Learning from the Master

Uberdude · #1

On Tygem and Fox there is a rumoured AI called Master beating up all the top pros (in blitz) with an unusual style and some interesting moves, so I thought I'd make a thread to study them (see thread and records on German publisher Herbsacker Verlag's blog).

It seems to generally favour the low over high approach to a 3-4 stone, but when it does and is 2-space-high pincered and the big knight table shape joseki follows I've seen it twice push through to take the corner and give a big wall, which afaik human joseki knowledge says is bad. Here is a game vs Kim Jiseok 9p (black) who probably played the unusual 2-space high enclosure at top left in imitation of Master who likes it. After getting the corner in sente Master likes to play the hoshi near the wall, and then played some very interesting light haengma of knight's jumps to n11 and m13 which reduce the influence of black's wall and even ask some questions about its liveliness with the threat of saving the cutting stone. If these are good moves, should black have played something like the o10 cap to preserve the influence or is that too easy going? This kind of centre balance of power seems to be something bots are good at, and I feel any advantage black gets with that corner not-joseki could well be erased in a single careless move on the outside.

Click Here To Show Diagram Code: [go]$$Wcm14 Game 41 Master (white) vs Kim Jiseok $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . X . . . . . . . . . 9 . . 7 . . | $$ | . . . , . . . . . , . X 0 8 O O X 5 . | $$ | . . . . . . . . . . . . . O X 1 3 4 . | $$ | . . . X . . . . . . . . . . X 2 X 6 . | $$ | . . . . . . . . . . . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . , . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . O . . . . . , . . . . . O . . . | $$ | . . . . . O . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

Click Here To Show Diagram Code: [go]$$Wcm24 $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . 0 . . . . . | $$ | . . . X . . . . . . . . . O . . O . . | $$ | . . . , . . . . . , . X X X O O . O . | $$ | . . . . . . . . . . . . . O X O O X . | $$ | . . . X . . . . . . . . . . X X X X . | $$ | . . . . . . . . . . . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . 3 . . . . . . | $$ | . . . , . . . . . , . . . . . 1 . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . 6 X . 2 . . . . . . . . . . . . . . | $$ | . . 5 . . . 8 . . . . . . . . . . . . | $$ | . . . O . 7 4 . . , . . . . . O . . . | $$ | . . . . . O . . . . . . . . . . . . . | $$ | . . . . . . . 9 . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

Click Here To Show Diagram Code: [go]$$Wcm34 $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . X . . . . . | $$ | . . . X . . . . . . . . . O 1 . O . . | $$ | . . . , . . . . . , . X X X O O . O . | $$ | . . . . . . . . . . . . . O X O O X . | $$ | . . . X . . . . . . . . . . X X X X . | $$ | . . . . . . . . . . . 3 . . . O . . . | $$ | . . . . . . . . . . . . . . 8 . . . . | $$ | . . . . . . . . . . . . O 7 4 . . . . | $$ | . . . , . . . . . , . . . 6 5 O . . . | $$ | . . . . . . . . . . . 9 . 0 . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . 2 . . . | $$ | . X X . X . . . . . . . . . . . . . . | $$ | . . O . . . X . . . . . . . . . . . . | $$ | . . . O . O X . . , . . . . . O . . . | $$ | . . . . . O . . . . . . . . . . . . . | $$ | . . . . . . . O . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

Here's another example versus Jiang Weijie 9p: again we see the immediate hoshi and some knight jumps for 34-38. The way it ignored approach to play 38 makes me think 37 tenuki could be a mistake. Are pros too keen to play normal opening approach moves when some fancy centre move is the key point? Tenuki to 48 turn allowing double approach was very interesting too!

