Interesting ugly shape lesson from bots

Uberdude · #1

Watching miracle97 (Byun Sangil 9p) get his usual beating from BensonDarr / PhoenixGo on Fox, one move stuck out to me, white to play. After approach at 1 white attached with 2, a move beloved by ddks that they used to get told off for and pros only played rarely, but is back in fashion with the bots, particularly with a side stone marked (and black already strong on lower side so white doesn't mind strengthening black there to solidify the corner/side) and Byun wedged, a standard resistance to white's plans. White then ends up with 2 cuts, how to defend?

Click Here To Show Diagram Code: [go]$$Bc BensonDarr (white) vs miracle97 $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . X O . . . . . . . . . . . . O O O . | $$ | . X O X . X . . . . X X O O . X O X . | $$ | . X X O X . . . . , X . X O . X X . . | $$ | . X O O O X X X . . . X . O . . . . . | $$ | . O O X X O O . . . . . . . . . . . . | $$ | . . O X . X X . . . . . . . . . . . . | $$ | . . O X X X O . . . . . . . . . X . . | $$ | . . . O O O . . . . . . . . . . . . . | $$ | . . O , . . . . . , . . . . . , W . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . 4 . . . . | $$ | . . O O O O . . . , . . . 2 3 O . . . | $$ | . . X X X . . . . . . . . 1 5 . . . . | $$ | . . . . . . X . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

The following pro game pattern search from waltheri shows the usual shape choices, most common is simple and easy solid connect at 1, or there's the tougher corner block (and then usually black cuts). There was an interesting game recently between Mateusz Surma 2p and Stanislaw Frejlak 7d where Mateusz chose the block but connect was better so they had a fun fight, see commentary at https://explorebaduk.com/2019/02/13/vad ... rejlak-7d/.

Attachment:

BD m97 pattern search.PNG [ 222.82 KiB | Viewed 2672 times ]

But BensonDarr picked a move never seen in waltheri, and one sure to give Otake a heart attack for being such an ugly shape:

Click Here To Show Diagram Code: [go]$$Bc Not a typo $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . X O . . . . . . . . . . . . O O O . | $$ | . X O X . X . . . . X X O O . X O X . | $$ | . X X O X . . . . , X . X O . X X . . | $$ | . X O O O X X X . . . X . O . . . . . | $$ | . O O X X O O . . . . . . . . . . . . | $$ | . . O X . X X . . . . . . . . . . . . | $$ | . . O X X X O . . . . . . . . . X . . | $$ | . . . O O O . . . . . . . . . . . . . | $$ | . . O , . . . . . , . . . . . , O . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . O . . . . | $$ | . . O O O O . . . , . . . O X O 6 . . | $$ | . . X X X . . . . . . . . X X . . . . | $$ | . . . . . . X . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

What's the idea? The most obvious is that compared to solid connection if black pushes at q3 white can safely block at r3. So instead a cut and fight is expected (during which r10 comes in handy) but it's really complicated. For more than that explore it yorself with lizzie.

I checked with a variety of bots, and many liked it too as #1 choice:
- LeelaZero #205 solid intuition, extend #1 within 1k playouts (solid connect -3%)
- LeelaZero #157 after 40k playouts, started out liking solid (<1% worse)
- LZ #157b (15 block 157 with 40b games) after 2k.
- LZ 191 (sees it as promising blue circle quite early but doesn't become #1 playouts until 8k)
- Elf v1 and v2 both like it within a few hundred playouts
- LMGX88, the 'Leela Master' trained on human games. After 20k playouts for connect ugly extend has 30 playouts, but then it realises it's good and by 47k total extend is #1
- MiniGo v15 990 cormorant. Doesn't discover it until 4k but by 10k is #1.

So I think it is very likely the best move, but one us humans with sense of good shape were blind to.

P.S. Speaking of Mateusz I just saw him on Fox (lost to FineArt) and he beat young Japanese hopeful Onishi Ryuhei 3p a few weeks ago, nice win!

ez4u · #2

Quote:

...So I think it is very likely the best move, but one us humans with sense of good shape were blind to.

Or ...us humans with our outmoded sense of shape...?

Uberdude · #3

Sure, Korean pros have been caring a lot less about good shape than the Japanese style and just playing moves that work for 20 years, but still this one eluded them. Or can anyone find an example?

