Knotwilg's practice

[Knotwilg]

The memory exercises continue, now the "list of tour de France winners" is being absorbed, containing 109 items.
On a whole other level, I'm also trying to master the two solos in the poprock masterpiece "Sultans of Swing".

Meta observations:

- Rote memorizations for quizzes are perfect fodder for car drives (commuting is the place & time)
- Guitar playing works best in an isolated, focused environment (in casu my second house in the Ardennes)

This leads me to think that a learning plan should not only focus on what & why but also on when & where. You can plan for "deliberate practice" as much as you want, if you don't take into account the conditions that will be necessary or beneficial to the activity, you'll likely end up with a miss.

[Knotwilg]

[Knotwilg]

BTW, since I've been visiting Taipei (I assume the past continuous is the way to describe I will do so again in the future) I've ventured into studying Chinese. I haven't advanced much beyond Ni Hao and Zai Jian (tones omitted), counting 1 to 10 and saying "Wǒ hěn gāoxìng lái dào zhèlǐ". First time I did, I got the obligatory accolades, while I probably said something like "I'm jealous for apple tree sleepwalking"

[Farodin]

Best of luck with it! I remember from another thread that you play guitar and so have a musical ear, in which case tones will come quite easy to you. At least they did to me, as a long-time piano player.

I've been studying Chinese for around 6 years already, so let me know if you need some resources/tips. Having been to Taiwan several times, I happen to be much more familiar with the Taiwanese accent & traditional characters than their mainland counterparts.

[Knotwilg]

Here's a kind of middle game joseki, arising from a san ren sei

$$W Invasion
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . O . . . . . |
$$ | . . . . . . . . . . . X . O . O O . . |
$$ | . . . O . . . . . , . . . . X X X . . |
$$ | . . . . . . . . . . . . X . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . X . . . |
$$ | . . X . . . . . . . . . . . . 8 4 . . |
$$ | . . . . . . . . . . . . . . 7 2 1 . . |
$$ | . . . . . . . . . . . . . 0 6 5 3 . . |
$$ | . . . . . . . . . . . . . . 9 X . . . |
$$ | . . . X . . . . . . . . . . . . . . . |
$$ | . . X , . O . . . , . . . . . X . . . |
$$ | . . X O . . . . . . O . . O . . B . . |
$$ | . . O . O . . . . . . . . . . O . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------

Click Here To Show Diagram Code

[go]$$W Invasion
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . O . . . . . |
$$ | . . . . . . . . . . . X . O . O O . . |
$$ | . . . O . . . . . , . . . . X X X . . |
$$ | . . . . . . . . . . . . X . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . X . . . |
$$ | . . X . . . . . . . . . . . . 8 4 . . |
$$ | . . . . . . . . . . . . . . 7 2 1 . . |
$$ | . . . . . . . . . . . . . 0 6 5 3 . . |
$$ | . . . . . . . . . . . . . . 9 X . . . |
$$ | . . . X . . . . . . . . . . . . . . . |
$$ | . . X , . O . . . , . . . . . X . . . |
$$ | . . X O . . . . . . O . . O . . B . . |
$$ | . . O . O . . . . . . . . . . O . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

When W1 invades, there is room to both sides, so Black rather connects. If W3 Black can block thanks to the presence of the marked stone.

$$W Double atari
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . O . . . . . |
$$ | . . . . . . . . . . . X . O . O O . . |
$$ | . . . O . . . . . , . . . . X X X . . |
$$ | . . . . . . . . . . . . X . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . X . b . |
$$ | . . X . . . . . . . . . . . . X X a c |
$$ | . . . . . . . . . . . . . . O X O . . |
$$ | . . . . . . . . . . . . . X X O O . . |
$$ | . . . . . . . . . . . . . . O X 3 . . |
$$ | . . . X . . . . . . . . . 4 2 1 . . . |
$$ | . . X , . O . . . , . . . 0 . X 7 5 . |
$$ | . . X O . . . . . . O . . O . 8 X 6 . |
$$ | . . O . O . . . . . . . . . 9 O . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------

