caranthir: @posetcay Ok, that's probably fine. ------- So, now the solution has three main branches. First is the easy one that also represents the strongest play for both sides and it's implemented. Second one is the straightforward punishment for opponent's blunder; implemented. (We break through, and if the opponent tried to prevent it, the side would collapse. This reason is not implemented into variations; it would be worth another small subproblem.) Third one is the alternative punishment - the other outside cut. More open and not implemented. I think it's ok to leave out the rest of it. I'll just accept the first move. ------ I am very strongly opposed to changing original starting positions of well established classic tsumegos. That this tsumego too has some "openness" is ok, let it be that way. Many of the classics are deep and by design don't necessarily have easily definable answers. I feel that many of them are more like essays than multiple choice tests, for the lack of a better metaphor.
posetcay: I think current solution tree is as good as it gets. The solution is a bit weird, but the only alternative would be to simply delete the problem. Just a minor change, I swapped K6 to L5 because it is the more natural move. This also eliminates the possibility of playing H5 at L5, leading to one fewer branch to worry about.
TwT: Dear admins, esp. Master Ivan, may you please have a look. As you may have noticed, caranthir gives his utmost to resolve issues being seen as "unresolvable".
Ivan Detkov: Am I right when I say the following? As it stands now, there is a solution based on escape but we are not sure. If we add a stone around O18, then there is no solution. It just started to feel like the task is broken.
caranthir: Ok, I edited the problem in the line of my comment below. ------- The problem can be enriched somewhat still if someone dares. Now in the alternative response, M16 instead of K15 would be another way to punish (globally almost as good). We'd get a wall on the side, the opponent crawling on the second line (opposite to Hrewsahghs' double-hane-and-connection-under proposal). Capturing that opponent group (and living with our inside stones) isn't very feasible but theoretically there's a ko fight if we wanted to play that way. Whatever the case, the wall on the side in sente is already enough compensation. ------ I'd be intimidated to edit in this latter (second order alternative) option, a stronger player should do it, if at all. I don't want to just put in Katago lines without thinking, and therefore the task would demand me considerable time. ------ @Ivan, I put also E18E19G19C19D17F19 in.
caranthir: Now when I look at it, I think it is also possible to make it a complete problem with the alternative response pretty neatly too without any modifications. Assume that E2E1G1C1K1. Now if the opponent wants to live, they'd have to play D3. Then we play J5; then for instance an continuation K4J4K6H5 and now we break trough into the center. These are all Katago blue moves starting all the way from K1 (C1 being the only bad move).
caranthir: @Ivan Detkov do you mean, E2E1G1C1D3 as a wrong answer? We could also have an alternative response E2E1G1C1 and thereafter K1 as a right player answer, with a note saying something like "opponent refutes oiotoshi but gets complications on the outside".
caranthir: I'd like to discard the current accepted solution (starting with G19 and proven to be faulty), and adopt the original solution outlined in the Sensei's article and supported by Katago, representing the strongest play for both sides. In other words, the following: E18E19G19F19C19, which establishes the oiotoshi. Now the problem is finished and in a real game the opponent should play elsewhere, reserving moves like K16 etc. as ko threats and forcing moves for later. Please state if anyone is against my idea.
Ivan Detkov: IMHO, this is fine. Just put this variant E2E1G1C1D3 in. It took some time for me to catch it.
caranthir: A continuation on my response to Hrewsahgs below. I argued that the problem is not broken but works as intended if taken the (most likely) whole board context into account. The strongest resistance from the opponent is to sacrifice and squeeze, making the outside solid. -------
That said, the starting move at G19 is wrong. (See also "wrong order of play" in the Sensei's article.) Using this order, not only does the opponent prevent oiotoshi, but we don't get to capture the J2 & K2 stones for the outside complications. The right starting move is at E2. Now the opponent will *not* make eyes for the inside group, but rather sacrifice and squeeze.
Hrewsahgs: [current position] + K19D17M16L15N17N18O18N16N19O17M19M15P18 I wonder if this would be an acceptable replacement to the main solution? The attacker can arguably do better than L15, but at least by that point it's whole-board related instead of purely local.
caranthir: The fact that opponent can indeed save the inside group if they really want has been previously noted: https://senseis.xmp.net/?GokyoShumyoSection5Problem12%2FSolution ------ This is doubtlessly NOT an oversight from the professional authors. The fact that forcing moves and complications on the outside exist if the inside group lives (given that we threaten oiotoshi in the right way, which makes us capture J18 & K18 if the oiotoshi is refuted - see my comments above), makes the option of sacrificing feasible. Indeed, if you put the position into Katago without anything else on the board, it will sacrifice and squeeze from the outside. I will stress: *this problem is not broken*.
Miko_boc: I wonder what happens if p2 is answered with r1 instead of capture at p1?
posetcay: P2 is a really nice refutation, I can not find a way for white to live after that. That means the problem is unsolvable as it stands now, so I changed the problem a little to fix it as best as I can (added O18 and M16), but I am open to alternative suggestions.
OK