Btw, this extend reminds me of this idea for defending both cuts* at once I thought I was so smart for discovering when I was ddk ;-)

I wonder if the bots ever think this is a good move!

Click Here To Show Diagram Code: [go]$$W $$ . . . . . . . . . $$ . . . . 1 . . . . $$ . . . . O . . . . $$ . . O O X O O . . $$ . . . X X X . . . $$ . X . . . . . X . $$ . . . . . . . . . $$ -----------------[/go]

* though if you play normal net to each cut black has the Shuei-style breakout tesuji.

Joaz Banbeck · #4

The theoretical refutation. :cool:

Click Here To Show Diagram Code: [go]$$Bc $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . X O . . . . . . . . . . . . O O O . | $$ | . X O X . X . . . . X X O O . X O X . | $$ | . X X O X . . . . , X . X O . X X . . | $$ | . X O O O X X X . . . X . O . . . . . | $$ | . O O X X O O . . . . . . . . . . . . | $$ | . . O X . X X . . . . . . . . . . . . | $$ | . . O X X X O . . . . . . . . . X . . | $$ | . . . O O O . . . . . . . . . . . . . | $$ | . . O , . . . . . , . . . . . , O . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . 7 . . . . | $$ | . . . . . . . . . . . . . . O . . . . | $$ | . . O O O O . . . , . . . O X O 6 . . | $$ | . . X X X . . . . . . . . X X . . . . | $$ | . . . . . . X . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

Bill Spight · #5

Thanks.

Waltheri indicates that this is a human blind spot. But I believe that John Fairbairn has shown us a similar human shape play, one that was not a descent to the third line, but an extension towards the side (on the 5th line? I don't remember exactly).

Also, I would not call this ugly shape, as it exemplifies the idea of shape as locally efficient play. It is a cousin of the keima connection, as the cutting stone is captured in a ladder. (OC, if Black peeps on the fourth line and White then makes a solid connection, which forms an empty triangle.

)

BTW, given the :wc:

stone, my thought was the hanging connection on the third line.

Uberdude · #6

Although when black cuts on the right white can capture black in a ladder, black would be happy to get a ponnuki outside, so white connected on the outside and the effect of the extension can be seen in that if white did that directly, black would rather play a for 7, white 8 and then b to live in the corner, but instead black cut at 7 to start an unreasonable fight.

Click Here To Show Diagram Code: [go]$$Bc tewari-ish $$ | . . O , . . . . . , . . . . . , W . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . 6 4 7 . . . | $$ | . . O O O O . . . , . . . 2 3 O 8 . . | $$ | . . X X X . . . . . . . . 1 5 a b . . | $$ | . . . . . . X . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

Here's Byun's attempts to set up a miai, expertly dodged by BensonDarr (top left sacrifice was also instructive, though LZ thought c10 was a mistake as white can capture cut with b8 hane so should o17 approach (black should o17 not 3-3).

Bill Spight · #7

Uberdude wrote:

Although when black cuts on the right white can capture black in a ladder, black would be happy to get a ponnuki outside, so white connected on the outside and the effect of the extension can be seen in that if white did that directly, black would rather play a for 7, white 8 and then b to live in the corner, but instead black cut at 7 to start an unreasonable fight.

Click Here To Show Diagram Code: [go]$$Bc tewari-ish $$ | . . O , . . . . . , . . . . . , W . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . 6 4 7 . . . | $$ | . . O O O O . . . , . . . 2 3 O 8 . . | $$ | . . X X X . . . . . . . . 1 5 a b . . | $$ | . . . . . . X . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

Here's Byun's attempts to set up a miai, expertly dodged by BensonDarr (top left sacrifice was also instructive, though LZ thought c10 was a mistake as white can capture cut with b8 hane so should o17 approach (black should o17 not 3-3).

Doesn't that imply this?

Click Here To Show Diagram Code: [go]$$Bc No nuki $$ | . . O , . . . . . , . . . . . , W . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . 7 4 . . . . | $$ | . . O O O O . . . , . . 8 2 3 O 6 . . | $$ | . . X X X . . . . . . . . 1 5 . . . . | $$ | . . . . . . X . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

Instead of this?

