It is currently Tue Feb 20, 2018 8:52 am

All times are UTC - 8 hours [ DST ]




Post new topic Reply to topic  [ 7 posts ] 
Author Message
Offline
 Post subject: The process of discovery
Post #1 Posted: Thu Nov 02, 2017 3:51 am 
Gosei

Posts: 1967
Liked others: 12
Was liked: 2886


I'm curious what people think about moves A and B by White here (the upper right quadrant is empty apart from these moves).

A is overwhelmingly the commonest move. Until modern times it was the only move ever played by pros, going back to mid-Edo times. In my experience, even amateurs routinely play it, though I suspect it's then less to do with understanding it and mostly because it's such an unusual move that it made a big impression on them when reading joseki books.

In any case, I don't claim to really understand it, and I have often wondered how top players of the past came to do choose this move without having tried the obvious B first (or, I repeat, ever!). Any suggestions?

Before you pronounce, let me point out that in recent times B has started to appear among top pros, so whatever reasoning the old pros used wasn't all that obvious after all. And the use of B post-dates the introduction of komi by a long, so is presumably not a special feature of old no-komi go, and it is not country-specific. Also, while the joseki is fairly common, it's not so common that you'd expect players to build up a lot of intuition in playing it. It seems the decision is based more on a priori reasoning than on a posteriori experience.

So, what is the reasoning, and why do you think the old pros so plainly thought they'd got it right first time?


This post by John Fairbairn was liked by 2 people: Gomoto, gowan
Top
 Profile  
 
Offline
 Post subject: Re: The process of discovery
Post #2 Posted: Thu Nov 02, 2017 5:48 am 
Dies with sente

Posts: 116
Liked others: 0
Was liked: 7
Rank: 2d
I don't claim to understand anything either, but my naive impression is that A stabilizes the corner from both sides, so no matter which side B choses W can hope to get some initiative later. B may have slightly more effect on the right side, but also leaves weakness and some potential sente exchanges for B from top. The rest of the board may also matter here.

If pro evaluation changes that may be because the various factors are weighted slightly differently now than in the past? Are there similarities in nearby stones in the cases where B is played?

Top
 Profile  
 
Offline
 Post subject: Re: The process of discovery
Post #3 Posted: Thu Nov 02, 2017 6:25 am 
Tengen

Posts: 4444
Location: Cambridge, UK
Liked others: 179
Was liked: 2205
Rank: UK 4 dan
KGS: Uberdude 4d
OGS: Uberdude 7d
The plus of B descent is it's better for territory on the right side/corner and reducing black eyespace or territory for the 2 stones, but a big minus is black has a nice peep at p18 which if you can get it in sente is very nice, and makes it even easier to sacrifice the 2 stones. The hanging connection avoids that (likely) forcing move, is more eyeshapey and makes the game more leisurely for white so saves on thinking, but I think if you want to play the 100% most efficient move then descent could sometimes be better but needs a lot of reading and judgement on future trades so hanging connection could be a cop-out for the "good-enough" lazy players. The sort of situations I would see as justifying the descent:
- black is not going to sacrifice the 2 stones because that side is too big globally so would extend and then you n17 (which is normally too cramped, if black were at k17 and you could more happily go one further to m17) which due to top side situation is not as sad as normal (e.g. severe follow up on l17).
- black does sacrifice the 2 stones, but the way you capture them means rather than answering p18 at q18 you can answer it with cleaning up the 2 stones (and q17 cut doesn't set up a painful semedori)
- black does sacrifice the 2 stones, and you do answer p18 at q18, but you capture them on a large enough scale that you are winning despite being forced at p18. Bonus if l17 + p18 group is weak and can be attacked for good profit later.

As for why B is getting more common, I'd expect it's the pros are now striving more for maximum efficiency (sometimes bordering on overplay) and paying less heed to the principles of old which let you play 98% Go but they want 100% or 101% now. Also the idea of leaving some bad aji or playing a move which had some bad aspects but you calculate that the game will flow in a direction that they are not exploited is something that AlphaGo seems to be teaching us (e.g. that early peep in joseki vs Lee in game 2 in many situations you could end up regretting it as bad aji keshi, but AlphaGo doesn't care and says "I know the game will go in a direction in which those bad possibilities don't happen", or the numerous early erasure or centre-focused moves like shoulder hits which it ends up efficiently combining or sacrificing).