Click Here To Show Diagram Code: [go]$$Wcm24 Game 35 Master (white) vs Jiang Weijie $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . 5 4 6 . . . . . . . . . . . . . | $$ | . . 7 X 3 X . . . . . . . . . . . . . | $$ | . . . O 1 2 O . . , . . . . . X . . . | $$ | . . . O X X . . . . . . . . . . . . . | $$ | . . 9 8 O . . . . . . . . . . . . . . | $$ | . . . 0 . . . . . . . . . . . . . . . | $$ | . . . X . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . , . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . O O X X X X X . , . . . . . O . . . | $$ | . . . O O O O . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

Click Here To Show Diagram Code: [go]$$Wm34 $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . O X X . . . . . . . . . . . . . | $$ | . . O . O X . . . . . . . . . . . . . | $$ | . . . O O X O . . 1 . . 2 . . X . . . | $$ | . . . O X X . 6 . . . . . . . . . . . | $$ | . . O X O . . . . . 3 . . . . . . . . | $$ | . . . X . . 8 5 . 7 . . . . . . . . . | $$ | . . . X . . 9 0 . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . , . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . 4 . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . O O X X X X X . , . . . . . O . . . | $$ | . . . O O O O . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

Click Here To Show Diagram Code: [go]$$Wcm44 $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . O X X . . . . . . . . . . . . . | $$ | . . O . O X . . . . . . . . . . . . . | $$ | . . . O O X O . . O . . X . . X . . . | $$ | . . . O X X . X . . . . . . . . . . . | $$ | . . O X O 4 1 3 . . O . . . . . . . . | $$ | . . . X . 2 X O . O . . . . . . . . . | $$ | . . . X . . O X . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . , . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . 5 X . . . . . . . . . . 9 . . . . . | $$ | . O O X X X X X . , . . 8 7 . O . . . | $$ | . . . O O O O . . . . . . 6 . 0 . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

Bonobo · #2

Thanks, Uberdude,

RobertJasiek · #3

Yes, Uberdude, the human tenukis are mistakes. Black must maintain the influence of his wall.

jeromie · #4

For those like me who are interested in this discussion but not immediately familiar with the joseki in question, I think this is the basic corner sequence under consideration:

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

a is the most common continuation according to Josekipedia (though that might be old); b is the move that Master has been playing.

Krama · #5

Maybe the sequence is bad actually but somehow fits Masters style of play. A lot of pros were known to play "not good" moves because it suited their style.

Also this is the reason people tell you not to copy the moves of chess engines, since sometimes they tend to play something no human would play. Human would choose a move that they can understand and play on with just like a machine plays it's own style because it knows how to continue from it.

Great times to be alive.

Uberdude · #6

Another of Master's favourite moves is the 2 space high corner enclosure from a 3-4, which is pretty rare to see in normal early-board situations in contemporary pro play. I wonder which pro will be the first to copy it?

. Will it become a new fashion, like AlphaGo game 5 opening or that peep?

Uberdude · #7

Another idiosyncratic move of Master (now confirmed as a new version of AlphaGo) is early 3-3 invasion of a 4-4 in an empty position. It's a common strategy to 3-3 invade once it makes 2 far extensions (particularly if they are low, so you don't see that much in strong players' games), but to do so against a lone 4-4 is pretty unusual. Note how it doesn't make the second line hane connect, presumably because that gives white eye shape. The result wall ends up getting attacked later, rather similar to the results above:

Click Here To Show Diagram Code: [go]$$Bcm19 $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . 0 . . O . . . . . . . . . . X . . | $$ | . . 9 O . . . . . , X . . X . X . O . | $$ | . . . . . . . . . . . . . . . O . . . | $$ | . . X . . . . . . . . . . O O . O . . | $$ | . . . . . . . . . . . . . . X . . . . | $$ | . . . . . . . . . . . . . . . X . O . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . , . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . 8 . . . . . . . . . . . . . . . . | $$ | . 7 6 . . . . . . . . . . . . . . . . | $$ | . 5 4 . . . . . . . . . . . . . . . . | $$ | . . 3 O . . . . . , . . . X . , X . . | $$ | . . 1 2 . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

And another one vs Gu Zihao 5p:

Bill Spight · #8

Uberdude wrote:

Another of Master's favourite moves is the 2 space high corner enclosure from a 3-4, which is pretty rare to see in normal early-board situations in contemporary pro play. I wonder which pro will be the first to copy it?

