I don’t know of any current opening book for the 19x19 board, so I made one. It explores hundreds of positions with computer analysis by KataGo! Because of this, it’s optimized for KataGo’s playing style, but it can be used in human analysis and might be of interest to many of you. With this book, you may discover what KataGo thinks is ideal in the most common opening patterns. It won’t disappoint you, unless you want something deeper than 10 moves. Unfortunately, I hit the size limit for puzzles and can’t add anything more. But I can still adjust values, add notations and so on if needed. Enjoy.
P.S.: If the position ends before 10 moves, try them in a different order. There’s many transpositions to save on tree size.
I would like to include in in the OGS community resources, but I’m not sure what the category is (or what I’m supposed to do, but that’s a “me” problem).
If I may weigh in on this pedantic question… I would suggest that it belongs under “Demo Board Tutorials”. Even though it uses the Go Puzzles platform, it is not really so much like a problem collection, but more like an interactive demo board, albeit with arbitrary navigation restricted for the sake of presentation. It could also be classified as a “Go Book” I think.
How did you add the score labels to the board in the puzzle format?
Are all the points with score labels supposed to have variations like move 1 tengen move 1 3-3? Or is it again like you were saying, that there would be more variations but
(Can’t play Nirensei as Black ) It doesn’t feel like the majority of points aren’t clickable that have a score label.
Some of the moves say incorrect, like playing a 5-3 point after the star point as black. Is that intentional? Maybe ending the sequence with a correct rather than incorrect would be more intuitive I presume it’s just the default to display incorrect when you reach the end of a tree without hitting a correct.
No. As I mention in the intro text, I only explore nodes that kept a score between 0.1 and 0.3 points, since 0.2 appears to be parity. Exponential growth of the tree size, as well as time constraints, prevented greater exploration. Still I can explore any variation if there’s enough interest. It just comes at the expense of some other variation, sadly. That’s also why the cap is 10 moves.
Check the transposition at Q16 D4. I explored nirensei there. But if you meant sanrensei, KataGo doesn’t score it favorably enough for my inclusion criteria (although I can make exceptions if the interest is there).
I could’ve sworn that wasn’t working for me the other day!
Maybe I was trying it with the number tool or something silly. I wanted to use like two letter A’s I think and I couldn’t get the label option to appear. Maybe I just pressed some other combination of button+click either.
) It doesn’t feel like the majority of points aren’t clickable that have a score label.