It is currently Wed Dec 02, 2020 9:26 pm

All times are UTC - 8 hours [ DST ]




Post new topic Reply to topic  [ 113 posts ]  Go to page Previous  1, 2, 3, 4, 5, 6  Next
Author Message
Offline
 Post subject: Re: reviewing SL articles using LZ and criticism
Post #61 Posted: Fri Dec 06, 2019 6:40 am 
Judan

Posts: 6621
Location: Cambridge, UK
Liked others: 417
Was liked: 3638
Rank: UK 4 dan
KGS: Uberdude 4d
OGS: Uberdude 7d
TelegraphGo wrote:

If there's more than 2 particularly valuable areas on the board, then playing with an eye towards all of them resolving is beyond my thought process, any thought process I can project onto AI, and any thought process I've ever heard. It's not (prove me wrong if you can) humanly possible to perfectly project that far, because your opponent is practically guaranteed an alternate move somewhere. Perhaps friends who have played each other many times before can think like this - but probes are a much better way for someone who knows exactly how their opponent will play to take advantage.

So why bother with the principle? I can't think of a single game I've seen where the only way to convince myself of a move was related to this at all. It's another way to think, but not a very powerful one, as far as I can tell. So why bother?


If you have sente then if there an odd number of roughly equal sized large gote moves then you get the last big move (tedomari) which is good. If there are an even number then opponent gets tedomari which is good for him. I'm just a 4 dan amateur but I do think about these things, though obviously it's not a perfect projection because the opponent has many choices and ways to resist. Once you are in the position of having an odd number of large moves then it's not particularly hard to play them out and get the last one, but the real skill comes in engineering the game to reach a position with the odd number so that you can get tedomari. In my games this often comes down to luck, but I think it IS possible to try to make this happen and I think that's the How John was getting at. I do remember reviewing some Lee Changho games a while ago and being struck how often he seemed to get the last big move of the opening. Was he just always the lucky one? I doubt it. Being consistently lucky is also known as skill :) . I didn't have any commentaries to see Lee's or others pros thoughts, but my impression was he certainly planned the flow of the opening with a view to making himself get the last big move. (There's a similar issue with dame parity in Chinese rules, is it just luck or can you engineer the parity to your advantage?)

To give a concrete example from my recent British championship match. I had just finished the joseki at the upper left and white approached at 18. I thought instead he maybe should have played at a, which was my fear in taking the top left corner after the pincer, but I decided corner was still good given that side was still awkward shape for him. So what to do? My extension at b is obviously a big move, stopping his extension there, developing my corner and protecting the 3rd line weak point of my 4th line shimari, and whilst not sente does have a nice followup of activating c14 which can become an attacking territory trasher. Back-off at c is also a fine move, prevents his double approach and also lessens any of his potential on right side. But I have learnt from bots that double approaches are not so scary, and also given he also has a stone at k3 already that can become inefficient when he double approaches (this is why approaches are now favoured over splitting moves, if he had only the approach stone and I ignored then double approach would be better followup than extend to k3) I could allow that. The top right corner is also a big place, though I didn't fancy approaching from the top as my f17 stone makes that side boring, 3-3 didn't appeal both for him forcing me to boring top side and giving thickness so probably approach on right side, but that felt smaller than left and lower side. So there are 3 main areas, though it's not clear if they are gote or sente.
Click Here To Show Diagram Code
[go]$$Bcm17 Andrew Simons (black) vs Andrew Kay
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . X X X . 1 . . . . . . . . . . . . . |
$$ | . X O O O . . . . , . . . . . O . . . |
$$ | . O O . . . . . . . . . . . . . . . . |
$$ | . . X . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . O . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . b . . . . . . . . . . . . . . . . |
$$ | . . a . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . X . . . . . . . . . . . . c . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . X . . . |
$$ | . . . X . . . . . O . . . 2 . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]


I decided that the e14 jump was a good way to exploit the white bad shape at c15 (this is why bots like to crawl not connect, but it doesn't work well with the 2-space high pincer AK loves) and could, in sente, give white some troubles to make my move at b more sente (I was pleased to see LZ agrees with me). AK followed a fairly standard sequence I anticipated (though 20-21 exchange not so common), which to my understanding used to be considered satisfactory for white to connect up, but my new bot-inspired thoughts is white is forced to make a bad shape connection on 2nd line and whilst the black stones could be considered heavy and bad exchanges they can be sacrificed or saved and used to attack white later, and crucially makes my c9 extension more severe.
Click Here To Show Diagram Code
[go]$$Bcm17 game contd
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . X X X . X 5 . . . . . . . . . . . . |
$$ | . X O O O 4 . . . , . . . . . O . . . |
$$ | . O O . . . . . . . . . . . . . . . . |
$$ | . . X 7 3 . . . . . . . . . . . . . . |
$$ | . 8 . 6 . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . O . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . X . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . X . . . |
$$ | . . . X . . . . . O . . . O . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]