. Will it become a new fashion, like AlphaGo game 5 opening or that peep?

That's my guess. The two space high approach became popular in the late 19th century. If the two space high approach is good, how can the two space high enclosure be bad?

Bill Spight · #9

Uberdude wrote:

Another idiosyncratic move of Master (now confirmed as a new version of AlphaGo) is early 3-3 invasion of a 4-4 in an empty position. It's a common strategy to 3-3 invade once it makes 2 far extensions (particularly if they are low, so you don't see that much in strong players' games), but to do so against a lone 4-4 is pretty unusual. Note how it doesn't make the second line hane connect, presumably because that gives white eye shape.

(Emphasis mine.)

Another joseki bites the dust?

Uberdude · **#10**

Bill Spight wrote:

Another joseki bites the dust?

Or maybe the birth of a new one? I think the full joseki could remain for situations with existing extensions because you don't mind making white stronger (though of course white might resist you taking the corner in sente by playing jump instead of hane or the double hane) but Master agrees the full joseki is good for white on an open board. But its new idea (or maybe not I didn't pattern search yet) is to play this partial version and tenuki.

John Fairbairn · **#11**

Quote:

Or maybe the birth of a new one? I think the full joseki could remain for situations with existing extensions because you don't mind making white stronger (though of course white might resist you taking the corner in sente by playing jump instead of hane or the double hane) but Master agrees the full joseki is good for white on an open board. But its new idea (or maybe not I didn't pattern search yet) is to play this partial version and tenuki.

The full joseki is surprisingly rare - fewer than 50 examples in the GoGoD database. That means the average pro will play it a vanishingly small number of times in his lifetime, and so presumably he will never bother to think about it very much. The hane may therefore be played because pros are on auto-pilot rather than for any good strategic reason (i.e they are thinking tactically, e.g. about avoiding ijime).

Cases of omitting the hane prior to Master/AlphaGo do exist, usually in connection with a ladder situation in the diagonally opposite corner. The first "pure" example is from 2006, and thereafter there are still only very few cases. One characteristic of most of these cases is that the adventurous tenuki player is a young low-rank pro (i.e. one less set in his ways?).

Not quite a propos, and it's just an impression (a strong one, but I didn't actually do a count) - playing the 3-3 invasion is generally common among Japanese and Chinese players, but Korean players seem usually to shy away from it.

Uberdude · **#12**

Going back to the first 2 games, how should black play to preserve the influence of the wall if tenuki is a mistake. Here were my comments on reddit, plus some diagrams

Uberdude wrote:

redditor wrote:

The joseki finishes locally with white sente, and AlphaGO invariably makes a two space extension from the solo white stone to neutralize influence. If the consensus is that white has it too good after the extension, then you have to conclude that the corner joseki is bad for black, since at no point does black have an opportunity to prevent the extension.

Like you say white ends with sente to make this first extension to hoshi. I don't think it's (much if at all) bad for black yet, though it's worth pointing out in the first game white has the adjacent corner so this hoshi stone is rather safer than in the second where it is a black corner. Maybe black should only go for this wall building variation (i.e. the table shape as white can choose to push through, or maybe even the original pincer) if he has the adjacent corner. My point was black then tenukid from this white move so white got to play the next move from this hoshi stone which turned the balance of power with the influence.