Click Here To Show Diagram Code

[go]$$W Double atari
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . O . . . . . |
$$ | . . . . . . . . . . . X . O . O O . . |
$$ | . . . O . . . . . , . . . . X X X . . |
$$ | . . . . . . . . . . . . X . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . , . . . . . , . . . . . X . b . |
$$ | . . X . . . . . . . . . . . . X X a c |
$$ | . . . . . . . . . . . . . . O X O . . |
$$ | . . . . . . . . . . . . . X X O O . . |
$$ | . . . . . . . . . . . . . . O X 3 . . |
$$ | . . . X . . . . . . . . . 4 2 1 . . . |
$$ | . . X , . O . . . , . . . 0 . X 7 5 . |
$$ | . . X O . . . . . . O . . O . 8 X 6 . |
$$ | . . O . O . . . . . . . . . 9 O . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

The double atari of B2 is a tesuji. White must live in [gote] with ''a'' to ''c''. Black has sente to play in the upper left.

Although White's invasion has succeeded, it's with small life and the upper half of Black's original sphere of influence is intact, even reinforced. The skirmish has also slightly damaged White's potential at the bottom. Black comes out with sente, so this is an acceptable result.

[kvasir]

Black comes out with sente, so this is an acceptable result.

Acceptable for black, yes. Isn't this just good for black? That is white could instead reduce, get stones in the center and then turn attention to something else with sente. Sente could be big in this game and in the diagram black has sente and is posed to make good use of it, black is dominant on the outside and the center, there are also soft spots that black can aim to pressure.

Just my 2cents, I haven't spent much time considering the position or if my initial idea on how to continue from the diagram makes sense

[Knotwilg]

Given my reflections in another thread on "How to improve at Go" I'm taking a challenge for 2023.

How do YOU think I should improve? First (serious) advice I'll take without any quibbles into January. And so on.

In absence of advice, I'll play longer matches, immediately review them from memory and also read a book I have never read (any tips welcome too). I may still review the review with KataGo but no longer take the lazy route.

[dust]

I may still review the review with KataGo but no longer take the lazy route.

Oh, I was going to suggest a 'lazy route'. Play a game or two against KataGo a day, and try to reduce the handicap stones (I'd suggest maybe starting at 5 stones as 2D).

It's quite likely you'll improve if you review your games, and you'll at the very least keep 'match fit'.

[Knotwilg]

I'll do what you suggested, dust!

In the meantime, here's already a self review, replayed from memory the day after
Then I'll add a KataGo review of review



And then the review of the review. Key points are 22 and 51

[John Fairbairn]

How do YOU think I should improve? First (serious) advice I'll take without any quibbles into January. And so on.

Dieter:
I may not be first but I'd still like to suggest an experiment that shouldn't interfere with any other path you take. You're a numbers guy. How about being a words guy for a while?

I'm keen to see the result because I don't play and so don't study, and that's not likely to change. Which means I can't do the experiment on myself. Also, my prime aim is to make historical commentaries available and entertaining. What I am about suggest is just a spin-off.

The background is that I am trying to plot the evolution of go theory through ages past by using master commentaries. Since only the Chinese produced such books, all my material so far has been literary Chinese (which, in practical terms, is a good thing, I think). To do this, I have had to turn into a little bit of a numbers guy in that I have made a corpus of all the texts I've used. I can use this in several ways. One is simply to discover which are the most frequent terms. Another is to discover at which phase (via move numbers) the terms tend to occur (so that I am able to confirm, for example, that boundary plays occur far more often and much earlier in the game than people who believe yose=endgame might think). Other uses are to hypothesise about nexuses of terms, which is what I would like to focus on here, and to plot changes in theory over time.

I have just finished the latest book in my Museum of Go Theory series, Brush, Ink, Go (Yi Mo), and have sent it off for proof-reading. This gave me a lot of data because (a) the compiler, a master player, selected 100 games with 26 different players covering a time span of several decades, and (b) he appended commentaries that were longer than in previous works. He also selected these 100 games from a "database" of 600 on the basis of their entertainment value and so there is great variety in the games.