But if I c9 now, I feared tenuki and double approach giving him something more urgent to do, so I wanted to settle the lower right in sente before c9. r6 back-off is obviously gote (though LZ says fine), so my plan was to use the m3 invasion as a sacrifice to resolve the lower right corner to a more comfortable position than allowing double approach in sente, then play c9 in semi-sente: if he answers I get tedomari at top right, if he plays top right I pull out the 3 stones to follow up c9 and am happy too. For example the diagram below would be going according to plan (32 is not a particularly good move that I would expect, I also considered d9 and g16 as more kiai, but illustrates my tedomari plan). As a bonus here m6 is a nice kikashi that could be a ladder maker for fighting from top left.
Click Here To Show Diagram Code
[go]$$Bcm25 tedomari plan
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . X X X . X X . . . . . . . . . . . . |
$$ | . X O O O O . . . , . . . . . O . . . |
$$ | . O O . . . . . . . . . . . . . . . . |
$$ | . . X X X . 8 . . . . . . . . . 9 . . |
$$ | . O . O . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . O . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . 7 . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . X . . . . . . . a . . . . 3 . . |
$$ | . . . . . . . . . . . . . 2 . . . . . |
$$ | . . . , . . . . . , 4 b . . . X . . . |
$$ | . . . X . . . . . O . 1 . O 5 . . . . |
$$ | . . . . . . . . . . . . . 6 . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]


However, if I play 27 low white may not be so obedient as to 28 kosumi but slide at q2. If I 3-3 then he gets a much more comfortable shape after L4 (even though I get sente) so I want to resist, he plays c3 and then I'm in a bit of a quandry, if I play r9 then he can k4 next, but if I k4 he attacks me on the right (miai and he already got corner profit).
Click Here To Show Diagram Code
[go]$$Bcm25 tedomari plan, white resist
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . X X X . X X . . . . . . . . . . . . |
$$ | . X O O O O . . . , . . . . . O . . . |
$$ | . O O . . . . . . . . . . . . . . . . |
$$ | . . X X X . . . . . . . . . . . . . . |
$$ | . O . O . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . O . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . 7 . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . X . . . . . . . . . . . . 3 . . |
$$ | . . . . . . . . . . . . . 2 . . . . . |
$$ | . . . , . . . . . , 5 . . . . X . . . |
$$ | . . . X . . . . . O . 1 . O . . 6 . . |
$$ | . . . . . . . . . . . . . . . 4 . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]


So I decided to play high on the corner (also considered 2-space for funzies but decided against, LZ says very slightly better) as then it's easier for me to ignore the slide (he has less severe attach on right after he takes 3-3 next) and also rather than saving m3 with kosumi I jump and that creates a follow-up at o6. This is what actually happened in the game. Trouble is I lost sente, but I have created new miai: c8 on the left and now r14 approach is better than before to develop right side for me, plus k3 is looking lonely.
Click Here To Show Diagram Code
[go]$$Bcm25 game contd
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . X X X . X X . . . . . . . . . . . . |
$$ | . X O O O O . . . , . . . . . O . . . |
$$ | . O O . . . . . . . . . . . . . . . . |
$$ | . . X X X . . . . . . . . . . . . . . |
$$ | . O . O . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . O . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . X . . . . . . . . . 7 . 3 . . . |
$$ | . . . . . . . . . . . 5 . 2 . . . . . |
$$ | . . . , . . . . . , . . . . . X . . . |
$$ | . . . X . . . . . O . 1 . O . . 6 . . |
$$ | . . . . . . . . . . . . . . . 4 . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]


AK's next move was a big surprise and I thought it was bad (LZ agrees worst move of game so far), the m3 attachment. I guess he was worried about some attack on k3 but I didn't see a good move for me there: if I play around m2 myself he can just extend to g3, and if I play from the side THEN he can m2, and my counter tesuji at n4 isn't great. So I just tenukid, My tedomari plan was now to approach on the right, if he answers o17 then I c9, if he answers that then I continue on the right with AI style r16 or old-style q11 and that is effectively opening tedomari so my plan since e14 jump has been successful. If he ignores c9 to play upper right then I'm happy to active the 3 stones and harass his wall. And if he pincers approach then I'm fine with right side fighting.
Click Here To Show Diagram Code
[go]$$Wcm32 game contd
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . X X X . X X . . . . . . . . . . . . |
$$ | . X O O O O . . . , . . . . . O . . . |
$$ | . O O . . . . . . . . . . . . . . . . |
$$ | . . X X X . . . . . . . . . . . . . . |
$$ | . O . O . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . O . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . X . . . . . . . . . X . X . . . |
$$ | . . . . . . . . . . . X . O . . . . . |
$$ | . . . , . . . . . , . . . . . X . . . |
$$ | . . . X . . . . . O . X . O . . O . . |
$$ | . . . . . . . . . . . 1 . . . O . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]


