When did my opponent make a decisive mistake? I am black here. I killed large bottom-center group, but white is lost even with that group alive.
MUCRNTRZJL.sgf (4.3 KB)
It’s much nicer to share SGFs with us by uploading them.
You can do that by opening the left hand hamburger menu (next to the OGS logo in the top left) and selecting “Library” then uploading your SGF
The result is like this .
Straight away you will get the “top 3 game changing moves” analysis, and if you are a supporter straight away you will also get the AI analysis.
I triggered an AI analysis for this game while I was at it. Since it was handicap, the AI might not have a lot to tell us, but now others can see, review this and leave comments more easily.
My own observation is that handicap is hard, but you can’t afford to lose all the corners and not create a single weak group for black.
In this game, white’s approach seemed to be “you can have the corners and I will have the middle” which I think is probably a losing proposition. It’s too easy to invade the middle when you have 4 strong corners to do it with and no weak groups.
HTH
EuG
1 Like
White used some slow moves in the opening.
Then a tipping point happened when black broke into the middle on the right (move 42-50). White just throws his stones to the wolves. With moves of this caliber, white can not expect to catch up even to a weaker player.
The AI evaluation is especially unhelpful. It seems to believe that white gained by leaving his stones in atari, and black lost by capturing them. This goes beyond what I feel is the “usual” apathy that Leela displays in decisive positions.
That being said, we should generally not expect to meaningfully find “mistakes” by white in a handicap game, with or without AI help.
Black is objectively almost guaranteed to win from the start. You enjoy the ultimate freedom to play whatever you like, since every move is almost equally good (or hopeless).
Ditto.
I thought the same thing, and said so in the AI thread.
However, it was pointed out to me that in a handicap game realistically we should expect 50% of the time for either side to win because we know that these games are played by unevenly matched players, which means that 50% of the time we should expect to find mistakes in the AI evaluation, and once the stronger player has gained a foothold, from that point the analysis should be interesting.