Of course, we have to remember that group tax applied and that must skew the results somewhat compared to modern play. As an example of the sort of way we can easily overlook how group tax runs through the whole game, tactics and strategy, consider that wedges are rather more common than in Japanese games. This make sense once you realise they lead to cuts, which trips the group-tax alarm buzzer. But I believe that is not too much of an impediment, and in any case is more than cancelled out by the fact that old Chinese go looks (to my untrained eye) remarkably similar in strategic terms to AI play. In other words, my hypothesis is that if we can learn something about old Chinese go, we can learn something deep about AI play.

The 100 games in Brush, Ink, Go have yielded 506 instances of comments on major concepts.

A whopping 104 relate to INFLUENCE. I am treating that as a nexus on its own. There is an awful lot to say about it, but that belongs in the book. For now, though, I think an experienced player can get by by remembering that it does NOT mean thickness (though it can cover thickness), and whenever it occurs, you should keep the word 'power' in the back of your mind.

The second nexus is rather large because it spans a range of what I will call ASSERTIVENESS. At one extreme is encompasses forcing moves (11 occurrences), and at the other there is 'restraint' which combines self-restraint with restraining the opponent, rather in the way that a honte does (2 occurrences). But in the middle are two concepts that each have over 30 occurrences. These are what I call CROWDING IN and PRESSURISING. Crowding in you might best understand as kettling, blocking off avenues of escape so that eventually the opponent's stones tend to crowd together. But at this stage he will have a fair amount of scope for varying or delaying his responses. Pressurising is when you are, as it were, in the opponent's face. He has to answer more or less at once but does have some scope for how he responds, unlike with a forcing move. For the purposes of the experiment, I think we can go with just thinking of all these terms as a group of assertive plays.

But, more controversially, perhaps, I would add the 'initiative' to this nexus. I think it is important to learn that sente (as we use it) is different from the much higher and more important concept of the initiative, which you have if you choose where the next play will resume, whether or not it is your turn to move. In that sense, having the initiative (in the true or Chinese sense) can even be like playing tenuki. But, one way or another, you are in control of where play resumes and so, in my mind, this - falls under the heading of ASSERTIVENESS. - 'Initiative' has a whopping 48 occurrences - you don't see sente mentioned that often strategically in modern commentaries!

This nexus is the one that most often makes me say, whenever I play over an old Chinese game, "Oh, just like AI."

The next nexus is DEALING WITH TERRITORY. One big item in this is 'encroachment' (57 occurrences) which I have talked about elsewhere. But the biggest element is what I call 'walling off', which deals with territory by explicitly making boundaries, and not just in the endgame - not even mostly in the endgame! I would be happy to argue that this is an area of that is sadly and almost entirely lacking in western play. Another item in this nexus I call SHARING. This is often used of the corners and the sides, i.e the area most likely to yield actual territory early in the games, and it refers to the action of divvying up the corner or side so that each player gets more or less his fair share. This has 24 occurrences but in one sense is more common than that, in that old Chinese joseki books show lots of cases of corner and side josekis that boil, down to sharing territory (as opposed to the old Japanese chestnut 'prog=fit versus thickness'). I also add to this nexus 'trades' and 'sacrifices'.

A fourth nexus is SAFETY. The number of references to settling a group (an evaluative term rather than a technique) is 34 (again compare that to modern commentaries), but we can also add to this nexus 'bases' (16 occurrences) and eye spaces (9).

A fifth nexus I call VANTAGE POINTS. It includes pure vantage points or big points such as Tennozan moves, of course, but also 'call & response' moves (11 occurrences) and what I call 'pitch-ins' (20 occurrences). These are lone moves into enemy areas (especially empty areas), and include invasions, splitting attacks (wariuchi) and even probes.

There are other possible nexuses but that will do for an experiment. The way I envisage using them is simply to list the elements of each nexus on a piece of paper (or on index cards) and to refer to the list as memory joggers when you play or review. The idea is to get all the elements thoroughly into your mindset so that they are available permanently and simultaneously. You can of course adjust the lists to suit your own interests. Part of the results of the experiment would be an assessment of to what degree this takes place and to what degree it helps you (not just your technical skill but you general alertness).

