Systematically determining and collecting information I'm missing to improve my gameplay

I have played almost 650 go matches and, while having reached rank 20 kyu before, I am fairly often stuck at 25 kyu; it is time I take a more thorough and effective approach to studying and learning how to play my favorite game. I don’t necessarily like taking a systematic approach, and if it takes away from the enjoyment of learning then I’ll stop. But I’d really like to improve and bring out the potential I know I have. Google searches have yielded some suggestions:

“Clearly identify what information is missing and its specific nature. This includes the ‘research questions’ [I’m assuming in terms of Go these are the areas of study in general such as invasions or the opening phase], objectives [strategy long term and short, tactics], or problems you’re trying to address [weaknesses in my playstyle or how to achieve a general type of goal] and pinpointing the data gaps.”
As well as
“Define objectives: Clearly outline what information you need to collect and why. This helps focus your efforts and ensure the data gathered is relevant.”

Next comes choosing appropriate methods and techniques for gathering data, such as reviewing my old games, reviewing old pro games, asking a higher ranked player, “group survey” (i.e. forum post), or maybe just observing other players.
The information being gathered, i.e. the skills I’m missing and areas of play I’m weak in, needs to be relevant and reliable, and accurate, consistent, and well-organized (I do take good notes at least).

(As a last ditch effort, I can start a new journal and “Clarify What I Know”, which is to document all of my existing information related to Go, which is supposed to help identify precise gaps and guide further efforts.)

Does anybody have any suggestions for any of this?
For example, looking back at a game of mine from earlier in the day, I saw that I am “missing” information on how to handle being invaded. This includes bases, making/preventing eyes, running into the center, etc.
What might be an effective and reliable way to ensure I learn the missing skill/information accurately? Technique or methods such as guidebooks, practice problems, a high ranked tutor? And lastly Google says the information learned has to be organized (my notes), analyzed to extract meaningful insights using appropriate tools, and finally acted upon by interpreting my findings to inform decisions.

I welcome any and all feedback. Thank you for reading, those of you who stuck it out.

2 Likes

Hi - I can really identify with what you’re saying. When I was between 25 - 20 kyu I found it really frustrating that there were very few resources for beginners which tried to answer the questions you’re asking.

It seemed like most of the resources available fell into two categories

  • very basic explanations of the rules / tutorials on how to play, puzzles and exercises that explored one small aspect of of a much larger game
  • resources for people who already understood the game which allowed them to get better

and very little in between. How do people go from taking their first baby steps, to really figuring out all the game mechanics - I mean besides losing your first 1000 games quickly, as the old saying goes

With that, I really tried my best to fill in those blanks, and started the 19x19 FOR BEGINNERS series here on OGS. Rather than trying to comprehensively tackle ALL the information, I tried to explain the big pieces, which all have to be understood and grasped individually, before you can begin to understand how they all interlock together to form a coherent whole.

All in all, it’s probably somewhere around 200+ pages of content (it’s hard to count, because so much of the space is taken up by images)

Introduction: Making Sense of Go

Part 1: Sente and Gote

Part 2: Settling Your Stones

Part 3: Playing a Balanced Opening

Part 4: Joseki Basics

Part 5: The Stages of the Game

Part 6: Shape

Take your time, please make sure to play through all of the included demo games, and enjoy. I hope this addresses some of what you were talking about.

4 Likes

Go wiki

2 Likes

Wow thanks @tonybe, I really appreciate you posting the whole series, and the tip about playing those first 1000 quickly, I forgot about that. The “Foundations of Go” book from Kiseido says not to even start reading until you’ve gotten several games under your belt.

In the mean time in between reading the 19x19 series, I was thinking of looking back on a few of my games and trying to see what my absolute biggest weakness is and working on that.

