Are the avalanches still joseki?

bayu · Post by **bayu** » Thu Nov 21, 2019 1:23 pm

Thanks for all the answers and game records!

I am happy they are not completely demoted to trick plays.

Yakago · Post by **Yakago** » Thu Nov 21, 2019 2:37 pm

Also FYI - if you create a situation where 'influency' stuff is happening on the outside and you have the ladder, KataGo will prefer high-approach -> avalanche variation

And then

It will not play the most complicated variations because of the descend to 2nd line that was mentioned earlier. But that's the choice of the player who has the corner.

Bill Spight · Post by **Bill Spight** » Sun Nov 24, 2019 11:23 am

Here is another game where we can regard the small avalanche as joseki.

GoGoD 1930-03-26j, Takahashi Shigeyuki, 3 dan (W) vs. Inoue Ichiro, 3 dan

Click Here To Show Diagram Code: [go]$$Wcm8 Small avalanche blunder $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . 7 . . | $$ | . . . O . . . . . . . . . a X 2 5 . . | $$ | . . . , . . . . . , . . . 4 O 1 X 6 . | $$ | . . . . . . . . . . . . . . . . 3 8 . | $$ | . . . . . . . . . . . . . . b 9 0 . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . , . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . c . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . X , X . . . . , . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

OC, Elf prefers the outside hane at a for

, but Takahashi's play was within the margin of error, IMHO — just barely, losing 4% to par. Somewhat surprisingly,

, the small avalanche hane, is Elf's top choice instead of the descent to 13. The rest of the plays are Elf's top choices until

, a traditional joseki play. Elf prefers the attachment at b. When I was learning go the joseki with

was considered slightly inferior, but the approach to the bottom corner at c would make up for that. Elf, however, thinks that

loses a whopping 19½% to par, which IMHO makes it a blunder.

Click Here To Show Diagram Code: [go]$$Wcm18 Minor error $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . O . . | $$ | . . . O . . . . . . . . . 1 X X O . . | $$ | . . . , . . . . . , . . . X O O X X . | $$ | . . . . . . . . . . . . . . . 2 O X . | $$ | . . . . . . . . . . . . . . . O X . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . , . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . X , X . . . . , . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

Elf also regards

as a minor error, losing 5½% to par. You might find its recommendation interesting.

Click Here To Show Diagram Code: [go]$$Bcm19 Elf's mainline variation for :b19: $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . 4 6 O . . | $$ | . . . O . . . . . . . . . O B B O . . | $$ | . . . , . . . . . , . . 2 X O O X X . | $$ | . . . . . . . . . . . . . 3 . 5 O X . | $$ | . . . . . . . . . . . . . . 1 W X . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . , . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . X , X . . . . , . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

Elf still likes the attachment. The result is that, while Black does not capture the

stone, White is forced to capture the

stones, and Black ends up with sente. Why didn't the humans come up with this?

xela · Post by **xela** » Mon Nov 25, 2019 1:22 am

I always thought that joseki meant something closer to "established pattern" rather than "utterly perfect in every respect". There are GoGoD games from 2019 with the avalanche pattern in it. If some pros are still happy to play it, then I'd say it's still joseki. (Is this a cue for a 30-page thread on translations of the word "joseki", to rival The Great Nakade Debate of 2013?)

For what it's worth, the chess world has had statistics for some decades showing, for example, that the queen's pawn opening has better winning stats than the king's pawn. Yet the king's pawn opening is still very much alive and well in contemporary play, and people are happy to choose a "slightly inferior" move that's a better fit for their playing style or their mood on the day.

Bill Spight · Post by **Bill Spight** » Mon Nov 25, 2019 3:10 am

xela wrote:I always thought that joseki meant something closer to "established pattern" rather than "utterly perfect in every respect". There are GoGoD games from 2019 with the avalanche pattern in it. If some pros are still happy to play it, then I'd say it's still joseki. (Is this a cue for a 30-page thread on translations of the word "joseki", to rival The Great Nakade Debate of 2013?)

For what it's worth, the chess world has had statistics for some decades showing, for example, that the queen's pawn opening has better winning stats than the king's pawn. Yet the king's pawn opening is still very much alive and well in contemporary play, and people are happy to choose a "slightly inferior" move that's a better fit for their playing style or their mood on the day.

Starting in the 1990s Go Seigen wrote books and articles about "21st century go". It is as though today's top bots have given us a glimpse of 24th century go, or even later, if there were no AI for go. Overnight, new joseki have appeared on the scene, and some joseki have become obsolete. Nobody plays the old 3-3 invasion of the 4-4, except in special circumstances, for instance. And the press from the 5-3 vs. the 3-4 has become common, either on its own or after a pincer.

What I am attempting to show is that the small avalanche is still joseki, while the large avalanche has become situational. And while the small avalanche may stick around,

in the above note is not just a small error, but a blunder or close to it, and will surely disappear as joseki. Joseki books (assuming they still exist!) of 10 years from now will look a lot different from joseki books today.