Another result I would like to see relates to the reasoning behind treating the elements as nexuses. It would be useful to know the extent of the benefit you get from any nexus. In other words, which are the most important nexuses.

There are some minor results that would be of interest. E.g. do you agree with me that this leads you to thinking more like an AI bot?

In any event, I don't see that this interferes with any other method of study, and I'd expect even a month would give some indication of whether it can help.

[Knotwilg]

Dieter:
I may not be first but I'd still like to suggest an experiment that shouldn't interfere with any other path you take. You're a numbers guy. How about being a words guy for a while?

Thanks for that suggestion too, John. I'll reply to your summary first: in my view, I already am a words guy. Whenever you speak about numbers guys, I don't feel addressed, although you seem to imply I am one. Yes, I use KataGo and yes I take its evaluation of point differences as a criterium. But that's just a small part of the way I approach Go and its study. I've been very involved in concepts and heuristics. We wouldn't have had all those debates about influence, thickness, English vs Japanese terms over the years if words would have little value to me. And would a numbers guy come up with something like "slonection"? But let me go into the details of your - highly appreciated - advice:

(...)
In other words, my hypothesis is that if we can learn something about old Chinese go, we can learn something deep about AI play.

I've seen this theme in your comments. It's an interesting thought, though not one that I can buy into just yet.

The 100 games in Brush, Ink, Go have yielded 506 instances of comments on major concepts.

A whopping 104 relate to INFLUENCE. I am treating that as a nexus on its own. There is an awful lot to say about it, but that belongs in the book. For now, though, I think an experienced player can get by by remembering that it does NOT mean thickness (though it can cover thickness), and whenever it occurs, you should keep the word 'power' in the back of your mind.

Well here we are again .

Since those debates, I have basically removed the term thickness from my repertoire of nexuses and replaced it with local strength, if I want to "play away from thickness" for example, or global strength, if I want to evaluate who is ahead and not just count territories. Those are useful concepts to me. Influence, to me, is a much more basic concept, that simply talks about the fact that stones in a certain arrangment (a wall, a pommuki) affect the open space nearby. Influence by itself is not "thick" or "locally strong enough" to be invulnerable to stones that "encroach" (on?) it. So a wall may use an extension, and a ponnuki on the side likewise, in order to start forming a framework of sorts. Essentially, while influence can be used to attack, it can also be attacked.

A good example can be seen in the game I just published here: the lower black corner is thick, so I should play away from it. White's not-really-a-ponnuki against it has influence but still I decide to attack it, by checking it towards my thick corner.

The second nexus is rather large because it spans a range of what I will call ASSERTIVENESS. At one extreme is encompasses forcing moves (11 occurrences), and at the other there is 'restraint' which combines self-restraint with restraining the opponent, rather in the way that a honte does (2 occurrences). But in the middle are two concepts that each have over 30 occurrences. These are what I call CROWDING IN and PRESSURISING. Crowding in you might best understand as kettling, blocking off avenues of escape so that eventually the opponent's stones tend to crowd together. But at this stage he will have a fair amount of scope for varying or delaying his responses. Pressurising is when you are, as it were, in the opponent's face. He has to answer more or less at once but does have some scope for how he responds, unlike with a forcing move. For the purposes of the experiment, I think we can go with just thinking of all these terms as a group of assertive plays.

Applying pressure is indeed a wording I have started using when thinking about attacking. "Attack" has this notion of killing, despite all the good advice we got that we shouldn't think about killing, rather "attack to make profit". It's more subtle and "applying pressure" conveys this subtlety. The profit can be influence or potential territorym but it can also come from "crowding in" as you say, which relates to the idea of overconcentrating the opponent, or surrounding them.

But, more controversially, perhaps, I would add the 'initiative' to this nexus. I think it is important to learn that sente (as we use it) is different from the much higher and more important concept of the initiative, which you have if you choose where the next play will resume, whether or not it is your turn to move. In that sense, having the initiative (in the true or Chinese sense) can even be like playing tenuki. But, one way or another, you are in control of where play resumes and so, in my mind, this - falls under the heading of ASSERTIVENESS. - 'Initiative' has a whopping 48 occurrences - you don't see sente mentioned that often strategically in modern commentaries!