Click Here To Show Diagram Code: [go]$$Bc Nuki? $$ | . . O , . . . . . , . . . . . , W . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . 8 . . . . | $$ | . . . . . . . . . . . . . 7 4 . . . . | $$ | . . O O O O . . . , . . 9 2 3 O 6 . . | $$ | . . X X X . . . . . . . . 1 5 . . . . | $$ | . . . . . . X . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

----

So Benson Darr is Kimmo?

Uberdude · #8

Yes, LZ would have extended as in the first diagram, Black gets the happy p6 atari to make the extend an empty triangle, but it's still useful in allowing white to safely jump to down to q2 when black m2 slides next.

Bill Spight · #9

Here is the similar shape that I was thinking about.

Click Here To Show Diagram Code: [go]$$Bc Similar shape from 1939 game $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . X . . . | $$ | . . O , . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . O . . | $$ | . . . . . . . . . . . . . . . O . . . | $$ | . . . . . . . . . . . . . . . O X . . | $$ | . . . . . . . . . . . . . . . X . . . | $$ | . . . . . . . . . . . . . . . B . . . | $$ | . . . , . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . X , . . . . . , . . . . . , . . . | $$ | . . . . O . . . . O . . . . . X . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

It appears in a 1939 game. Here is the post by John Fairbairn. viewtopic.php?p=204278#p204278

gowan · **#10**

When we said a move was good shape in the past it meant that the move had properties like efficient, light, has good follow-up, etc. For an arrangement of the stones to be beautiful there is no necessary intrinsic appearance. Classic bad shape such as the empty triangle has always been known to be occasionally a good move. We human players depend on a variety of heuristics to decide whether a move is good or not and so-called good shape is one of them. The strong AI bots don't need the heuristics so a grasp of shape is not meaningful.

Ian Butler · **#11**

Uberdude wrote:

Sure, Korean pros have been caring a lot less about good shape than the Japanese style and just playing moves that work for 20 years, but still this one eluded them. Or can anyone find an example?

Btw, this extend reminds me of this idea for defending both cuts* at once I thought I was so smart for discovering when I was ddk ;-)

I wonder if the bots ever think this is a good move!

Click Here To Show Diagram Code: [go]$$W $$ . . . . . . . . . $$ . . . . 1 . . . . $$ . . . . O . . . . $$ . . O O X O O . . $$ . . . X X X . . . $$ . X . . . . . X . $$ . . . . . . . . . $$ -----------------[/go]

* though if you play normal net to each cut black has the Shuei-style breakout tesuji.

What is the Shuei-style breakout tesuji?

is it extend both cutting stones upwards so you capture white 1 and the stone below it?

Uberdude · **#12**

Ian Butler wrote:

What is the Shuei-style breakout tesuji?

If you cut and they play the normal net, then push out of one net like an idiot and you can break out of the other thanks to it now being atari.

Click Here To Show Diagram Code: [go]$$B $$ . . . . . . . . . $$ . . . 6 8 9 . . . $$ . . 4 5 O 7 2 . . $$ . . . 3 O 1 . . . $$ . . O O X O O . . $$ . . . X X X . . . $$ . X . . . . . X . $$ . . . . . . . . . $$ -----------------[/go]

Note that instead of the tight net white can try the looser knight move net of 2 so that push can be answered with pullback 6, but the problem is black could just play 3 at a and white is short of libs to hane. If there were white stones at b/c then this might work though.

Click Here To Show Diagram Code: [go]$$B $$ . . . . . . . . . $$ . . 4 . 6 . 2 . . $$ . . . 5 O . . . . $$ . . . 3 O 1 a . . $$ . c O O X O O b . $$ . . . X X X . . . $$ . X . . . . . X . $$ . . . . . . . . . $$ -----------------[/go]

This is based on a famous tesuji played by Honinbo Shuei, I first came across it in the book 'Beyond Forcing Moves', which explains its probey genius.
See move 92 here: https://online-go.com/game/425048 (I replayed some of my favourite pro games when I won the OGS title qualification tournament to play against myself as the title holder!).

Also Charles Hibbert played a similar fun idea in the 2016 British championship, see move 103 in http://britgo.org/files/bchamp/2016/gam ... _clone.sgf

Uberdude · **#13**

Check out number 7 of https://explorebaduk.com/2019/03/20/2690/

Knotwilg · **#14**

No, check out 28

Interesting ugly shape lesson from bots

Who is online