Edit: For some perspective, today the early tengen on the 19x19 has to be considered obsolete, even though it is still sometimes played. When I was learning go people said that with komi Black won a tengen opening only about ⅓ of the time. Elf's assessment agrees, giving the early tengen by Black a loss of around 13%. Do we think that a play that loses 20% by Elf's reckoning is going to survive? I have found a Go Seigen game where he makes a mistake (according to Elf) that leads to a large avalanche. Elf estimates the loss from that play at 12%, close to its rating of early tengen. How does the large avalanche survive that level of loss?

lightvector · Post by **lightvector** » Mon Nov 25, 2019 6:21 am

Well, there's always some chance, however slim, that buried in the resulting fighting or in the future aji of the position is a blind spot that is much bigger than 20%. Bots have had blind spots bigger than that...

Bill Spight · Post by **Bill Spight** » Mon Nov 25, 2019 7:48 am

lightvector wrote:Well, there's always some chance, however slim, that buried in the resulting fighting or in the future aji of the position is a blind spot that is much bigger than 20%. Bots have had blind spots bigger than that...

True. Two human blind spots include making the 4-4 corner opening obsolete in even games for more than 300 years, AFAICT, and making the mini-Chinese obsolete for around 200 years.

IMO not playing the 4-4 was simply prejudice against it as suitable only for handicap play or old-fashioned as a setup stone. Yes, play was more territorial than it had been under the setup stones, but the 5-4 was played, and it is less territorial than the 4-4.

As for the mini-Chinese, it was played in handicap games with an open corner adjacent to one with a 4-4 stone. I don't know why it became obsolete. AFAICT, it was revived after the Chinese opening became popular. A few years ago I chanced across it with a number of examples in one of the ancient go books online.

Applebaps · Post by **Applebaps** » Wed Nov 27, 2019 12:00 pm

In this age of AI, everyone now speaks the language of AI. Reviewers talk about "win rates" and percentages based on what the computer tells them, as if it was their idea. 90% of the posts on this forum are just AI-related.

It's probably for the best, since AI are the strongest players now and we're not really qualified to say what's optimal without their help. But it still kills a lot of the joy, imo.

Personally, I say you should play what you want to play and let your opponent show you why it's wrong, if they can. Disclaimer: I'm DDK

xela · Post by **xela** » Wed Nov 27, 2019 3:26 pm

Don't worry, the novelty will wear off soon enough.

It's as if some new players have just turned up at our go club. They're very strong but inarticulate. So we're all figuring out the best way to interrogate them, how to get them to talk sense.

It's only a few months since we've been able to put superhuman software onto a home PC at low cost, so it's still all very new. Give it a year and it will be just one thread of the discussions, not an obsession.

But please do stick around during that year and keep enjoying the non-AI content! If I've got time after work I'll post a (human) tsumego problem for you: I saw something interesting in a book yesterday. Yes, a printed paper one. They're still out there!

Applebaps · Post by **Applebaps** » Wed Nov 27, 2019 4:04 pm

xela wrote:It's as if some new players have just turned up at our go club. They're very strong but inarticulate. So we're all figuring out the best way to interrogate them, how to get them to talk sense.

I like this analogy a lot!

John Fairbairn · Post by **John Fairbairn** » Fri Dec 13, 2019 11:41 am

Having translated Fujisawa Hideyuki's book on the avalanche I took a special interest in a couple of the comments here and an a related thread. I even made time to write a long article giving chapter and verse for my own claims, but just as I finished I got the infamous Windows blue screen. I haven't the will to do it all again. You'll have to trust me when I say I made my assertions on the basis of five joseki dictionaries, plus commentaries. I'll just give the headlines here.

First, Gomoto said:

nuff said
(Go Seigen and Alpha Go agreed on this move, the matter is closed )

$$B
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . X X . . . |
$$ | . . . , . . . . . , . . . . O O X B . |
$$ | . . . . . . . . . . . . . . . . O . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------
Click Here To Show Diagram Code
[go]$$B $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . X X . . . | $$ | . . . , . . . . . , . . . . O O X B . | $$ | . . . . . . . . . . . . . . . . O . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . , . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . , . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

This was meant to prove that a fairly standard human move, the connection at the 3-3 point, in place of the circled sagari, was wrong.

Various problems with this claim. One is that Go never played it, but he did play the 3-3 point (in 1931). The sagari was first played against Go by Fujisawa Hosai in 1958, and very many top pros have played it, too, well before AlphaGo.

Trying to back the claim that there are things that bots are teaching us that we have to "unlearn", Bill took this up later and said:

Well, now the bots tell you to play below, instead.

G
$$Bc
$$ . . . . . . |
$$ . . b a . . |
$$ . . O X . . |
$$ . . O X . . |
$$ . O X . . . |
$$ . . 1 . . . |
$$ . . . . . . |
$$-------------+
Click Here To Show Diagram Code
[go]$$Bc $$ . . . . . . | $$ . . b a . . | $$ . . O X . . | $$ . . O X . . | $$ . O X . . . | $$ . . 1 . . . | $$ . . . . . . | $$-------------+[/go]

Do you say, Oh, no! I couldn't? Or do you say, Thanks?