For me "having sente" is having the right to play elsewhere, "taking sente" is claiming to have such right and "keeping sente" is playing in a way that the opponent can't "take sente" next but needs to answer.

This nexus is the one that most often makes me say, whenever I play over an old Chinese game, "Oh, just like AI."

That's interesting. I don't know about old Chinese games but continuing on the previous, I see AI play elsewhere (taking sente) more often than conventional/amateur wisdom, not because groups are stronger than we think but because they are lighter than we think. AI are much more versatile in their decisions to sacrifice groups, because they foresee how this will happen in a way that the opponent needs to invest a lot to capture even large groups, and the sum of the forcing moves they got out of it, plus the benefit of having taken sente, outweighs the sacrifice.

The next nexus is DEALING WITH TERRITORY. One big item in this is 'encroachment' (57 occurrences) which I have talked about elsewhere. But the biggest element is what I call 'walling off', which deals with territory by explicitly making boundaries, and not just in the endgame - not even mostly in the endgame! I would be happy to argue that this is an area of that is sadly and almost entirely lacking in western play. Another item in this nexus I call SHARING. This is often used of the corners and the sides, i.e the area most likely to yield actual territory early in the games, and it refers to the action of divvying up the corner or side so that each player gets more or less his fair share. This has 24 occurrences but in one sense is more common than that, in that old Chinese joseki books show lots of cases of corner and side josekis that boil, down to sharing territory (as opposed to the old Japanese chestnut 'prog=fit versus thickness'). I also add to this nexus 'trades' and 'sacrifices'.

I am in the process of adopting encroachment as a better concept/wording than reduction, just like I'm still lacking words to dissociate an invasion that just intends to live small and take away potential territory, from an invasion that intends to weaken the structure around it.

The other idea about "sharing" can be novel to me and deserves more thought.

A fourth nexus is SAFETY. The number of references to settling a group (an evaluative term rather than a technique) is 34 (again compare that to modern commentaries), but we can also add to this nexus 'bases' (16 occurrences) and eye spaces (9).

There's a good example of this in the same game review. Move 137 - the one I didn't play.

A fifth nexus I call VANTAGE POINTS. It includes pure vantage points or big points such as Tennozan moves, of course, but also 'call & response' moves (11 occurrences) and what I call 'pitch-ins' (20 occurrences). These are lone moves into enemy areas (especially empty areas), and include invasions, splitting attacks (wariuchi) and even probes.

Too vague for me to respond to right now.

There are other possible nexuses but that will do for an experiment. The way I envisage using them is simply to list the elements of each nexus on a piece of paper (or on index cards) and to refer to the list as memory joggers when you play or review. The idea is to get all the elements thoroughly into your mindset so that they are available permanently and simultaneously. You can of course adjust the lists to suit your own interests. Part of the results of the experiment would be an assessment of to what degree this takes place and to what degree it helps you (not just your technical skill but you general alertness).

OK - I can do this with the concepts (nexuses) offered. The benefit I see is finding more and better candidates. In general I think my playing lacks depth rather than width, i.e. "seeing sequences" more than "seeing candidates", but it doesn't hurt to check that assumption by accepting your offer.

Another result I would like to see relates to the reasoning behind treating the elements as nexuses. It would be useful to know the extent of the benefit you get from any nexus. In other words, which are the most important nexuses.

I can do that by including them in the reviews, hence get a count. Just pleasing the number guy in you

Them =

1. "Influence" (and "local strength")
2. "applying pressure" and "crowding in" (= surround to overconcentrate)
3. being "assertive" in taking sente (by treating groups more lightly than I would usually do)
4. "encroach" (and the two flavors of invading, which I'll call "invade to live" and "invade to weaken" for now)
(I won't use the "sharing" nexus just yet, before I get a grasp of what is meant, or how it could be useful)
5. "safety" (not too sure about this one just yet, as it seems to go against the "assertiveness" idea)
(I won't use "vantage points" just yet, for similar reasons as "sharing)

That's 5 for you.

There are some minor results that would be of interest. E.g. do you agree with me that this leads you to thinking more like an AI bot?

