beginning but I originally bought it based on recommendations from people here.

Sorry for double posting I need to correct myself. Yanagisawa did not say easy problems he said easy collections. So interpret that how you will. I think the general point reamains the same but the difference is not inconsequential as any collection will naturally contain a range of problems.

I can get behind easy collections ; )

As just a hobby go player I like to feel good when I'm pursuing my dreams. 50% failure rate in problems does not feel good to me. I'm more a 33%- to 25% kind of guy. Also reminds you to stay sharp but lets you feel good about yourself, too.

Splatted wrote:On the subject of easy tsumego, I've concidentally just started translating a short video by Japanese pro Yanagisawa Satoshi 5p (柳澤理志) on exactly this subject. It's part of a 4 part series of videos, 5 to 10 mins each, focused on how to use tsumego to improve. I intend to translate all of them and will share them on the forum if there's interest, but for now you may want to know that the first (and longest) video is about why you should focus on easy tsumego.

Because of copyright issues, I am not sure that posting a full translation here would be considered fair use. But surely summarizing his ideas would be valuable.

The reasoning is that more complicated tsumego are built on top of the simpler ones. This is an entirely literal statement. The positions found at the beginning of beginner tsumego can also be found 1 or 2 moves in to the solutions of a more advanced ones, and those advanced problems will again be found 1 or 2 moves in to the solutions of yet more advanced tsumego. It is by building this foundation step by step that one learns to solve more advanced problems.

This reasoning has been the basis of many textbooks, probably for centuries. But research has shown that it is at best a half truth. Within limits, the order of learning does not matter. You don't have to build knowledge and understanding up logically, step by step. In the case of tsumego, suppose that I can get around half of 3-5 kyu problems right, and most of those I miss are because they are built upon 8-10 kyu problems that I would also miss, or never learned thoroughly. That does not mean that I have to go back and work on 8-10 kyu problems. Overlearning the problems that I miss will do just as well, if not better. Overlearning is a review technique. See https://senseis.xmp.net/?Overlearning

He also stresses that it's important to repeat problems and always read out the answer thoroughly. This applies even when you remember the answer from last time you solved it.

Good point about reading out the solutions to problems you know. (Seeing the solution sequences is OK, too. ) IMO, thoroughness is very important. See https://senseis.xmp.net/?GoProblemsTheFudgeFactor

So my opinion is that although SoDesuNe's experience of improving more from hard problems because they forced him to read is likely very common, the important difference was not the problems themselves, but actually that he was reading in one instance and not the other. In other words I think people are focusing on the wrong variable.

There is a lot to studying and practicing go problems. A lot of people put emphasis on reading. Spaced repetition, not mentioned between us now, is also important, and well discussed on L19. But the emphasis on easy problems is, IMHO, misguided. The 50% rule is based on psychological research dating back decades. The research on overlearning goes back even further.

None of this is to disparage Yanagisawa or his videos. I look forward to your write-ups of them.

At the risk of pointing out the obvious, there's more than one way to do it! Some people learn better from constantly challenging themselves, some benefit more from repetitive drill to reinforce the basics. (And some even get good results by alternating these strategies, one month easy, one month hard or something.)

Unfortunately there's not a lot of controlled scientific studies on long term learning. (In the lab, you can see how much someone improves or retains over a few hours, or even days, but it's much harder to study what happens over a five year period.) So most of the evidence is anecdotal. But I believe that both approaches have about equally good results, for those who have the motivation to stick with it. Of course if the problems are so easy that you get bored and give up, or so hard that you get frustrated and give up, it's not a good long-term plan for you.

I thought this was becoming a derailment so have moved the tsumego discussion here: viewtopic.php?f=15&t=17396

xela wrote:At the risk of pointing out the obvious, there's more than one way to do it! Some people learn better from constantly challenging themselves, some benefit more from repetitive drill to reinforce the basics. (And some even get good results by alternating these strategies, one month easy, one month hard or something.)

Unfortunately there's not a lot of controlled scientific studies on long term learning. (In the lab, you can see how much someone improves or retains over a few hours, or even days, but it's much harder to study what happens over a five year period.) So most of the evidence is anecdotal. But I believe that both approaches have about equally good results, for those who have the motivation to stick with it. Of course if the problems are so easy that you get bored and give up, or so hard that you get frustrated and give up, it's not a good long-term plan for you.