Some things wrong with this, too, though not all directly in the quotes above. One claim is that pros say the 3-3 move is joseki, and that's what we have to unlearn. My "chapter and verse" was to prove that pros haven't claimed it was joseki. They use phrases such as "common shape", and what they said specifically was that Black can make this move if he wants to avoid complications such as ladders and kos, and if he does he does not necessarily get a disadvantage. Nothing about equality, It is in fact a fuseki move. There are a couple of josekis that do follow from this starting point. It seems often overlooked that almost all josekis have fuseki moves within them, and not all josekis start on move 1.

But the main mistake is that the bots haven't proved the 3-3 move is either wrong or bad.

It is true that LZ gives the sagari a very slight preference over connection on an otherwise empty board - 45.6% to 44.8% after 18k playouts for the sagari alone, which I've been given to understand in any case lies within the margin of error.

But in the following position, LZ gave the 3-3 connection at A as best at 46.2% (17k playouts) and the sagari didn't get a rating. The second best was the nobi at B and the small avalanche hane at C got a good work-out early on before fading. The play envisaged on the rest of the board seemed to be all in the lower left. Note that White has a stone in the lower right, so the complications of ladders (which Black, in pro opinion, wants to avoid) would generally favour White. LZ seems to be thinking like a pro.

In the following case, however, the sagari A was the top move after 26k playouts (46.0%).

What I found interesting was that the presumed subsequent play was now mostly in the lower right, and the variations shown there looked very ladderacious (may be my imagination as I'm not sure how that ties in with LZ's alleged weakness in ladders).

So what I infer from all that is:

1. The pros have been misrepresented in their opinions, in part, and overall are actually supported by LZ.

2. We don't have to unlearn the 3-3 move yet. We just have to learn to play it for the right reasons (like Go Seigen).

And I can't help but note again that a slew of joseki books and commentaries by humans gave very clear and useful advice using words. What LZ offered could be made sense of, but only by using the words and principles already learnt from the human corpus.

At the very least the matter is not closed.

Knotwilg · Post by **Knotwilg** » Fri Dec 13, 2019 12:34 pm

Like Sorin (on another level OC) I long stopped playing the avalanche. When I approach high and they attach, I'll hane. When I attach and they play into my arms and hane, I'll connect. Case closed. I don't like long corner patterns that dominate a quarter of the board. For similar reasons of complexity and "longness" I don't do early 3-3 but rather approach. So, John, I'm much more conservative in my play than in my study of the game

I'm with Gomoto on the term "joseki" and interpret it as "established corner pattern". Not all established corner patterns are the best choice in each situation and some have gone out of fashion altogether. At my level, the margin of error in choice of corner pattern is irrelevant in the light of the errors I make in the middle game. That doesn't mean the evolution of positional judgment doesn't interest me, on the contrary!

gowan · Post by **gowan** » Fri Dec 13, 2019 2:07 pm

If I recall correctly, the heyday of the avalanche pattertns would have been in the 1950's through the 1970's. Around the time of Rin Kaiho and Ishida Yoshio, I think, many games in the Honinbo title matches involved avalanche patterns.

I wonder whether the long corner patterns that dominate at least a fourth of the board could be seen as a form of kettei uchi? After the large pattern has finished a lot of the shape in that area has been settled.

Bill Spight · Post by **Bill Spight** » Fri Dec 13, 2019 7:09 pm

xela wrote:It's as if some new players have just turned up at our go club. They're very strong but inarticulate. So we're all figuring out the best way to interrogate them, how to get them to talk sense.

Is goboarding legal?

Bill Spight · Post by **Bill Spight** » Fri Dec 13, 2019 7:29 pm

John Fairbairn wrote:But the main mistake is that the bots haven't proved the 3-3 move is either wrong or bad.

Not guilty, your honor! I think you are reading too much into brief comments.

That said, your post is excellent!

At the very least the matter is not closed.

Indeed not.

However, as I explore the Elf commentaries, I don't remember finding one where the 4th line push that creates the large avalanche is a good play. I think there is a Go Seigen game where the 4th line stones are already there and then the hane is played against them. So I'm pretty sure that the large avalanche is, at best, situational, rather than joseki. The earliest I might be able to write an article on the large avalanche would be next fall, if anybody cares.

Life In 19x19

Are the avalanches still joseki?

Re: Are the avalanches still joseki?

Re: Are the avalanches still joseki?

Re: Are the avalanches still joseki?

Re: Are the avalanches still joseki?

Re: Are the avalanches still joseki?

Re: Are the avalanches still joseki?

Re: Are the avalanches still joseki?

Re: Are the avalanches still joseki?

Re: Are the avalanches still joseki?

Re: Are the avalanches still joseki?

Re: Are the avalanches still joseki?

Re: Are the avalanches still joseki?

Re: Are the avalanches still joseki?

Re: Are the avalanches still joseki?

Re: Are the avalanches still joseki?