This post by Uberdude was liked by 2 people: Bill Spight, gowan
Top
 Profile  
 
Offline
 Post subject: Re: The process of discovery
Post #4 Posted: Thu Nov 02, 2017 6:50 am 
Lives in sente

Posts: 778
Location: Littleton, CO
Liked others: 258
Was liked: 220
Rank: KGS 3k
Universal go server handle: jeromie
I certainly can’t answer better than Uberdude, but I agree that B seems to gain territory at the cost of flexibility. My impression is that flexibility has been highly valued by most professionals, but now more are experimenting with maximizing efficiency even if it means eliminating future possibilities.

Top
 Profile  
 
Offline
 Post subject: Re: The process of discovery
Post #5 Posted: Thu Nov 02, 2017 9:14 am 
Judan

Posts: 6653
Liked others: 1560
Was liked: 2471
John Fairbairn wrote:


I'm curious what people think about moves A and B by White here (the upper right quadrant is empty apart from these moves).

A is overwhelmingly the commonest move. Until modern times it was the only move ever played by pros, going back to mid-Edo times. In my experience, even amateurs routinely play it, though I suspect it's then less to do with understanding it and mostly because it's such an unusual move that it made a big impression on them when reading joseki books.

{snip}

So, what is the reasoning, and why do you think the old pros so plainly thought they'd got it right first time?


I first learnt about A when I started reading go books (aside from Lasker and Korschelt) and magazines, at around 4 kyu. I still remember the reasoning given. While B takes the corner and attacks the two Black stones on the right, it leaves White vulnerable to attack from the top side, as Black can threaten the cut and the 2x3 eye in the corner is not alive. OTOH, A also attacks those two stones and secures the corner, although on a slightly smaller scale. Most importantly it secures life for the White group and thus supports a play against the single Black stone on the top. Black will normally have something in the top left corner, so such a play will usually be an invasion or reduction.

Every time such a position has occurred in my own games, I have played at A. Often the top side stone has been closer, which makes A even better. ;) However, looking at this position anew, I lean towards B, probably with subliminal influence by AlphaGo. I doubt if AlphaGo would play A unless it was already ahead. ;) Waltheri ( http://ps.waltheri.net/ ) also shows C, which was played by Yamashita Keigo. Yamashita lost that game, but C looks like an AlphaGo move to me. :) At first blush, the Waltheri winning percentage stats seem to support A, which has a win rate of around 60%, while B has a win rate of around 45%. But C has a win rate of 100%, while the Yamashita game is the only one that shows up, and Yamashita lost. Something is wrong with Waltheri's win rate. :(

_________________
Don't cry for me, Sergeant Tina.

Top
 Profile  
 
Offline
 Post subject: Re: The process of discovery
Post #6 Posted: Thu Nov 02, 2017 9:52 am 
Lives in sente

Posts: 1232
Liked others: 273
Was liked: 324
Rank: 5d
GD Posts: 1000
I haven't seen any commentary on move A versus B, but it seems to me that, obviously depending on the larger scale position, very often B would be considered leaving bad aji. Descents like B occur in many other situations, such as, for example, some variations stemming from the two space high pincer to a one space high approach move to the 3-4 corner move. In one variation in that joseki the descent move may involve ladder aji after a cut. In all these situations I doubt whether we can say whether the hanging connection or the descent is better without knowing to what extent it is bad aji.

Top
 Profile  
 
Offline
 Post subject: Re: The process of discovery
Post #7 Posted: Thu Nov 02, 2017 11:00 am 
Tengen

Posts: 4444
Location: Cambridge, UK
Liked others: 179
Was liked: 2205
Rank: UK 4 dan
KGS: Uberdude 4d
OGS: Uberdude 7d
Btw, Mr GoCommentary.com (a strong Chinese amateur) did a video on a related shape (with 5-4 press joseki in progress in adjacent corner) in which the issue of descent vs hanging connection came up and a key issue is either player taking the l17 key point. https://youtu.be/Ofu1xRwE2_4


This post by Uberdude was liked by: Gomoto
Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 7 posts ] 

All times are UTC - 8 hours [ DST ]


Who is online

Users browsing this forum: Majestic-12 [Bot] and 2 guests


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