So everything was going well (and LZ agrees I've taken a lead) and according to the generate tedomari plan, until he played a bad move at 36 which tempted me off plan and into trouble. 36 is a rather weird shape, leaving a cutting point and not having a strong follow-up on top as g16 would. My instinct was to leave this as his problem for later (because even if he spends one move around g12 isn't not an aji-free capture) and just tenuki to a/b for tedomari. This was the sort of plan I had been reading back on e14 jump for people in kibitz wondering why I played slowly then. But then I made my first serious misjudgement and thought I could try to get b11 in sente for b12/c13 and that would be a good exchange, but it's not.
Click Here To Show Diagram Code
[go]$$Bcm33 game contd
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . X X X . X X . . . . . . 2 . . . . . |
$$ | . X O O O O . . . , . . . . . O a . . |
$$ | . O O . . . . 4 . . . . . . . . . . . |
$$ | . . X X X . . . . . . . . . . . 1 . . |
$$ | . O . O . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . O . . . . . . . . . . . b . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . 3 . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . X . . . . . . . . . X . X . . . |
$$ | . . . . . . . . . . . X . O . . . . . |
$$ | . . . , . . . . . , . . . . . X . . . |
$$ | . . . X . . . . . O . X . O . . O . . |
$$ | . . . . . . . . . . . O . . . O . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]


So I decided to do that and then pulled out the stones. I'm annoyed with myself because even at the time I knew this could ruin my good opening with unnecessary early fighting; I should have just stuck to the tedomari plan. It's hard to keep a clear head.
Click Here To Show Diagram Code
[go]$$Bcm37 game contd, tedomari plan ruined
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . X X X . X X . . . . . . O . . . . . |
$$ | . X O O O O . . . , . . . . . O . . . |
$$ | . O O . . . . O . . . . . . . . . . . |
$$ | 2 1 X X X . 5 . . . . . . . . . X . . |
$$ | . O . O . . . . . . . . . . . . . . . |
$$ | . 4 . . . . . . . . . . . . . . . . . |
$$ | . 3 . O . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . , . . . |
$$ | . . X . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . X . . . . . . . . . X . X . . . |
$$ | . . . . . . . . . . . X . O . . . . . |
$$ | . . . , . . . . . , . . . . . X . . . |
$$ | . . . X . . . . . O . X . O . . O . . |
$$ | . . . . . . . . . . . O . . . O . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]


Edit: fixed quote


This post by Uberdude was liked by 3 people: Bill Spight, Gomoto, Knotwilg
Top
 Profile  
 
Offline
 Post subject: Re: reviewing SL articles using LZ and criticism
Post #62 Posted: Fri Dec 06, 2019 7:02 am 
Honinbo

Posts: 10554
Liked others: 3501
Was liked: 3313
Knotwilg wrote:
I have performed a master edit and a rename of the SL page leading to this discussion. Please check if you find this rendition satisfactory.

https://senseis.xmp.net/?TedomariExercise1

John Fairbairn wrote:
I think the result is disastrous...the LZ move does NOT answer the question posed as the theme of the page: How to get the last big point. LZ is just answering the question: what is the best move in this position?


Well, neither does the original content. Nothing I have seen answers that question. "If you see the last large place to play in the opening, play there," is about as close as we get, right? And the original content doesn't even do that, since it's recommended play leaves the right side wide open! :shock:

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

My two main guides in life:
My mother and my wife. :)

Everything with love. Stay safe.

Top
 Profile  
 
Offline
 Post subject: Re: reviewing SL articles using LZ and criticism
Post #63 Posted: Fri Dec 06, 2019 7:04 am 
Honinbo

Posts: 10554
Liked others: 3501
Was liked: 3313
:blackeye:

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

My two main guides in life:
My mother and my wife. :)

Everything with love. Stay safe.


Last edited by Bill Spight on Tue Dec 10, 2019 1:39 pm, edited 2 times in total.
Top
 Profile  
 
Offline
 Post subject: Re: reviewing SL articles using LZ and criticism
Post #64 Posted: Fri Dec 06, 2019 7:43 am 
Honinbo

Posts: 10554
Liked others: 3501
Was liked: 3313
TelegraphGo wrote:
If there's one obviously extra valuable area of the board, we don't need to tell ourselves that the temperature is about to decrease to convince ourselves that it's the most efficient area to play. It's the place to play by definition.


If only things were that obvious. The proverb applies in the more usual case in the opening, where there are a number of reasonable candidate plays. There are many examples from komi games of the 1950s and later where Takagawa, who knows an oba when he sees one, made a mistake by playing the textbook oba, according to Elf.

The fact that there are several reasonable alternatives is a big clue that the temperature is not about to drop any time soon.

----

OT edit: Latin scholars will note my use of the historical present. :lol:

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

My two main guides in life:
My mother and my wife. :)

Everything with love. Stay safe.

Top
 Profile  
 
Offline
 Post subject: Re: reviewing SL articles using LZ and criticism
Post #65 Posted: Fri Dec 06, 2019 7:57 am 
Gosei
User avatar

Posts: 1731
Location: Ghent, Belgium
Liked others: 268
Was liked: 798
Rank: KGS 2d OGS 1d Fox 4d
KGS: Artevelde
OGS: Knotwilg
Online playing schedule: UTC 18:00 - 22:00
I created another thread to discuss the particular "proverb"/advice to "take (the last) big point(s) in the opening". It's wooly :)

I'll look into Uberdude's similar attempt here.

