Sadaharu: Added more options for the last few moves. Also I realised that white's answer on the top right could be better, but I'll just leave it as it is for now.
Lyde: I think Black can play E1 instead of J12, Black can play C10 or E2 instead of E7 and Black can play E7 right away instead of D7. Instead of connecting at D9, also A8 and A10 work and even make fewer Ko threats for if Black wants to try N5...
caranthir: 1) E1 should be answered with J12. Opponent gains. ------ 2) C10, E2 and E7 are equivalent. ------- 3) E7 and D7 are equivalent. ------ 4) A whole lot of moves are equivalent to connecing at D9, including yours. ------- 5) I could make these changes (except (1)), but when I look at this with Katago, there are literally dozens of more valid permutations that could also be added. Starting from the very first move. -------- Practically speaking I find making a complete (or even near-complete) solution tree for these endgame problems a near impossibility. And that tree would be a total nightmare. I'm sorry. I'll think about it. -------------------- I think that for complicated endgame problems, the only viable format would be to play against a bot (with enough playouts and/or some pre-analysis of the problem already put in). I's very different from tesuji or life and death problems, where manually curated sequences rule (because often the point of resistance sequences therein is to pose difficult-to-see challenges for the solver and to demonstrate the funcionalities of the shapes in question, and not so much to optimise points (most ironically a real player in a real game should very often just to simply play away and not even waste ko threats)).
1) E1 should be answered with J12. Opponent gains. ------ 2) C10, E2 and E7 are equivalent. ------- 3) E7 and D7 are equivalent. ------ 4) A whole lot of moves are equivalent to connecing at D9, including yours. ------- 5) I could make these changes (except (1)), but when I look at this with Katago, there are literally dozens of more valid permutations that could also be added. Starting from the very first move. -------- Practically speaking I find making a complete (or even near-complete) solution tree for these endgame problems a near impossibility. And that tree would be a total nightmare. I'm sorry. I'll think about it. -------------------- I think that for complicated endgame problems, the only viable format would be to play against a bot (with enough playouts and/or some pre-analysis of the problem already put in). I's very different from tesuji or life and death problems, where manually curated sequences rule (because often the point of resistance sequences therein is to pose difficult-to-see challenges for the solver and to demonstrate the funcionalities of the shapes in question, and not so much to optimise points (most ironically a real player in a real game should very often just to simply play away and not even waste ko threats)).