I don't have that expectation and I don't think it's a fair one. Understanding why AI chooses a move, perhaps, but it's still our human interpretation of those choices, not replicating its thinking.

In any event, I don't see that this interferes with any other method of study, and I'd expect even a month would give some indication of whether it can help.

Sold!

[John Fairbairn]

Dieter:

Thanks for being game enough to take this on, even with your reservations. I treat you as a numbers guy BTW because you said you were a mathematician. That, in my experience, often implies a mindset that differs radically from a person who works with the effect of words, which is what I mean by a words guy. It is not meant to imply that the one can do things the other can't, or vice versa. It's just about the starting mindset.

I see the difference in your statement:

For me "having sente" is having the right to play elsewhere, "taking sente" is claiming to have such right and "keeping sente" is playing in a way that the opponent can't "take sente" next but needs to answer.

I may have quite the wrong impression, of course, so please forgive me if I am. But what I see is that, despite the use of sente in three different and potentially confusing ways (and there are more), you are seeking a neat and tidy way to differentiate them. But that leads to ignoring the word I'm using: "initiative", presumably because that would make a mess of things. I sense a similar resistance when it comes to "invasion" which is part of encroachment. You seem to have detached it, however. That may make things tidier, but it's plain wrong. Encroachment != reduction (it just includes it). My wider point is that it is natural for a words guy to embrace the messiness of adding extra terms such as initiative. If you have the appropriate mindset, you can learn a lot from sifting through the midden (and, in that vein, you can regard nexus as a posh word for mess, or if you don't want to be either posh or vulgar, as a lady's handbag).

To go back to initiative. Recall Muhammad Ali and rope-a-dope. The other boxer had sente. He was the one coming forward. Ali was the one backing on to the ropes. The other guy was the one throwing the punches. Ali had to defend (but smiled or talked as he did it). The other guy was given sente, had sente, and kept sente. But Ali had the initiative. He was dictating where the punches were traded, he was dictating how much energy each boxer expended. And so on. He had the 'upper hand' = initiative (and note, incidentally, the not-so-subtle difference between upper hand and voorhand).

Chinese "initiative" relates mainly to having the upper hand. It's what bots have - they care little about sente. If we look at how the term initiative is applied in practice, there is an inexact but still pretty reasonable Japanese equivalent BTW, and it's not sente. It's amashi. In other words, amashi = rope-a-dope.

[dust]

For Assertiveness and Initiative, I wonder if there is a meaning around shaping the overall flow of the game and how it proceeds?

I'm reminded of the extreme level of this in the classic title games between Cho Chikun and Takemiya, where neither player compromised in trying to steer the flow of the game.

[John Fairbairn]

For Assertiveness and Initiative, I wonder if there is a meaning around shaping the overall flow of the game and how it proceeds?

I'm reminded of the extreme level of this in the classic title games between Cho Chikun and Takemiya, where neither player compromised in trying to steer the flow of the game.

I think you are on to something, though I'm not quite sure what.

Around the time of Cho and Takemiya's dominance, it was very common to talk about not following the opponent's orders (chumon), and this resistance to the opponent was often described as kiai (matching fighting spirit). Maybe it was common whenever Takemiya played because his style was the opposite of the likes of Cho and Kobayashi. In any event, these words seem to have almost disappeared from recent commentaries (even before AI). I'm not at all sure what has replaced them. Maybe they are just taken for granted now.

That doesn't mean everyone plays the same way now. My sense is that the style of play you see in title games or league games nowadays is influenced by AI but is generally still quite sedate/"human". But the younger players, who probably study with AI even more, seem obsessed with disrupting the flow of the game - even the flow of their own moves! Is this a lack of skill or have they mastered chaos theory?! One possibly related thing I have noticed is that sacrifices have got bigger.

[Knotwilg]

So far I tried to play a couple of games with 7 "nexuses" in mind. It's too soon to call any result but both of them ended in bad fights for me.

I do find that the "sharing" comes to mind most and perhaps I'm taking that too much as "taking my share" in an area where the opponent is strong, which is why I get under pressure so much.

I also find that by thinking of concepts I'm getting even lazier in reading ahead.

But hey, it's training for a reason!