Re: current development

[Timothy Y. Chow]

On Sun, 8 Dec 2019, Joseph Heled wrote:
> Agreed, but from a practical point of view, not caring about 
> non-reachable positions and positions with a very low probability is 
> good enough for a playing-bot.

We probably shouldn't clutter the mailing list further with this debate, 
but *from a practical point of view*, what I most want from a bot is to be 
able to analyze decisions that come up when I play humans and that are not 
obvious.  These positions include "propositions"---artificially 
constructed positions that are fun to play.  Also, in my playing group, 
when we play for fun, we often intentionally steer the game into crazy 
backgames.  These might be "low probability" by some definitions, but they 
aren't low probability in real life.

Here's a true story about one of my friends.  After he purchased XG, the 
*very first position* he entered into XG and asked for its opinion about 
was a "wild" position that had arisen in one of our fun games.  He was 
disappointed that XG had no clue about the position (as was evident by its 
wildly swinging verdicts when one changed the evaluation level).

Tim