Re: current development

[Joseph Heled]

I think that the idea that backgammon is "solved in any way or form" is ridiculous, and part of the Dunning Kruger effect. 

Even racing is not "solved" in any meaningful way when you have cube actions to consider. 

Right now I am writing a book that I hope will help to dispel this notion. It is not about BG but much simpler race games, ones that can actually be solved.

To do a "AlphaGammon" would require, at a minimum, one or more googlers who have worked on AlphaGo or AlphaChess and are willing to contribute time and resources. 

That is my personal opinion, that is, I would not go into such an endeavor without it.

-Joseph

 

On Thu, 5 Dec 2019 at 08:08, Timothy Y. Chow <address@hidden> wrote:

On Thu, 5 Dec 2019, Joseph Heled wrote:
> I had the same idea the day I heard they cracked go, but just saying
> something is a good idea is not helpful at all in my book.

Well, other people may have other books.

Also, it's my impression that many people *don't* think this is even a
worthwhile idea to pursue.  Backgammon is already "solved," is what they
will say.  It's true that "AlphaGammon" will surely not crush existing
bots in a series of (say) 11-point matches.  At most I would expect a
slight advantage.  But to me, that is the wrong way to look at the issue.
I would like to understand superbackgames for their own sake, even though
they arise rarely in practice.  Furthermore, if we know that bots don't
understand superbackgames, then the closer a position gets to being a
superbackgame, the less we can trust the bot verdict.

Some years ago, Paul Weaver created a "backgame quiz" with some
interesting backgame problems.  Most were not what I would call
superbackgames, though a couple were getting close.  The only rollouts I
ever saw anyone else do were with standard rollout settings.  I have been
running (eXtreme Gammon) rollouts with much stronger settings, and in at
least one case I've gotten some noticeably different results.  I think
this is interesting.  I also don't fully trust even the strong rollout
settings that I'm currently using, and would welcome a bot that I would be
more willing to trust.

This thread was asking about whether there is any current development of
GNU.  If someone poses that question, and nobody even mentions that
"AlphaGammon" would be of interest, then the impression can be created
that nobody cares about that.  I want to combat that impression.  So just
because I'm not prepared to invest in the supply side of the equation
doesn't mean that it's of no value to exhibit the existence of the demand
side of the equation.  In my book, at least.

Tim