What is the real limit of being self-taught?

General Go Discussion
1

First of all, my question is addressed to experienced players, professionals and coaches, in general, to everyone who has experience in Go and is serious about Go. But this does not mean that everyone cannot speak out here.

So, let’s say there is adult person who started playing Go. This person learned the rules and basic principles of the game, but did not go further in theoretical training. This person does not read textbooks, tutorials, or any specialized literature, does not watch educational videos, does not solve problems, and does not use the services of coaches or more experienced players. This person is simply playing, trying to independently, without the help of computer programs, analyze the games played and extract gaming experience from them, nothing more. Let’s say this person plays 1-3 games a week, depending on his other activities.
I think there are quite a lot of such players, Go lovers.

Now, actually, the questions…

With this training regimen described above, what level of play will such a player reach, say, in 10 years? At what level do such players most often stop? At what level will he need to study theory or hire a coach to further advance?

I understand that there are capable players, naturally gifted, talented, even geniuses. My question is not about such players, but about ordinary, average amateurs.

If anyone has statistics on my questions, I would be grateful for sharing. I also thank everyone in advance for their substantive answers.

2 Likes
2

10 games / day not allowed?

1 Like
3

Frankly speaking, I can’t imagine how an adult who is not professionally involved in Go can play 10 games a day… :face_with_monocle:

1 Like
4

I seen A LOT accounts like that and played like that in the past myself, while watching anime at the same time

2 Likes
5

ACTUALLY using time on review of your own lost games, but NOT using AI in 2024 is a very strange goal.

1 Like
6

Perhaps this goal may seem a little less strange if we assume that in this way a person strives to develop his own thinking skills.

3 Likes
7

I would assume that most players don’t even review, so absolute minimum among average person who also don’t waste much time on Go is 9k

among average people who also play A Lot, it would be at least 5k

1 Like
8

I think you’d be restricting your thinking skills by trying do to everything yourself.

Great ideas don’t come often, and to some people they never come. But you can get practice in where to search for great ideas by learning other people’s great ideas.

Understanding what experienced people have learned offers you far more tools to apply to your own analysis and allow you to think outside of the box. This is what develops your thinking skills, a combination of working by yourself on problems and of experiencing a large quantity of ingenious solutions to problems (that you’d probably never would encounter on your own).

Think about it in terms of other stuff you could learn. Would you become a great composer if you never listened to music from other composers? Would you become a great writer if you never read a book yourself? Would you become a great philosopher if you never knew what other people were thinking?

10 Likes
9

Thank you! Great statistics! :+1: :+1:

10

Thank you. I completely agree with you. But in the case that I described, we are not talking about achieving any heights, but rather about determining the limit of one’s own abilities.

1 Like
11

your hypothetical experiment also needs limit on max rank of your opponents.
else:
you can challenge 9 dans, resign, then remember their opening… it would be like using AI

1 Like
12

I don’t think this works. Even if a strong opponent plays only normal openings and you somehow remember all the answers, the chances are almost zero that the game develops into something you’ve already seen after, say, 50 moves.

Also, AI had to play about 5 million games before it got strong in this way, and a major component in the recent strong AIs is the use of convolutional networks, that could be compared to human “intuition” and allows it to make assessments of positions it has never seen.

1 Like
13

you can create collection of joseki from them. And you can mimic their style partially

14

But there’s more to Go than joseki…

Also, you might lose to weaker opponents who make joseki mistakes, but you have no idea how to punish them.

2 Likes
15

to compare human and AlphaZero, we would need to get humans who never seen Go before, and make them play each other Only.

1 Like
16

I was talking about Leela and AlphaGo, not the Zero’s

1 Like
17

not zero seen all pro games

18

its known fact that current AI needs much more data than human to learn something, its not known how to fix it. But AI on strong computer can run much faster than human. So it can play millions games in 3 days and reach 10 dan.

2 Likes
19

Minus the fact that I’ve read some books and a few tutorial videos, all the other things you mentioned apply to me, so I’d say 4k.

The stipulation that you cannot be reading books or watch videos and tutorials makes that target group very small. It is very unlikely that you love a hobby, you like to spend time on it each week, yet somehow, for some unknown reason, you have no drive to learn anything about the hobby you love, from other people sharing the same hobby. :thinking:

6 Likes
20

I have a friend in his thirties to whom I’ve taught a bit of Go. Just the basics, not even shapes or something similar.
We sometimes play 9x9 at his place; I give him a 5-stone handicap, and it’s more or less a 50/50 chance who wins.

Thing is, he rarely plays, if he plays at all.
Maybe he would improve a bit by reviewing games on his own. (Most of the time, he doesn’t even realize something is amiss), but I don’t think that’s a good way for his level to learn. It’s just try and error without fun and without an opponent.

Problem is, he won’t even do that.
Players like my friend usually stop the moment external motivation disappears—in this case, me. We only play when I have time, and he only plays with me. He says “it’s a nice game, I like it” and that’s that.

There are often other pleasant things to do in life, so why bother with a “nice” but new game that requires a lot of effort to get better?

So I think the training regime you talk about here is completely faulty to begin with, and I can’t really imagine players, who do this.

PS: I think there are a lot of Go lovers that don’t really actively play Go. They still like the game, may be interested in some events, see Go somewhere and tell you “ooh, i know that game. It’s fun”, but may not ever put a stone on a goban again in their life. I know at least 4 of such people

6 Likes