2 Likes

There’s one more thing I would keep in mind if I were you

If you think about the experiential learning needed to improve in Go as similar to learning a martial art, like kung fu

The differences in fighting a yellow belt, a brown belt, and a black belt aren’t just quantitative (i.e. more hours of practice, a greater number of skills learned) they’re also qualitative - the way someone mentally and strategically approaches a given battle will be very different, what sort of responses they’re able to come up with on the spot, and the types of risk they’ll be willing to take on would be hugely different

The differences between a 20 kyu, a 10 kyu, and a 1 dan will be metaphorically similar

It’s not just that the 1 dan has the reading, shape, and contact fighting knowledge to handle multiple weak groups and come out alive, it’s that they have experienced living through many such situations before, and know how to not get overwhelmed juggling all those confusing priorities

So yeah, as cool as it would be if we could somehow download all the information into our brains Matrix style, and just wake up saying “I know how to play Go”

But when the rubber hits the road, you’ll have to temper the intellectual knowledge of how a given game mechanic should work with the living reality of having to apply those abstractions to real life play with flawed, imperfect, unpredictable opponents

So yeah, take your time and make your way through the skill levels one by one. Good luck

PS - if you’d like someone to take a look at one of your previous games and give you some pointers on specific things to study, please let me know, and I can take a look

4 Likes

@tonybe honestly what I really want to do most is just play, over and over, night and day. And often I do. I suppose I have been worried that I won’t improve just by playing.

Plus, I can’t really remember 90% of what I read anyways. Some of the basic concepts maybe. I do puzzles and used to practice a lot of tesuji; I think because of my inability to absorb book information I should train my intuition to play better on the spot.

In any case, thanks again for the advice, I am taking it to heart

EDIT – @tonybe it would be awesome if you could look over a game for me, I’ll find one and send it to you

4 Likes

Proverbs are so valuable. You can try to understand what they mean, remember the most important, try to follow them, try to understand more, remember the next importand ones … and so on.
That and some tips/ reviews from stronger players gain me the most knowledge as beginner I think.

3 Likes

Yeah, feel free to double down on playing. Maybe play some games against better human opponents (15 - 10 kyu) or a few low ranked AI (like GnuGo) just so you get a better idea of what sorts of things are possible and impossible within the game

Also, if you’re not aware of basic things like what makes a living group, or corners > sides > middle, you’ll just keep making the same mistakes over and over, and it may take months for that knowledge to dawn on you organically

So yeah, maybe a little from column A and a little from column B

But yes reviewing your games is a great way to improve. The OGS AI review is very useful for this, because it often identifies the moves with the biggest win % change, and shows you other options

Yes, happy to look at some games, offer a review

1 Like

@Tschej I like proverbs a lot too, “Divide and Conquer”, advice to live by. I bought the $100 book Essential Go Proverbs and have looked at some of them, maybe 50 or 60 in it. “The Ponnuki is Worth 30 Points” (I think). They are easy to remember

EDIT/ADDITION – @tonybe I do enjoy reading the material, and I must have almost every beginner book on the market, I’ll review one today.

I used to have AI, I am so broke at the moment though, so next month.

I’m in the process of figuring out how to send you a link to one of my games now

3 Likes

Go to your OGS profile

Find the game in your history on your profile

Open the game in your browser and copy paste the URL into the body of a comment in the forum

2 Likes

Game that I lost, opponent was 24k, I was 25 at the time.

2 Likes

Is that an average game of yours?
I liked it very much. I have the feeling I understand (almost) every move of yours - and that alone is a big thing from a 25k game. My respect. If you continue to play like that and improve on the crucial mistakes, I foresee a big jump in rank.
(Now I put you under pressure, so you have to perform hehehe)

3 Likes

The usual advice (though I disagree with how it’s often stated) is about 100 games. 1000 is a lot more. There’s a saying about 1000 games to 1 dan, and that’s about how long it took me. So I don’t think you should blitz your way through the 1000 expecting to still be 20kyu at the end. Indeed I made a conscious choice to slow down and think when I was around 20-15k and it helped me improve.

