Thank you very much everyone for the sharing & discussion!
(I should delete this Kogo’s sgf I downloaded gods know how many years ago )
This is interesting, I’ve always played B in this situation. TIL
This is also interesting: I always thought 5 and H5 are “pushing the cart from behind” which is bad for black? 4th-line territory feels good enough for me to accept the incoming relentless attack on the left side.
You’re right! I was missing Black’s 7 should push at H5 instead
Pfff it’s just Skynet what does it know
I used to be fan of TheCaptain & Bloodarena on KGS
===
Offtopic the “deviant” term: I guess whether it’s negative or not depends on cultural experience. DeviantArt for example is a great inde art community. To me it’s just “not following the mainstream”.
Pushing the cart from behind is bad if you don’t gain enough or if the stones to be attacked are light. In this case, the stones on the left side are stones that need to be saved, so giving black unnecessary sente is a big risk.
Of course if black doesn’t attack well the risk may pay off, but it’s generally better to just take the stone when you can.
Checking Waltheri’s database, the outside atari is played about 8 times more often than the inside atari. I found only one game where the opponent immediately pulled out the cutting stone [¹], so it seems that pros tend to play the inside atari only when it seems unattractive to pull out the cutting stone immediately.
Although I suppose it is already different from the basic joseki because of the presence of an extra white stone at K17 that might cause the outside atari (G17) to be slightly overconcentrated. I suppose that avoiding overconcentration is an additional factor for pros to decide which atari they pick.
My KataGo agrees with Lee ChangHo’s inside atari at F16 (although the score difference with the outside atari G17 seems less than ~0.1 point). KataGo doesn’t quite agree with black pulling out immediately, thinking it loses ~0.3 points.
But both players make clearly bigger mistakes of ~0.5 points in the game continuation, so I suppose that in this position the difference between the inside and the outside atari is too subtle even for players of this calibre.
Some KataGo variations from the position on move 22. All are very close games with black having a lead of 0.0 - 0.3 points, as is the actual game continuation up to move 38:
We use the word deviation for an itinary replacing the main one (like when there is an accident). This one is secondary and usually less convenient, more convoluted as the original one.
In our context of josekis, maybe “variants” is a word more commonly used?
But first let me show you some findings on the moves leading up to the position:
On move 5 KataGo slightly prefers a 3-3- invasion instead of the F3 approach.
Also, black’s D3 attachment on move 7 seems slightly premature. KataGo prefers tenuki.
And on move 8 KataGo prefers white F4 over pushing through with E2.
So we are here now with black leading by about 0.3 points:
If white instead extends (which loses a bit more than a point), black should not just push and give white territory, but use better haengma to create a sort of “body” for his floating group by pressuring white on both sides. For example:
Hmm not peeping D7 come out very weird to me: in this variant white’s left group looks much better. And although D7 aji remains to prevent white cut at F6, B’s 8 also serve the same purpose.
I suppose black’s aim is to settle his center group more or less in sente to keep the initiative.
When neither player makes a significant mistake, a balanced result like this should be expected. Black shouldn’t expect to get it all, and neither should white.
I feel like if you don’t like this variation for black, then maybe it’s being played too early. Black can also play something similar later when the adjacent corners and sides have more stones.
Or if white defends in the centre black gets two moves in a row somewhere, that is White played F4, black somewhere else, white plays an extra move in the centre to settle and black again plays elsewhere.
The two moves should probably be enough compensation in the opening I imagine.
If you peep and continue like in your previous diagram with pincer on left side, then (apart from needing to push on 4th line a bit more to avoid net, and white can hane up to 5th so you give white lots more on lower side) you are fundamentally making 2 groups on the outside rather than 1, which is a bit too much to handle. It’s not game over too much and of course white could mess up in the fighting and let you get a good result, but it’s more sensible to stick with only 1 weak group on the outside and make some eyeshape for it, and later you can play checking extensions/pincers on the side, moves like c11 and n3, once your centre group is reasonably strong. And topologically speaking there is still value in the 2 white side groups being separated even through they have decent eyespace on the sides for now, because later fighting on those sides does carry some opportunities to harass them (e.g. imagine some trade where white kills q4 but black gets a giant wall along N in exchange, then h2 and l2 are nice follow ups and that group might die).
I had not heard of Bloodarena - though the username suggests a style of play -
but there is a SL article on TheCaptain : The Captain at Sensei's Library
TheCaptain was 5d KGS before retirement. BloodArena was 7d. Both are quite famous for aggressive playstyle usually starting a fight very early by choosing fighting variation at first corner approach.
I prefer BloodArena personally, they have a habit of feeding a big dragon in a temporary dead state, leading to multiple cuts in surrounding groups that escalate into whole-board fights
)
-