In the first game black played the jump in lower left and white answered with the elegant large knight jump to the centre. After some more sente exchanges at the lower left black split white from the corner and white played the small knight and black then cut and started a fight. So my musings are should the jump of 25 have been something like q7 on the right directly. Once white got those 2 extra moves with the hoshi the wall felt sad. Another idea would be to answer the q10 hoshi with o10 cap. This feels rather O Meien style to me, and maybe the sort of move AlphaGo likes, but it also feels a rather slack exchange to allow white a 4th line for 6th line exchange. But if white then defends at r6 black can play the e6 jump to develop the huge top left through centre moyo, and if white plays around the lower left then black can split the hoshi from the corner and make o10 work in the attack so maybe it's ok.

In the second game black's counter pincer of 35 feels good and white jumps out with the small knight move, but he then tenukis for an approach of 37 which I think is likely a mistake: continuing the fight on top looks urgent gives how nice Master's 38 is. I'm not sure where, maybe k13?

redditor wrote:

Rather than exchanging the hane on the second line and forcing white to capture, Black crosscuts first, and allows white to exchange the hane for a turn. Presumably this allows black to maintain more influence?

Yes, black gets stronger influence by blighting the outside white stone more, but white also gets more territory. The problem is if this is better for black than when he gets the second line sente hane-connect, white could simply not play it and then black may well get it in sente after the cut.

First game:

Click Here To Show Diagram Code: [go]$$Wcm24 split, then what? 26 knight's move again? then a or b? $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . X . . . . . . . . . O . . O . . | $$ | . . . , . . . . . , . X X X O O . O . | $$ | . . . . . . . . . . . . . O X O O X . | $$ | . . . X . . . . . . . . . . X X X X . | $$ | . . . . . . . . . . . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . 3 . . . . . | $$ | . . . , . . . . . , . . . . . 1 . . . | $$ | . . . . . . . . . . . . . b . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . 2 . . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . O . . . . . , . . . a . O . . . | $$ | . . . . . O . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

Click Here To Show Diagram Code: [go]$$Wcm24 cap? variation 1 $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . X . . . . . . . . . O . . O . . | $$ | . . . , . . . . . , . X X X O O . O . | $$ | . . . . . . . . . . . . . O X O O X . | $$ | . . . X . . . . . . . . . . X X X X . | $$ | . . . . . . . . . . . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . , . . . . . , . . . 2 . 1 . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . X . 4 . . . . . . . . . . . 3 . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . O . . . . . , . . . . . O . . . | $$ | . . . . . O . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

Click Here To Show Diagram Code: [go]$$Wcm24 cap? variation 2. Maybe 27 at a? $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . X . . . . . . . . . O . . O . . | $$ | . . . , . . . . . , . X X X O O . O . | $$ | . . . . . . . . . . . . . O X O O X . | $$ | . . . X . . . . . . . . . . X X X X . | $$ | . . . . . . . . . . . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . , . . . . . , . . . 2 . 1 . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . 3 . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . 4 . . . | $$ | . . X . . . . . . . . . . . . . a . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . O . . . . . , . . . . . O . . . | $$ | . . . . . O . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

Second game:

Click Here To Show Diagram Code: [go]$$Wm34 Is this the vital point? how about counter at a? $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . O X X . . . . . . . . . . . . . | $$ | . . O . O X . . . . . . . . . . . . . | $$ | . . . O O X O . . 1 . . 2 . . X . . . | $$ | . . . O X X . . . . . . . . . . . . . | $$ | . . O X O . . a . . 3 . . . . . . . . | $$ | . . . X . . . . . 4 . . . . . . . . . | $$ | . . . X . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . , . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . O O X X X X X . , . . . . . O . . . | $$ | . . . O O O O . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

Uberdude · **#13**

John Fairbairn wrote:

The full joseki is surprisingly rare - fewer than 50 examples in the GoGoD database. That means the average pro will play it a vanishingly small number of times in his lifetime, and so presumably he will never bother to think about it very much. The hane may therefore be played because pros are on auto-pilot rather than for any good strategic reason (i.e they are thinking tactically, e.g. about avoiding ijime).