Top
 Profile  
 
Offline
 Post subject: Re: reviewing SL articles using LZ and criticism
Post #66 Posted: Fri Dec 06, 2019 8:24 am 
Honinbo

Posts: 10554
Liked others: 3501
Was liked: 3313
Uberdude wrote:
If you have sente then if there an odd number of roughly equal sized large gote moves then you get the last big move (tedomari) which is good. If there are an even number then opponent gets tedomari which is good for him.


One thing I have picked up from the ELf commentaries — I could be wrong — is this:

If there is one large gote move, then you can get it;
If there are two roughly equal large gote, then they are miai, and you expect that you get one and your opponent gets one;
If there are three such moves, you can probably get the last one;
If there are five such moves, you have a good chance to get the last one. ;)

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

My two main guides in life:
My mother and my wife. :)

Everything with love. Stay safe.

Top
 Profile  
 
Offline
 Post subject: Re: reviewing SL articles using LZ and criticism
Post #67 Posted: Fri Dec 06, 2019 12:45 pm 
Tengen

Posts: 4337
Location: North Carolina
Liked others: 477
Was liked: 718
Rank: AGA 3k
GD Posts: 65
OGS: Hyperpape 4k
I've had a hard time wrapping my head around what people were even arguing about. lightvector take seems so correct it seems almost banal.

I think a useful question is: "if we didn't have this page, would we create it and illustrate it knowing that the only engine we've tested thinks the page's 'good' sequences are bad?"

I doubt it. But once we have the page, something inside us resists removing or qualifying the example.

_________________
Occupy Babel!


This post by hyperpape was liked by: bernds
Top
 Profile  
 
Offline
 Post subject: Re: reviewing SL articles using LZ and criticism
Post #68 Posted: Sat Dec 07, 2019 9:51 pm 
Lives in gote

Posts: 430
Location: Adelaide, South Australia
Liked others: 173
Was liked: 224
Rank: Australian 2 dan
GD Posts: 200
hyperpape wrote:
But once we have the page, something inside us resists removing or qualifying the example.

Qualifying: no problem, in this case we've actually done that. But removing? Sensei's Library is in part a library. Real world libraries keep the old books as well as the latest and greatest. It's good to have a record of what people used to think and how much we have (or haven't) learned since.

Top
 Profile  
 
Offline
 Post subject: Re: reviewing SL articles using LZ and criticism
Post #69 Posted: Sun Dec 08, 2019 3:22 am 
Lives with ko

Posts: 131
Liked others: 1
Was liked: 18
Rank: AGA 4 dan
Universal go server handle: telegraphgo
I had a better post but deleted it accidentally and don't have the time to go as in-depth. :(

In response to Bill's "if only Go were that obvious", occasionally Go is that easy. Sometimes one area really is just bigger than the others - taking the last empty corner when the game started with three 3-3 openings, for example. I think my point was the same as yours, though, that the principle doesn't work well when there's only one reasonable spot left. The case where there are many remaining positions on the board is of more use/interest.

To Uberdude - I love the commentary of your thoughts during the game. I learned something about miai and sente that I hadn't realized. But was it really that important to your thought process that there were three big areas? My impression was that you were trying to make good miai as much as you can, and not that you were trying to take gote at a specific timing where the temperature decreased. Suppose after your ideal tedomari variations both players suddenly noticed that there was a fifth empty corner somewhere, and all your thoughts about how many big places left had reversed parity. I feel like the positions where you aimed to get tedomari would still would be good for you, because in getting to that spot you got yourself the best exchanges you could by making W's moves difficult, and you can still choose to try to take sente or probe whether you can create another big point elsewhere.
Basically, I think that Uberdude's control of initiative is what pulled him ahead in the opening, and maybe not so much the musing on getting the last big move. Perhaps I'm being too rigid, stubborn, set in my ways of not considering temperature decreases until they're very close to happening. In modern games, though, it just seems so superbly likely that someone will disrupt the flow (with either surprisingly bad or surprisingly good moves) that temperature analysis from 3 areas deep seems superfluous.

Top
 Profile  
 
Offline
 Post subject: Re: reviewing SL articles using LZ and criticism
Post #70 Posted: Tue Dec 10, 2019 10:59 am 
Dies in gote
User avatar

Posts: 64
Liked others: 129
Was liked: 26
Rank: DDK Life
Universal go server handle: Applebaps
Online playing schedule: Evenings M-F UTC-9, Saturday mornings
Knotwilg wrote:
This criticism surprises me because it suggests we should continue to rely on pro advice even if it is refuted today by LZ, because we are allegedly unable to understand LZ (or bots in general). I would say that we have equal tentative understanding of Dosaku, Shusaku, Go Seigen or AlphaGo, while on the other hand we aren't totally void of Go insights, so we can do the interpretation of either with varying degrees of confidence. Although the shin fuseki was annotated by Go Seigen, still a large part of our understanding of it comes through our own mental process, as it should, and therefore, we should not refrain of making similar attempts to understand unannotated bot analysis.


