Uberdude has been doing a great service recently in highlighting trends he spots in the play of AIs. He is the Keats of go:

Much have I travell'd in the realms of gold,
And many goodly states and kingdoms seen;
Round many western islands have I been
Which bards in fealty to Apollo hold.
Oft of one wide expanse had I been told
That deep-browed Homer ruled as his demesne;
Yet did I never breathe its pure serene
Till I heard Chapman speak out loud and bold:
Then felt I like some watcher of the skies
When a new planet swims into his ken;
Or like stout Cortez when with eagle eyes
He star'd at the Pacific — and all his men
Look'd at each other with a wild surmise —
Silent, upon a peak in Darien.

Slightly guiltily, I like to point out that humans often beat AIs to the punch, in the same way that Vasco Núñez de Balboa was overlooked in favour of Cortés.

Here's a good example. Ubderdude spotted interesting uses by AlphaGo of the nozomi on the 5th line. It is Iwamoto being just as interesting 80 years ago.


If you want to test yourself, go to move 29, then try guessing White 30.






Iwamoto played his nozomi because he did not want to see Black turn the upper side into thickness, but he regretted his choice as he didn't really know how to continue. It seems to me that that is the crux. AIs don't necessarily come up with better ideas than humans, and certainly find little, that is truly new. But they do seem to know how to continue in the centre. I look forward to someone getting a better handle on that.

That move reminds me of a shape Matthew Macfadyen 6d seems to like: the same large knight's move to the 5th line but as black from his 4-4. He likes to use it where black is building moyos on both sides so similar to this but where I might consider a more normal move like the one point jump to d14 which makes the 3-3 harder as well as reinforcing the moyo. Matthew's plan tends to be to allow the opponent to invade one side and then fight, but turn the other into cash and make it bigger with the 5th line move. My feeling is it is often not really a sound way to play, but it's an unfamiliar shape to most and leads to fights which are easy to mess up against Matthew unless you are super strong.

I'm not familiar with the term "nozomi" (only as a kind of shinkansen). Is this a typo or a synonym for "nozoki" or a different kind of move?

Ah, thanks.

The meaning depends on the kanji, of course. I'm more familiar with the meanings from normal use: wish, desire, hope (望み) or challenge, confrontation (臨み).

Once your attention is drawn to something you never normally notice, you seem to see it everywhere afterwards.

Here is an example not of nozomi, but of 5th-line play nevertheless.



Just in passing first, Black 17 is a kind of 5th-line move that was very common in old games, and when I first started to play go I was truly mystified by such moves. Over 50 years later I am still truly mystified.

But revenons à nos moutons: Genjo playing White against Chitoku. The significant move is White 24. Yoda says the expected move would be K17, but Genjo was of Meijin stature and played on the 5th line. He did tend to favour open spaces anyway, but then he couples this with another 5th-liner at 26. Suddenly the game, to me, looks like an AI playing White.

It seems like a great idea, but once again the centre play proved too difficult. It worked to some extent - the Black group at the top lived with a measly 2 points and White did get most of the centre, and it does look towards exploiting Black's mistake at 19 (should be on the upper side - Yoda). But he couldn't make that work in combination with what was happening elsewhere on the board and he ended up losing by a whopping 11 points.

and remind me of me. {sigh}

_________________
The Adkins Principle:

At some point, doesn't thinking have to go on?

— Winona Adkins

I think it's a great idea to talk during sex, as long as it's about snooker.

— Steve Davis

Something that makes appeal to me is it makes it harder for black to attack W's shape at o15 as a way of managing his group; W can resist more strongly now.

The point O15 was certainly raised in the commentary, but Black got there nonetheless, and that was part of the reason for White's loss.

It was a strategic loss - no obvious blunders. To try to convey what happened in the final position shown above, White attacked both the left side and upper side groups but in neither case was it true ijime and the result was that his left corner group had an open skirt on the left side (i.e. no certain territory or eyes) when Black descended to B15 to make eyes. Because of Black O15 the right White side group was also weak (and it also meant that White's forces against Black's corner there were almost meaningless).

The result of these joint weaknesses was that White ended up fairly soon (move 56) having to connect his two groups at the point to the right of 26. This was connecting on a dame point, in effect giving Black a free move. Since White was also thus in gote, Black got to the lower right first - he chose to just dump some aji there and then extend on the lower side, as he feared giving White more thickness if he played the corner out at once.

In other words, all the parts of the board were truly inter-related, a tough scenario even for a meijin. White's task is to make something out of nothing. Black's task is to make nothing out of something - a whole lot easier.