You mention of long term learning and controlled studies reminded me of Bartlett's classic, Remembering ( https://archive.org/details/Bartlett193 ... g/mode/2up ). A few years ago I heard a psychologist mention one of Bartlett's results as something that was newly discovered. "We now know. . . ." By chance, Bartlett ran into one of his original subjects some years later, who agreed to try to recall what he had studied before. He recalled some things and misremembered others. Bartlett also pointed out the difficulty of controlling the conditions of a psychological study, because of individual differences. The same external situation might be viewed and felt differently by each person.

Your point about motivation is important. A study method that leads to burnout or to unrealized expectations is counterproductive.

It is important to realize that the 50% rule does not mean always challenging yourself, in terms of doing problems. The reason is that you combine it with overlearning. For instance, suppose that you do problems for 30 min. each weekday on the subway. You start on Monday and do 15 problems, missing 7 or 8. Then on Tuesday you do 15 more, ditto for Wednesday and Thursday. By Friday you have around 30 problems that you missed to review for overlearning. This takes you a little less than 30 min., but you still missed 10 of them. You still need to overlearn all of them next week, because you have to get the 20 that you got right on Friday one more time, Also, you will have to get the 10 that you got wrong twice at least three more times, and that's if you get them right next time. So next week you start your overlearning on Thursday. As you can see, overlearning can quickly build up so that it takes a lot of your time. maybe even most of it. You are not constantly doing challenging problems.

On overlearning: Many moons ago I bought a book by an amateur pianist, who talked about his experiences and practice. One thing that he did when working on a piece was to practice the parts that he was originally unable to play to the point that they became the strongest part of his performance. He made the analogy to the body healing a broken bone; it makes it extra strong. I liked that idea, and have found it useful, myself.

Speaking of tsumego, I've searched a bit of the Chinese written articles (my mother tongue is Chinese) and I've found an answer coming from Lee See Dol himself.
This article is from China's Sina Sports Web news here

虽然他曾经屹立在巅峰的位置，但围棋对于他来说仍然是一个“不可知的存在”。李九段说，“业余棋手的问题中，最让我难以答复的问题是，‘为什么要下在那里’。那只是根据多年下棋的感觉下到那里，仅此而已。现在问我为什么要下在那里，让我实在不知道应该怎样回答才好。”

　　“围棋，很多时候已经不是二选一，而是三选一。许多时候，只是凭感觉而选择。如果不是简单的棋形，我其实算不了那么深。”

The translation:
"Although Lee was once on the top of the game's competition, to him Go is still a game that is of "unknowable existence". He said: "Among the questions from Go amateurs, the hardest one was "Why should I play here?". To me, the decision was made based on my experience of many years playing Go, that's all. Now even if you ask me again why should I play there, I will still have a hard time explaining why." "Go, in many times is not a choice between 2 options. It could be 3. Many times my decisions were made based on how I felt about them. If they were not simple shapes, I wouldn't be able to read that much deeper."

There's a comment concluded that the difference of strength between players is often on the "intuition" itself. Strong players know and choose the most effective / efficient moves to read, but weaker players have no idea where to start reading and may end up with unproductive reading.

So I believe, the reason for tsumego and Life&Death is to teach shapes, tesujis to improve the "intuition'.

I've taken in some advice on my posts and concluded with these goals, so here's the weekly update. Not much changes on my routine, just a little more focused and I removed too many distractions that does not motivate me to keep doing it. I'm starting to play games against human players instead of playing against bots. I'm looking forward to reach a certain rank progression in coming weeks

GOAL

1. To reach 5k by July **Currently: 10k ~ 8k**

2. To finish the below books by this year (3 rounds for problems; 2 rounds for theories)
THEORIES
a) Shape Up! for a Stylish Baduk
b) Attack & Defense
c) All About Thickness
d) The Breakthrough to Shodan
c) Milton Bradley's Improve Fast in Go
Tesujis / Life & Death
a) Weiqi 1000 Tesujis
b) Weiqi 1000 Life and Deaths
c) Lee ChangHo's Tesuji Collection

zafuri95 wrote:GOAL

1. To reach 5k by July

Just a thought to this goal: Rank measures progress, not skill.

If you like to know whether you got stronger, I'd prefer a goal along the lines of: I can correctly solve ~75% of the problems in Weiqi 1000 Tesuji in under one minute. This you can actually track sensibly.

When you go by rank, there are a lot of variables, which are out of your control: Actual strength of your opponent, state of mind of your opponent, blunders by your opponent, time losses/lag, rank shifts, whether or not you play handicap games and so on.