Granted, this is me speaking from my teens-kyu POV, but while I do agree that studying bots can yield fruits to help us grow stronger, I just don't think it's that useful for a lot of us down here in the DDK range, or even SDK range. I don't think the reasoning behind this critique is well-phrased, but I do ultimately agree with their conclusion and sentiment.

Bots are incredibly strong, but so is Honinbo Shuusaku, so is Lee Chang-ho, so is Takemiya Masaki, so is Kageyama Toshiro. To zero in on Takemiya in particular, I've had more than a few strong players tell me not to emulate the cosmic style, or even to build moyos in general, because "it doesn't work anymore, AI has disproved it."

Here's the thing, though: I'm a mid-teens-kyu amateur, not a 9p lol. If I'm looking to improve in the here and now, there are many many things I can do to get a little stronger (or a lot stronger!), just focusing on improving my fundamentals. Little-big things like direction of play, rudimentary reading of fights, learning when to use and when to deviate from joseki, life and death, there's a lot.

I *could* spend my time listening to AI and trying to interpret it, or I could just continue to study "outdated" or "disproven" styles of play and particular kinds of moves etc, to get stronger, and then worry about AI stuff when I get on that level.

IMO there's nothing wrong with studying from a teacher whose strength you can't personally emulate, just like there's nothing wrong with studying from a teacher whose techniques have been "disproven". The more data we have, the better. The problem comes in when people emulate without understanding and get themselves into hot water in their games; that's true whether your sensei is Leela or Chikun. When stones hit wood, do you know WHY you're jumping into the 3-3 early? Or are you just doing it because that's what Leela does? Do you know WHY you're playing the sanrensei? Or are you just doing it because Takemiya does (or rather, because he used to)? As human beings, developing the strength to follow up on those ideas requires us to understand them at a deeper level than just blind emulation.

For instance, when you jump into the 3-3 early, yes you take territory, but you give your opponent big influence. Can you handle that influence? Are you strong enough to deal with the consequences? If not, then don't jump into the 3-3. If you are, okay go ahead and bleep bloop. It's the same reasoning any player should use when examining any joseki, imo.

The "disproven" advice and ideas of the old masters (and current pros) can still be incredibly useful for those of us who aren't at the bleeding edge of dan-level technique and understanding, basically. No human on earth is qualified to say what a "good go move" is when bots' ELO ratings are five times as high as our strongest pros, lmao. But does that mean we shouldn't put stones to wood? Hell no. There's great value in the struggle, even a futile one.

Also, a lot of "outdated" techniques still work just fine against human opponents. I know from firsthand experience that psychologically-motivated moves/styles often render an opponent weaker than they otherwise would be. It won't work on a bot, but I'm not playing bots, so who cares?

_________________
Joseki (n): 1. Japanese term meaning "when Jo lives in seki."

Top
 Profile  
 
Offline
 Post subject: Re: reviewing SL articles using LZ and criticism
Post #71 Posted: Tue Dec 10, 2019 1:54 pm 
Honinbo

Posts: 10554
Liked others: 3501
Was liked: 3313
Applebaps wrote:
When stones hit wood, do you know WHY you're jumping into the 3-3 early? Or are you just doing it because that's what Leela does? Do you know WHY you're playing the sanrensei? Or are you just doing it because Takemiya does (or rather, because he used to)? As human beings, developing the strength to follow up on those ideas requires us to understand them at a deeper level than just blind emulation.

For instance, when you jump into the 3-3 early, yes you take territory, but you give your opponent big influence. Can you handle that influence? Are you strong enough to deal with the consequences? If not, then don't jump into the 3-3. If you are, okay go ahead and bleep bloop. It's the same reasoning any player should use when examining any joseki, imo.


If you only make plays you understand, you will consistently play beneath yourself.

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

My two main guides in life:
My mother and my wife. :)

Everything with love. Stay safe.


This post by Bill Spight was liked by: dfan
Top
 Profile  
 
Offline
 Post subject: Re: reviewing SL articles using LZ and criticism
Post #72 Posted: Tue Dec 10, 2019 3:44 pm 
Dies in gote
User avatar

Posts: 64
Liked others: 129
Was liked: 26
Rank: DDK Life
Universal go server handle: Applebaps
Online playing schedule: Evenings M-F UTC-9, Saturday mornings
That's... a strange way of looking at it, haha. What does that mean?

