I recently played a few games against Amybot-ddk.
I already played against it some time ago and didn’t like it very much. Amy sometimes plays weird moves and has serious issues with life and death. I managed to win games where I was behind by killing groups that I shouldn’t be able to kill. I only did that because of Amy’s mistakes.
But then I realised that I was missing a point: if Amy is weaker than me, why was I behind in those games? Maybe Amy is quite strong at fuseki and middle game… I mean stronger than me! Is that the case?
I decided to investigate further and I discovered something:
- I am a bit too greedy
- I always fear to miss opportunities
- I can’t properly evaluate if I am behind or ahead
The results are that I cut and fight when there’s no need to. Also Amy is quite good at punching me when I do that. (So we could add to the list “I’m not good at fighting”, but that’s matter for another study)
Summary of my games against Amy was: quite even fuseki, useless fighting in middle game, going behind, recovering just because of bot’s weaknesses.
That’s bad.
So I stated this goal: “I must always be ahead”.
That is “my winrate must always be above 50% or my score must be always positive since the start of the game”.
I found out I can do that by playing “basic” according with Dwyrin’s definition (which I already summarised in another thread).
“I’m not trying to kill” was my mantra.
I managed this way to reach my goal a couple times and I think I’ll keep doing this sort of training, since I feel it makes me stronger at “keep calm and play your game”, taking under control my “cut first, then read” issues.
All that was possible only thanks to katago analysis: that’s how I realised how I was wasting my points getting involved in useless fights. That’s also how I’m able to check eventually if I was consistent with my goal during the game.
So I must thank two bots: one for letting me crush it repeatedly and the other for letting me understand what was really happening in my games.