Re: Bryce is \"In the Well\"
[ QUOTE ]
[ QUOTE ]
Hello Bryce,
One of the top Go players in the world once remarked that if there were a "perfect" (i.e., game theoretically optimal) Go player, he would estimate that he'd need a 4-stone handicap to play against him.
[/ QUOTE ]
I was actually having a similar discussion with a friend of mine tonight.
Bryce correct me if I am wrong here but doesn't a GT optimal line yield an EV of 0? i.e it keeps you from being exploited but also keeps you from exploiting another player?
So if the above statement is correct then wouldn't +EV play always be exploitative?
[/ QUOTE ]
Regarding the last statement, that's correct. "Optimal play" is sometimes used as two different terms. Some people use it to describe game theory perfect play which is EV neutral (I usually refer to this as game theory optimal, or GTO play, which can be good for shutting opponents down in areas where they are strong). Usually when I use the term optimal play I mean applying the best possible line to the most accurate assumptions possible (always a tricky point to define this, since it's a non-deductive process, so it's never 100%) yielding the highest average rate of return.
Now, to answer the first question, I'm not entirely sure. I'm familiar with what balanced play looks like in a lot of areas and can shut a lot of spots down, but my understanding of many of the larger areas is far from comprehensive. The reason it's hard to give a solid answer on this is that, since all this isn't quantified yet, there's always the potential for some stuff I don't know I don't know (as opposed to stuff I know I don't know) that could throw things way out of whack.
To give you an idea, though, I'm still learning new stuff every day.
|