_________________
Joseki (n): 1. Japanese term meaning "when Jo lives in seki."

Top
 Profile  
 
Offline
 Post subject: Re: reviewing SL articles using LZ and criticism
Post #73 Posted: Tue Dec 10, 2019 3:58 pm 
Gosei

Posts: 1634
Location: Earth
Liked others: 609
Was liked: 279
intuition > consciousness

Top
 Profile  
 
Offline
 Post subject: Re: reviewing SL articles using LZ and criticism
Post #74 Posted: Wed Dec 11, 2019 12:13 am 
Lives in gote

Posts: 430
Location: Adelaide, South Australia
Liked others: 173
Was liked: 224
Rank: Australian 2 dan
GD Posts: 200
Applebaps wrote:
Granted, this is me speaking from my teens-kyu POV, but while I do agree that studying bots can yield fruits to help us grow stronger, I just don't think it's that useful for a lot of us down here in the DDK range, or even SDK range. I don't think the reasoning behind this critique is well-phrased, but I do ultimately agree with their conclusion and sentiment.

I agree with about half of what you've said here...

Applebaps wrote:
Bots are incredibly strong, but so is Honinbo Shuusaku, so is Lee Chang-ho, so is Takemiya Masaki, so is Kageyama Toshiro. To zero in on Takemiya in particular, I've had more than a few strong players tell me not to emulate the cosmic style, or even to build moyos in general, because "it doesn't work anymore, AI has disproved it."

Those strong players are giving you the right advice for the wrong reasons. People were telling kyu players not to try cosmic style even last century, long before AIs. The issue is that when we amateurs look at a Takemiya game, often the only thing we really notice is the ginormous moyo. We don't pick up on how the moyo is built "actively", with double-purpose moves where each black stone not only contributes to the framework, but also puts some pressure on white, so that the moyo grows organically out of the position. So we try to play our own "cosmic styles", but do it single-mindedly, using passive moves. It works against weaker opponents, but we're perpetually surprised when we meet stronger players and our moyos just dissolve.

There's a saying -- I can't remember the source -- along the lines of "to play go, you need to look in two directions at once". It's about playing efficiently, getting the maximum amount of work out of each stone. It's easy to forget this if you obsess to much over cosmic style. No problem trying it once in a while, but if you try to play that way in every single game you won't learn as much. More on that below...

Applebaps wrote:
Here's the thing, though: I'm a mid-teens-kyu amateur, not a 9p lol. If I'm looking to improve in the here and now, there are many many things I can do to get a little stronger (or a lot stronger!), just focusing on improving my fundamentals. Little-big things like direction of play, rudimentary reading of fights, learning when to use and when to deviate from joseki, life and death, there's a lot.

Absolutely. And it's really nice to have AI on tap so that if there isn't a stronger player in the room at the moment, you can still get a quick check on those topics.

Applebaps wrote:
For instance, when you jump into the 3-3 early, yes you take territory, but you give your opponent big influence. Can you handle that influence? Are you strong enough to deal with the consequences? If not, then don't jump into the 3-3. If you are, okay go ahead and bleep bloop. It's the same reasoning any player should use when examining any joseki, imo.

Yep, well said.

Here's another example. Many years ago I learned this sequence from 38 Basic Joseki:
Click Here To Show Diagram Code
[go]$$Bc
$$ . . . . . . |
$$ . . b a . . |
$$ . . O X . . |
$$ . . O X . . |
$$ . O X 1 . . |
$$ . . . . . . |
$$ . . . . . . |
$$-------------+[/go]

I already knew that black ''a'' or ''b'' instead of :b1: were the "real" joseki moves. But :b1: is a way of staying away from the complications of the avalanche. I remember the first time a dan player tried to "avalanche" me in a handicap game, and how disappointed he was when I didn't play along! (But I still spent time studying some variations of the avalanche, even if I wasn't brave enough yet to try them in a serious game.)

Applebaps wrote:
Also, a lot of "outdated" techniques still work just fine against human opponents. I know from firsthand experience that psychologically-motivated moves/styles often render an opponent weaker than they otherwise would be. It won't work on a bot, but I'm not playing bots, so who cares?

Now here's where we depart ways. Yes, some club players do things that they know aren't quite right, but it works against amateur opponents. And if you fight fire with fire, you might be able to match the other players in your club. But you'll never surpass them until you learn to stop setting fire to things! If instead you learn to overcome your opponents' trick plays with honest moves of your own, then you're walking a longer road (you won't catch up quite as quickly), but you're walking a road that doesn't lead to a dead end. I suspect you're hanging around these forums because you don't want to just reach a comfortable level and then stop learning, you want see how much of the deep potential of this game you can explore.


This post by xela was liked by: Applebaps
Top
 Profile  
 
Offline
 Post subject: Re: reviewing SL articles using LZ and criticism
Post #75 Posted: Wed Dec 11, 2019 3:29 am 
Oza

Posts: 2762
Liked others: 16
Was liked: 3867
I'm struggling with two things here.

Let us first remember the topic of the thread: reviewing SL articles using LZ and criticism. I'm not quite sure whether that's 'reviewing AND criticism' or 'using LZ and criticism'. I assumed the former. It seems others understood the latter. But I think the real topic has revealed itself, here and in other threads, as 'old go wisdom is old wives' tales'.