3 Likes

Thanks - I’ll take a look.

Ok, so I’m really confused - because at this point, there are SO MANY amazingly useful AI tools out there available for absolutely free. You shouldn’t need to spend any money to benefit from these

Here are some that I would recommend:

GnuGo - if you download this free .SGF editor from PandaNet, it comes bundled with a version of the GnuGo AI, which plays somewhere between 14 kyu (at the lowest setting) to around 8 kyu at the highest setting. As a 24-18 kyu player, I found it a very useful learning tool because - unlike the TPK opponents I was playing - it was never very aggressive, but it also never made mistakes in things like shape, counting liberties, knowledge of direction of play, stages of the game. It allows one to request infinite UN-DOs, and try all kinds of things to see how different approaches to a given problem allow for different results.

Katrain - this is like having an infinitely patient 9 dan Go teacher living on your computer (note: you need a fairly good CPU and graphics card, but if you do it can be very useful)

Katagui - if you DON’T have a decent gaming PC at home, here’s a version of KataGo that you can access through a browser. It is very good for reviewing 19x19 games, because you can upload SGF files, go through them move-by-move, and get other suggestions for what you could have played at that point.

At that point, the biggest barrier in actually learning from KataGo / Katrain is that the advice it is trying to give you is too high-level - meaning it will encourage you to try a more risky, fight heavy, capture heavy style than you may be ready for as a 20 kyu player.

As such, one learning tool I’ve employed is playing AI vs AI games where I’ll pit the GnuGo AI against KataGo, and use KataGo as a set of “training wheels” to help me see where the best moves are for each position, learning to see how KataGo approaches a given board position, and trying to extrapolate that into my own games

Lastly, if you don’t want to play these AI on your computer, you can always play them as accounts on OGS:

Here’s the GnuGo user on OGS

There are lots of other bots out there you can challenge, but the majority of them are stronger than 10kyu. Let me know if you’re interested and I can suggest some others.

But yeah, no money needed

2 Likes

OK, last post in this thread, and I’m done for the night - I promise : D

So my favorite Go proverb gets right at the heart of the complexity of the game:

If Black has all four corners, Black should resign - if White has all four corners, Black should resign

I takes a long time to understand the interplay of territory and influence, of the differences between playing aggressively and slyly taking advantage of the underlying efficiencies of the game, of the deep understanding of direction of play and the stages of the game that allows a top player to adapt to any situation - but yeah, this proverb captures it with such a wonderful economy of words (pedant note: even though I quoted the longer version, there’s a very tongue in cheek barb in the tone, and I really like it)

2 Likes

@Tschej I’d say that’s probably an average game of mine these days, I feel the respect and I’m grateful. You’ve got mine too, I hope (and expect) to one day soon have that same skill of understanding the moves on that level.

I thrive on the pressure :smiley:

1 Like

'm going to download all of those AI tools immediately after I read that proverb with focused attention, but of course I gotta look at the game @tonybe reviewed before that. I’m itching to do all of it.

That last paragraph you typed was exactly what I needed to hear. It calmed that part of me that said I am making no progress and I never will and I should give up because I’m terrible.
What a stupid voice to listen to

“Divide and Conquer” - Proverb - Not exactly in this case, but, divide myself from all that anxiety about progress and rank and just play and learn the thing I probably love most in the world :slight_smile:
Let confidence gain more territory in my brain and capture the invading group of… you name it, those mental stones that always bring us down XD

2 Likes

I wrote this on a very similar topic a few weeks ago - I can really relate to that feeling, and it’s interesting how interfacing with the learning curve of Go helped me handle that feeling in other areas of my life

1 Like

I hate to bug you again @tonybe but I’m not sure which link to click on for GnuGo. Its embarassing to have to ask you for help, but I dont want to download something that will mess up my laptop.
WIndows.

Asking for help with this doesn’t feel good…

1 Like

2 Likes