Whilst I agree you see this joseki in a higher percentage of weaker players' games, not having pros play it doesn't mean they don't think about it much. Just like you don't often see a pro game with a kill at the vital point of a bulky five or straight three doesn't mean they didn't think about a variation that lead to it (often by at a glance recognition), see it was bad for them and thus not play those moves. Similarly the 3-3 invading pro player will presumably not play the 3-3 when this simple joseki gives him a bad result, but will when it gives him a good result, and thus the 4-4 player will often choose a different variation to avoid it if it is bad for him. Those case where it is balanced (or the players disagree who it is good for, maybe one values territory more and the other values influence more) will be less common, hence the scarcity. On the other hand weaker players will often play the 3-3 when the wall is better, or build the wall when the territory is better so it's more common in their games (and it's one of the first joseki many people learn).

John Fairbairn wrote:

Cases of omitting the hane prior to Master/AlphaGo do exist, usually in connection with a ladder situation in the diagonally opposite corner. The first "pure" example is from 2006, and thereafter there are still only very few cases. One characteristic of most of these cases is that the adventurous tenuki player is a young low-rank pro (i.e. one less set in his ways?).

Very interesting!

John Fairbairn wrote:

Not quite a propos, and it's just an impression (a strong one, but I didn't actually do a count) - playing the 3-3 invasion is generally common among Japanese and Chinese players, but Korean players seem usually to shy away from it.

Perhaps this is partly down to Koreans liking to add the small knight's move to a 4-4 more than the others, so bare 4-4 situations arise less for them?

dfan · **#14**

I looked at the first diagram in this thread with Crazy Stone (which obviously is much weaker than Master!) and it had some interesting ideas. Here's the sequence given in the diagram, which I assume happened in most if not all of the games with this joseki:

Click Here To Show Diagram Code: [go]$$Wc $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . X . . . . . . . . . 9 . . 7 . . | $$ | . . . , . . . . . , . X 0 8 O O X 5 . | $$ | . . . . . . . . . . . . . O X 1 3 4 . | $$ | . . . X . . . . . . . . . . X 2 X 6 . | $$ | . . . . . . . . . . . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . , . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . O . . . . . , . . . . . O . . . | $$ | . . . . . O . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

After which White played at the middle right star point, reducing the influence Black's wall and even making it a target.

Crazy Stone would like to leave out the :b4:

-

and

-

exchanges and go straight to the cut:

Click Here To Show Diagram Code: [go]$$Wc $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . X . . . . . . . . . 5 . . . . . | $$ | . . . , . . . . . , . X 6 4 O O X . . | $$ | . . . . . . . . . . . . . O X 1 3 . . | $$ | . . . X . . . . . . . . . . X 2 X . . | $$ | . . . . . . . . . . . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . , . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . O . . . . . , . . . . . O . . . | $$ | . . . . . O . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

Now if White takes advantage of the absence of those exchanges to move down the right side, Crazy Stone is happy to strengthen its wall while being pushed from behind:

Click Here To Show Diagram Code: [go]$$Wc $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . X . . . . . . . . . O . . . . . | $$ | . . . , . . . . . , . X X X O O X . . | $$ | . . . . . . . . . . . . . O X O O . . | $$ | . . . X . . . . . . . . . . X X X 1 . | $$ | . . . . . . . . . . . . . . . O 2 3 . | $$ | . . . . . . . . . . . . . . . . 4 5 . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . , . . . . . , . . . . . , 6 . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . O . . . . . , . . . . . O . . . | $$ | . . . . . O . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

Meanwhile, leaving out those two exchanges before playing the cut means that Black now has a nice counter if White tries to play her wall-nullifying move and then respond in the corner in the same way as before:

Click Here To Show Diagram Code: [go]$$Wc $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . 5 . . . | $$ | . . . X . . . . . . . . . O . . 4 6 . | $$ | . . . , . . . . . , . X X X O O X 3 . | $$ | . . . . . . . . . . . . . O X O O 2 . | $$ | . . . X . . . . . . . . . . X X X . . | $$ | . . . . . . . . . . . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . , . . . . . , . . . . . 1 . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . O . . . . . , . . . . . O . . . | $$ | . . . . . O . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

