🚀 go-pugleaf

In <1992Sep26.014734.15691@cbnewsl.cb.att.com> ed@cbnewsl.cb.att.com (edward.m.hummel) writes:

>In article <1992Sep25.220728.26568@unixg.ubc.ca> ramsay@unixg.ubc.ca (Keith Ramsay) writes:
>>As an aside, I'd like to point out that the problem with repeated
>>positions is not an essential problem, so long as there are only
>>finitely many possible positions. ...
>>
>>For large enough n, W_n and B_n have to stabilize.  The remaining
>>positions are ones in which neither can force a win, i.e., either can
>>in principle guarantee a draw. 

>The point is that even though either can guarantee a draw, it cannot
>be proven that the value of the position is a draw.  At least, I haven't
>heard a proof, or seen one posted here.  So, if there are a finite
>number of such positions, and the values (although indeterminate) are
>constant, then chess is solvable, BUT we don't know the value of the
>initial position.  (We know it as a function of a finite number of
>undetermined constants.) 

I am apparently missing the point over these 'loops'.
Why do you feel that they must be assigned a value other than a draw?
It seems clear to me that these loops are drawn according to the accepted
rules of chess.  If neither side is willing or capable of making an
alternative move, the game is clearly drawn, and should be valued as such.
To determine whether an alternate move should be made is simply a matter
of using the mini-max approach as previously mentioned...

If you want to start assigning values other than 'drawn' 'won' and 'lost'
then you could, I suppose, also start saying that white lost, but 'not by
much'...  But this complicates the game...  The goal is (usually) to win,
not to stomp your opponent as hard into the ground as you possibly can.

---Xeno