Brilliant long ijime sequence

[John Fairbairn]

The game below contains perhaps the most brilliant ijime (bullying) sequence I have ever seen, and it includes a couple of nifty tsumego themes by none other than the God Of Tsumego, Maeda Nobuaki.



Go to move 84. In this position White has virtually no territory (not even komi) and even has a couple of dicey groups. Black has settled territory and safe looking groups - except that he has six of them.

White 84 strikes me as a very AlphaGo-esque move, with White having to work out its relationship with virtually every part of the board.

See how, over the space of over EIGHTY moves, White accrues tiny increments of territory (by bullying) while allowing Black virtually none, and in the process White even firms up his two dicey groups.

At the end of the 80+ moves, the sixth Black group succumbs to the proverb that tells us a sixth group must die. But note also the tsumego themes in the top half of the board and they way they exert a domino effect from group to group.

For good measure, at move 84, given both the balance of territory and the status of the White group at the bottom, many players would just connect at O2, as Maeda pointed out. He gave an unusually long comment about 84, which perhaps indicates he was rather proud of it. However, he was aided by a poor Black 85 - after that it was a downward spiral. Black should have omitted the 85-86 exchange and gone straight to 87.

I do hope such a coruscating example is not wasted in what currently appears to be more a graveyard than a forum

[Uberdude]

Nice example John, I like how the attacks ricochet around the board. A question on terminology though: why do you call this ijime/bullying rather than attacking? To me the English word bullying conjures up feelings of forcing the opponent to respond submissively, teasing a group that's already fairly safe but is sad to defend, rather than the large-scale attacking strategy seen here (though black is forced to submit several times).

I do hope such a coruscating example is not wasted in what currently appears to be more a graveyard than a forum

I've been quiet recently as I was on holiday in Thailand, and next week am off to the Isle of Man Go festival. Coincidentally the 4th International Collegiate Go Tournament was held in Bangkok whilst we were there so we crashed the closing dinner (my wife knows some of the local Thai organisers) and bumped into last year's British champion Junnan Jiang and Stephen Hu. The tournament was won by a Japanese player, 2nd was a Chinese and 3rd was Korean ex-insei studying in USA Mark Lee who swept the 2014 US Open beating quite a few low-dan pros, so that gives an idea of the high level. The semi-finals of the Chinese Changqi Cup (also sponsored by the Ing foundation) were held there too so I got to see Jiang Weijie 9p (who was in rather higher spirits than Chen Yaoye 9p who he had just beaten by a whopping 13 points) joining in the Thai conga (Tan Xiao 9p beat Gu Lingyi 5p in the other):

JiangWeijie.jpg (169.55 KiB) Viewed 7575 times

[FuriousGeorge]

This is a fascinating game. Normally at my level (being so low) such aggression is thoughtless and without a plan. I'm extremely confident that Nobuaki had a plan, but what level was calculated and what level was emergent over those first 80 moves?

...and to high level players, how solid are your plans entering a game? Do you say, 'I'm going to try to split them into many groups' or 'I will play defensively and force them to overextend' or does it take shape as the moves proceed?

In a way, this game does feel like taunting, like Nobuaki just wanted to see what would happen if he kept prodding.

[John Fairbairn]

but what level was calculated and what level was emergent over those first 80 moves?

Maybe you mean the second 80 moves (after 84), but these players had 11 hours each and used every minute (not to mention whatever thought they also gave overnight, so I think we can safely assume an awful lot was calculated.

Incidentally, the game went on till around move 250 before Black resigned, and in that subsequent White killed the Black group at the top and White killed the White one on the lower left side (a trade but the sort of trade it's hard to make in modern speedy go (unless you are AlphaGo). What brought on the resignation was that either killed or bullied the lower right corner. It's not clear because the record is defective at that point.

[Bill Spight]

Wonderful, John! Thank you.

In addition to , there are a number of interesting plays. I have selected a few to comment on.

At first glance, it looks like with Black may be out of the woods, but is a very nice play. True, it looks like White will have to play somewhere around there, but that seems to be the perfect spot, working with later White plays, including

White 104. Which is one of those plays about which you say, Oh, of course — once you have seen it.

White 120 is another wonderful attacking move that is in just the right spot.

Earlier moves of note:

which lets Black into the corner and takes gote, besides.

which is what I think of as bullying.

which takes away the eye potential of the Black group.

the large scale attack, which sets the stage for everything else. Attacking on a large scale makes it hard for Black to throw the F-03 stone away. And, indeed, he does not, instead, replying with J-04. But those stones soon become heavy. OC, the extension to M-03 instead invites White to jump to K-05, which would be very painful. I wonder: Is this a place for an AlphaGo attachment at K-04?

[jeromie]

These were my thoughts on the board state at 84 (before looking at the remaining moves):
Black Groups:
Upper left corner - Safe, but enclosed and susceptible to forcing moves. Relatively small amount of territory, especially if white gets to move here first.
Left side - Safe, but not much territory. Lots of second line moves.
Bottom left - In danger. if it gets cut off from the other black groups there will be lots of forcing moves available.
Bottom right corner - Looks safe, but with some possible peeps / other threats that could cause trouble
Top center - Depending on who moves here, this group doesn't look solid.
Upper right corner - Solid territory, but the tail on the side could get cut off.

White Groups:
Upper left corner - Looks safe
Lower left corner - Should be safe with the 3-3 stone, but it could definitely come under attack.
Bottom - Should be able to live, but could get sealed in if black gets a move on top.
Right - This looks weak to me. It's not in immediate danger, but it's definitely a target for attack.
Top right - Alive, but there are forcing moves that can be used against it.

So by my (weak) accounting, black has two groups that could come under attack and two that have definite forcing moves. This seems like a good time to begin a bullying strategy, and it doesn't surprise me that white was successful. (Though of course I could not successfully do that myself in this situation, and probably wouldn't have noticed the board state if I didn't know what was coming.)

Thanks for the post, John. I always find your contributions interesting!

[alphaville]

Go to move 84. In this position White has virtually no territory (not even komi) and even has a couple of dicey groups. Black has settled territory and safe looking groups - except that he has six of them.

We must be using different definitions of "territory" - what do you mean when you say white has none, "not even komi"? I count 30 points or so for white.

[Kirby]

Go to move 84. In this position White has virtually no territory (not even komi) and even has a couple of dicey groups. Black has settled territory and safe looking groups - except that he has six of them.

We must be using different definitions of "territory" - what do you mean when you say white has none, "not even komi"? I count 30 points or so for white.

I'd count about 16 for white at that position.

[Javaness2]

Where was Black's mistake?
Was the idea ending in :black:83 simply too slack? Did Black need to pressure the group at Q7 instead?
Was :black:85 start closing down connections? My reflex is a move at L6, but I can't claim to have read anything out in any depth.

Once :white:M12 appears, with the useful looking :white:N6,O6 already in place, the alarm bells start to become audible.