So White has to respond differently to :b2:

:

Click Here To Show Diagram Code: [go]$$Wc $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . X . . . . . . . . . O . . 3 5 . | $$ | . . . , . . . . . , . X X X O O X 4 . | $$ | . . . . . . . . . . . . . O X O O 2 . | $$ | . . . X . . . . . . . . . . X X X . . | $$ | . . . . . . . . . . . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . , . . . . . , . . . . . 1 . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . O . . . . . , . . . . . O . . . | $$ | . . . . . O . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

and Black is healthier than before as well as having more territory.

I'm not really qualified to assess this analysis so I'm curious what thoughts stronger players have.

Uberdude · **#15**

dfan, you/CrazyStone are right that if black cuts first then white pretty much needs to continue with the 2nd line hane or else the corner suffers a lot: I'd overlooked pulling out of the atari and the 2 stone edge squeeze. So in that case black can cut first to ensure he gets a thicker wall but white gets more territory in return. Crazyone pushing again on the 2nd line is surprising though, the human joseki (rarely played though by pros) as I knew it is to tenuki (or defend the corner with s16 as there is some bad aji inside with s16 r18 otherwise so I guess CS is trying to defend against that in sente, but too much 2nd line crawl for my taste).

Uberdude · **#16**

Another interesting trait of Master is that if you approach its 4-4, it answers high, 3-3 then after pushing down it likes to tenuki instead of choosing blocking on one side or another. This 3-3 invasion is very common in the Chinese opening these days and normally black splits from the approach (e.g. http://ps.waltheri.net/database/game/72524/), but in the AlphaGo self-play games released several months ago we saw it liked to block the side so a few pros tried that (and I saw some using it in the ama Gold Cup), e.g. http://ps.waltheri.net/database/game/73339/. Another nice idea is to play some approach moves elsewhere as probes and depending on how the opponent answers you choose which side to block (e.g. here white answers with low move on top side, so that becomes less interesting, so black blocks the right side: http://ps.waltheri.net/database/game/74171/). Ke Jie tried this in a game but his opponent just ignored the approach to directly defend with the tiger mouth shape (or hanging connection, not sure what to call it). He lost that game, and I wondered if that defence was too good for white: viewtopic.php?f=13&t=13513.

>> Edit, see post below, including the sgf here caused some rendering bugs.

Anyway, the recent Master vs Ke Jie game above also saw this shape and Ke made the corner tiger mouth, allowing black to make his own tiger mouth from the shoulder hit at bottom left. Now I can't help feeling if the corner move is slow. Quite a lot of humans seem to play it vs Master (I will include more later) and for sure it is a huge move, kind-of reverse sente, but it gives Master a chance to play a speedy and flexible opening. If the thinking behind it is "normally black blocks here so I will take advantage/punish Master by connecting here" then maybe it's wrong. In Ke's game he did later invade the top side taking advantage of the lack of a base of a 4-4, but Master managed the situation well and ended up trading and killing the invader.

Just looking at the local shape, I think diagram 1 is worse for white than diagram 2. The hanging connection does have some slight advantage with regards to the top side, but the thinner connection to the approach has all sorts of bad aji (things like a-d). There was a spectacular example of this in fact from another AI recently: viewtopic.php?f=10&t=13921.

