[computer-go] New version of Crazy Stone

Sat Jun 3 17:58:23 PDT 2006

I don't have a link, but I read a description many years ago.  They looked
at the variance in score between players of the same strength, and found
that weaker players have higher variance.  They plotted the variance against
the strength, and extrapolated to zero variance, which would be perfect
play, at about 4 stones above top pro.

David

-----Original Message-----
From: computer-go-bounces at computer-go.org
[mailto:computer-go-bounces at computer-go.org] On Behalf Of Jim O'Flaherty,
Jr.
Sent: Saturday, June 03, 2006 3:06 PM
To: computer-go
Subject: Re: [computer-go] New version of Crazy Stone

David,

I'm curious to understand how such a study might work.  Do you have any
links or references to the studies?  I am genuinely curious.

Since perfection is computationally outside any sort of human or device we
have discovered/invented thus far, what kind of basis establishes a
"perfection horizon" against which the pros could be measured?

I am coming from a frame that there is something here I can learn, not that
the studies are erred or fallacious.

Jim

----- Original Message ----
From: David Fotland <fotland at smart-games.com>
To: drd at mit.edu; computer-go <computer-go at computer-go.org>
Sent: Saturday, June 3, 2006 4:17:55 PM
Subject: RE: [computer-go] New version of Crazy Stone

Statistical studies suggest that the top pro players are about 4 handicap
stones from perfection.  Since there are only about 2 stones between the top
amateurs and the top pros, this is quite a large gap.

David

> -----Original Message-----
> From: computer-go-bounces at computer-go.org 
> [mailto:computer-go-bounces at computer-go.org] On Behalf Of Don Dailey
> Sent: Saturday, June 03, 2006 9:52 AM
> To: computer-go
> Subject: Re: [computer-go] New version of Crazy Stone
> 
> 
> Although I'm not very knowledgeable about Go as a player,
> I would agree with the statement that human ability is
> vastly overestimated.    This has been the case in so
> many fields that I am sure it's the case here too.
> 
> Sometimes I think that experts in a particular fields are
> the worst ones to ask.   It's often the case that there is
> a kind of arrogance involved, where they place themselves
> and their achievements too high on an absolute scale.
> 
> I have asked and pondered this question with chess skill
> too.   Even standard checkers on an 8x8 board is not
> really mastered by humans and it is far simpler than
> chess, and chess is far simpler than GO.   
> 
> I did a simplified version of checkers once on a 6x6
> board where it was possible to achieve incredible search 
> depths.
> 
> In my study I learned that it was possible to get 
> continued playing improvements up to the point I could
> reasonable test.   That was a few years ago and perhaps
> 6x6 checkers is now doable with perfection by computer -
> but it dramatically illustrated to me that if you could
> still make improvements on such a simplified game,  then 
> the top computer chess programs must still be incredibly 
> far away from perfect play.   
> 
> Now you can get chess programs that are as good as the
> best players for your home PC.   So my guess is that human
> chess players are as far from perfection as masters are
> from beginning players.    If that is true of chess, I'm
> sure it is at least as true, or more - for go.
> 
> It may be that at levels well beyond human play it does not 
> make as much sense to measure playing ability in terms
> of "stones."    Perhaps there is huge range of skill within
> that final "stone" of playing ability - where a 1 stone 
> handicap makes it impossible to win but without the handicap 
> you can dominate a near perfect player who is still not quite 
> as nearly perfect as you are!
> 
> 
> - Don
> 
> 
> 
>  
> On Sat, 2006-06-03 at 05:28 -0700, steve uurtamo wrote:
> > > How many stones would a top pro need against a
> > > perfect player? (3 or 4
> > > has been bandied about as the answer).
> > 
> > that's an excellent question, although i think
> > that it dramatically overrates human playing
> > strength.  to think about it a different way,
> > how many stones would a top pro player today
> > be able to give to a top pro player from 50
> > or more years ago?  even in just the fuseki,
> > so many newer josekis (including chinese and
> > korean styles) have been created that a pro
> > from another time wouldn't have the knowledge
> > to deal with, i think that at least a stone or
> > two would be fair.  to imagine that we're now
> > only 3 or 4 stones away from perfect play is
> > a great stretch of the imagination about our
> > progress in the last 50 years, i think.
> > 
> > s.
> > 
> > __________________________________________________
> > Do You Yahoo!?
> > Tired of spam?  Yahoo! Mail has the best spam protection around
> > http://mail.yahoo.com <http://mail.yahoo.com/>  
> > _______________________________________________
> > computer-go mailing list
> > computer-go at computer-go.org
> > http://www.computer-go.org/mailman/listinfo/computer-go/
> 
> _______________________________________________
> computer-go mailing list
> computer-go at computer-go.org 
> http://www.computer-go.org/mailman/listinfo/computer-go/
> 

_______________________________________________
computer-go mailing list
computer-go at computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://computer-go.org/pipermail/computer-go/attachments/20060603/8552299d/attachment.htm