The Greatest Go Tournament Run Possible: The #1 Player 

New hoodie, new haircut. Looking good

I think that it changed depending on how many playouts, it would realize that shin jinseo was losing about 10 moves earlier on 1 million playouts compared to 10k iirc

@@telegraphgo thanks, very interesting! Impressive that you remember all these months later 😅

coming from a chess background, im so used to engines being the king of endgames, what with tablebases etc

Absolutely. I actually am feeling different about these games over time as I analyze more of them as well. I feel like I am improving a lot at understanding pro players dreams and desires within the game. Different players have different dreams - I feel like I understand Nakamura Sumire better than other top players, and Park Junghwan is my favorite player (full stop, but also) to analyze when I want to refine my sense of what is worth going for. With some experience of doing it, I actually prefer not to analyze Shin Jinseo. His judgement of what is good and bad is better than his peers but using that tends to lead his games down unexpected and irreplicable routes. His dreams are significantly more obscure than his peers. His style is significantly more flexible. The reason why I prefer not to study this is that I feel like I'm more in need of learning large scale victory techniques than a relentless onslaught of very hard-won little victories.

Thank you for your reply. It is generous of you to be open and sharing, and your ideas are thought-provoking. I am a 14kyu player and to improve I have focussed on techniques described in books. Now, however, I want to try reading and analysing games. Here's wishing you success in the New Year! @@telegraphgo

Not really, the winrate was about 50% and points difference was 0.0, not 0.5 in either player's favor. I could research endgames until the AI agreed on a decisive result with me.

Of course that's a consideration, but it's always wrong unless the territory is more important than the center moves. It's not like you're ever really afraid 2 free center moves (first one and the one in exchange for black taking the corner) will ultimately result in negative value for the center boundaries in practice.

@@telegraphgo you underestimate my ability to worry

black k6 is just 2 points (at j6 and h5) compared to white k6, best not to waste ko threats when white takes an equal size endgame.