Click Here To Show Diagram Code: [go]$$B diagram 1 $$ ----------------+ $$ . . . . . . . . | $$ . . . . . . O . | $$ . . . . . O . . | $$ . . X . X X O . | $$ . . . . . . d . | $$ . . . . . O . . | $$ . . . . . b . . | $$ . . . . . a c . |[/go]

Click Here To Show Diagram Code: [go]$$B diagram 2 $$ ----------------+ $$ . . . . . . . . | $$ . . . . . . . . | $$ . . . . . O . . | $$ . . X . X O . . | $$ . . . . . . . . | $$ . . . . . O . . | $$ . . . . . . . . | $$ . . . . . . . . |[/go]

These shapes are also related to ignoring the slide which you can make some tewari analysis with against 3-3 joseki:

Click Here To Show Diagram Code: [go]$$B $$ ----------------+ $$ . . . . . . . . | $$ . . . . . . . . | $$ . . 3 . . 6 . . | $$ . . . . 1 . 4 . | $$ . . . . . . . . | $$ . . . . . 2 . . | $$ . . . . . . . . |[/go]

1-3 is joseki (if 3-3 is an okay opening move), but then 4 is probably aji-keshi but 5 is a slack answer as blocking at a is stronger. Black a for b would then likely be a bad exchange.

Click Here To Show Diagram Code: [go]$$W $$ ----------------+ $$ . . . . . . . . | $$ . . . . . . b . | $$ . . 2 . . 1 . . | $$ . . . . 4 a 5 . | $$ . . . . . . . . | $$ . . . . . 3 . . | $$ . . . . . . . . |[/go]

P.S. Haylee just did a review of this game here:

Uberdude · **#17**

Here's the Ke Jie vs Master game:

And Park Junghwan also making the connection:

And then Master as white answering Chen Yaoye's approaches and not taking gote to make the connection, Chen then blocked (and later Master did the attach on top of the Chinese opening stone to sabaki there as also seen in AlphaGo self-play games).

Another game with the same opening (rotated) as vs Park Junghwan above, but now Ke Jie is black and when he develops the moyo on move 13 Master dives in to make sabaki rather than connecting and allowing an even bigger moyo like Park did. It even ignores the block there later! So flexible.

Fan Tingyu making the connection, Master plays fast-paced opening. Later closes the side and then does the 2nd line poke at the thinness and ends up cutting the approach stone off!

pookpooi · **#18**

so anyone can sum up v.25(rumored) style yet, from these 60 games...
As v.13, Fan Hui said it's calm and conservative while Ann Younggil described it as slightly territorial but well-balanced.
In v.18 Michael Redmond commented that it's more aggresive and unorthodox.

Uberdude · **#19**

Uberdude wrote:

Another of Master's favourite moves is the 2 space high corner enclosure from a 3-4, which is pretty rare to see in normal early-board situations in contemporary pro play. I wonder which pro will be the first to copy it?

. Will it become a new fashion, like AlphaGo game 5 opening or that peep?

Well, I suppose Huang Yunsong did it before Mi Yuting in the Mingren semifinal yesterday!

Click Here To Show Diagram Code: [go]$$Bcm1 Huang Yunsong 6p (black) vs Mi Yuting 9p $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . 8 . . . . . | $$ | . . . 2 . . . . . , . . . . . 1 . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . 9 . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . , . . . . . , . . . . . 7 . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . 6 . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . , . . . . . , . . . 5 . , 3 . . | $$ | . . . 4 . . . . . . . 0 . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

Click Here To Show Diagram Code: [go]$$Bcm11 $$ +---------------------------------------+ $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . 9 . . . . . . . . | $$ | . . . . . . . 4 . 1 8 . . O . 5 . . . | $$ | . . . O . . . . . 7 6 . . . . X . . . | $$ | . . . . . . . . . 0 . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . , . . . . . , . . . . . X . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . O . . . . . . . . . . 3 . . . . | $$ | . . . . . . . . . . . . 2 . . . . . . | $$ | . . . , . . . . . , . . . X . , X . . | $$ | . . . O . . . . . . . O . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ +---------------------------------------+[/go]

dfan · **#20**

People following this thread now or in the future may like to know that there is another interesting discussion in progress of Master's style in the thread translation of a 9d pro commentary, starting with post #36.

Learning from the Master

Who is online