The first thing I don't properly understand is the very strong urge to debunk all old wives. In real life, those old wives were your mothers and grandmothers, and they did a pretty good job of getting you where you are today. Much of the time they used their intuition and not their consciousness, made plenty of mistakes, too. But, if you could live your life again, would you still like to be brought up at home, or would you prefer to be brought up in a germ-free isolation ward of a hospital? A mother's love or modern technology? Of course, you might want your mother to make sure you got your measles vaccination, and you might even look down smugly on those 'benighted' parents who fear such vaccinations. But I feel pretty sure most people would prefer some mix of old and new, and would not fret over a lack of perfection. Would you rather have a happy life of 70 years or an unhappy life of 100 years? There's no right answer, of course, except the one you choose for yourself. But why assume the right answer for you is the right answer for everyone else? Or, more pertinently, why assume it's a good idea to keep telling everyone else their own answers are wrong?

It's ironic that go is a game of co-existence. Too many go players don't practise what they play.

The other thing I can't understand is the implicit assumption that everyone wants to be a 13-dan. Well, I don't. I wouldn't object to it, but I don't want to work at it. But I do enjoy watching other people work. That's normal. That's why we have 100,000 people and millions more on sofas at home watching 22 people kicking a ball around on a soggy pitch with a gale blowing hard. Soccer has players and fans. Go has players and fans.

Like fans everywhere, go fans love to learn a little bit more about their passion. But it doesn't have to be facts such as 'bots say new move 41 in this joseki has a win rate of 65% whereas the old move was 50%'. It might be a bit of trivia such as 'Takemiya loves ballroom dancing.'

Even if they do think their passion would be enhanced by a bit of technical knowledge, they know they don't have to try the sorcery of new moves, or to try to do PhD-type research with no grant. There is still plenty of scope for even dan players to get much stronger by using the easy-to-understand old wives' tales. They're wrong? Oh, tell me why? You can't? OK, come back when you can. 20 years time? OK, but in the meantime I've got s live to live.

I think also it is forgotten too often that go is a mind sport. Not, for most people, a computing exercise. What many fans enjoy is witnessing the battle not of one joseki or fuseki or tesuji over another but of one human mind over another. That is often better understood through fellow humans' knowledge - old wives' tales.

I suspect the art in the Sistine Chapel stirs most of us much more than the admittedly brilliant artwork in the latest Disney of The Lion King. I thought it was an awful, soulless film. But that was just for me. I can well understand that many people would enjoy the experimenting in the film, on and off the screen. What I would totally fail to understand, though, is if a computer artist went to The Vatican and daubed graffiti over da Vinci's work and told us this is not how paintings should be done. That perspective is wrong. It has a view rate of only 50%; the computer would give you a view rate of 60%. You need an XYZ123A GPU, not a sable paintbrush. And why is he showing angels - angels don't exist?

I am certain that every artist on the Disney films studied how to draw using old men's heuristics of egg shapes and circles, went through phases of drawing boring fruit, and ended up studying da Vinci and his ilk. I suspect many will have felt awestruck as they looked at the Old Masters. At the very least, I suspect very, very few would ever dismiss the Old Masters as old wives.


This post by John Fairbairn was liked by: columbo
Top
 Profile  
 
Offline
 Post subject: Re: reviewing SL articles using LZ and criticism
Post #76 Posted: Wed Dec 11, 2019 3:49 am 
Gosei
User avatar

Posts: 1731
Location: Ghent, Belgium
Liked others: 268
Was liked: 798
Rank: KGS 2d OGS 1d Fox 4d
KGS: Artevelde
OGS: Knotwilg
Online playing schedule: UTC 18:00 - 22:00
Let's borrow an idea from Alan Turing here, now that we're talking AI.

Suppose you could at all times ask Go Seigen's opinion and he would still be able to tell you if a move is good or bad, and by how much, but not "why". Would you ask him?

Now suppose Go Seigen and an AI are in the same room and you ask them the same questions. You don't know who's answering and by the quality of their answers you are unable to distinguish who says what. Would you still ask?

Now Go Seigen dies and the room is left with only AI. Would you still ask?

In any of these stages, people would probably stop asking. Maybe the first stage is already deterring you.

Personally, I would ask, ask, ask. Not because I want to become 13 dan - I will never. But I like finding out the truth, or getting as close to the truth as I can. The same reasons why I'd read pro opinions apply to why I'm feeding positions to AI. The answers are different but I find that rather enlightening than frightening.

Top
 Profile  
 
Offline
 Post subject: Re: reviewing SL articles using LZ and criticism
Post #77 Posted: Wed Dec 11, 2019 4:40 am 
Oza

Posts: 2762
Liked others: 16
Was liked: 3867
I think there is an abiding problem with L19 for people like me: there are too many mathematicians and computer scientists, who all each reinforce each other's way of thinking. There are people, apparently in a minority, who are interested in viewing the world in different ways.

So, to answer one of your questions:

Quote:
Suppose you could at all times ask Go Seigen's opinion and he would still be able to tell you if a move is good or bad, and by how much, but not "why". Would you ask him?


Probably not. First, I would assume that his explanation would be over my head. As confirmation of that, he tries to explain why particular moves are good or bad in his books on 21st century go and I don't really understand that. Superficially it feels like I understand, but it's not in a useful way. Furthermore, I assume even top pros didn't really understand what he says either, because that's why they couldn't beat him. He inspired top pros, but I don't really think he taught them any principles. The only way they learnt, I feel, was by another 10,000 hours of practice. Or, to unquote Gomoto: consciousness > intuition.

Second, what I would be most interested in from GSG is not why move A is allegedly better than B, but why he played it. Was it psychology? Was it because he'd had a bad night's sleep? Was it because he played too fast and boobed? If so, why did he play too fast. Or did he just see more than the opponent? And in Go Seigen's case it's also essentially how I wrote my books about him, and as they have been popular books, I feel safe in claiming that this is how many fans view the game.

The search for the ultimate truth sounds very noble and all that. But most of us prefer a bit of gossip, and human frailty to gossip about too.

I infer the hard-boiled AI fans and I are on completely wavelengths. Fine. I'm comfortable with that. I even toy around with a bit of AI myself. But I wouldn't feel comfortable with my wavelength being shut down, or being drowned out by propaganda loudspeakers. That would seem a very North Korean way of doing things to me.


This post by John Fairbairn was liked by 2 people: Applebaps, columbo
Top
 Profile  
 
Offline
 Post subject: Re: reviewing SL articles using LZ and criticism
Post #78 Posted: Wed Dec 11, 2019 4:45 am 
Oza
User avatar

Posts: 2305
Location: Tokyo, Japan
Liked others: 2195
Was liked: 1298
Rank: Jp 6 dan
KGS: ez4u
I think I am more on JF's side. What is truth? You are part way through a game. You have the opportunity to ask for LZ's choice for the next move. However, you are not allowed to see any of the variations and will not have any further recourse to LZ throughout the rest of the game. LZ returns a move. Should you play this move or an alternative that you choose based on the heuristics that you have previously studied and utilized up to now?

If you play LZ's recommendation, how do you plan to continue to choose your future moves in the game?

Note that I am not trying to attack you. My question is serious. I have great difficulty in using the ideas from AI games. It is easy to sit passively and watch LZ generate variations and moves in Lizzie. It is a very different challenge to take those ideas and turn them into consistent and successful play in real life.

And on Gomoto's saying, Consciousness becomes intuition (through the 10,000 hours).

_________________
Dave Sigaty
"Short-lived are both the praiser and the praised, and rememberer and the remembered..."
- Marcus Aurelius; Meditations, VIII 21


This post by ez4u was liked by: Applebaps
Top
 Profile  
 
Offline
 Post subject: Re: reviewing SL articles using LZ and criticism
Post #79 Posted: Wed Dec 11, 2019 6:12 am 
Honinbo

Posts: 10554
Liked others: 3501
Was liked: 3313
xela wrote:
Here's another example. Many years ago I learned this sequence from 38 Basic Joseki:
Click Here To Show Diagram Code
[go]$$Bc
$$ . . . . . . |
$$ . . b a . . |
$$ . . O X . . |
$$ . . O X . . |
$$ . O X 1 . . |
$$ . . . . . . |
$$ . . . . . . |
$$-------------+[/go]

I already knew that black ''a'' or ''b'' instead of :b1: were the "real" joseki moves. But :b1: is a way of staying away from the complications of the avalanche. I remember the first time a dan player tried to "avalanche" me in a handicap game, and how disappointed he was when I didn't play along! (But I still spent time studying some variations of the avalanche, even if I wasn't brave enough yet to try them in a serious game.)


Well, now the bots tell you to play :b1: below, instead. :)

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?

_________________
The Adkins Principle:
At some point, doesn't thinking have to go on?
— Winona Adkins

My two main guides in life:
My mother and my wife. :)

Everything with love. Stay safe.

Top
 Profile  
 
Offline
 Post subject: Re: reviewing SL articles using LZ and criticism
Post #80 Posted: Wed Dec 11, 2019 6:17 am 
Judan

Posts: 6621
Location: Cambridge, UK
Liked others: 417
Was liked: 3638
Rank: UK 4 dan
KGS: Uberdude 4d
OGS: Uberdude 7d
I think John's point about people enjoying Go as fans rather than (or as well as) players answers ez4u's point about the difficulty of how to apply LZ's lessons when you are playing a game yourself. You simply don't need to! Exploring pro games, problems, exploring variations etc with LZ, such as I did here on book by Fujisawa Shuko, is an interesting and enjoyable activity for me (and I hope other readers too) without requiring validation that I can play as well as LZ in my own games. Just look at that f12 attachment, it's so cool! And by exploring the variations I can see its purpose and followups depending how the opponent answers. If I can imagine such moves in my own games then that's a bonus, but it's already a positive without me playing.

Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 113 posts ]  Go to page Previous  1, 2, 3, 4, 5, 6  Next

All times are UTC - 8 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group