Re: computer-go: perfect play

[William Harold Newman <william.newman@xxxxxxxxxxx>]

On Wed, Sep 27, 2000 at 01:20:54PM +0100, TMCooper wrote:

> On a related matter that has come up.  I am no worse than 192 stones worse
> than GoDevil or GoGod, if I am allowed to place my stones myself.  My
> strategy is to place my handicap stones in a grid (waffle) pattern to fill
> rows 2,5,8,11,14 and 17, and the columns of the same numbers.  Then I will
> pass unless my big group is put into atari, at which point I will capture
> some enemy stones and continue as before.  At the end of the game I will
> have a very comfortable win.
> 
> I think that this is a proof.  I am sure that others can find much higher
> lower bounds for their strength.  Does anyone know of any better ones? Or
> can they come up with any themselves?

I was at the Go Congress two years ago when Martin Mueller beat
ManyFaces at a high handicap (27 stones? 29 stones? I forget..) and I
proposed something like this at that time, to illustrate how
remarkable it is to lose at such a high handicap.

     A B C D E F G H J K L M N O P Q R S T
  19 . . . . . . . . . # . . . . . . . . . 19
  18 . . . . . . . . . # . . . . . . . . . 18
  17 . . . . . . . . . # . . . . . . . . . 17
  16 . . . . . . . . . # . . . . . . . . . 16
  15 . . . . . . . . . # . . . . . . . . . 15
  14 . . . . . . . . . # . . . . . . . . . 14
  13 . . . . . . . . . # . . . . . . . . . 13
  12 . . . . . . . . . # . . . . . . . . . 12
  11 . . . . . . . . # . # . . . . . . . . 11
  10 . . . . . . . . # # # . . . . . . . . 10
   9 . . . . . . . . # . # . . . . . . . . 9
   8 . . . . . . . . . # . . . . . . . . . 8
   7 . . . . . . . . . # . . . . . . . . . 7
   6 . . . . . . . . . # . . . . . . . . . 6
   5 . . . . . . . . . # . . . . . . . . . 5
   4 . . . . . . . . . # . . . . . . . . . 4
   3 . . . . . . . . . # . . . . . . . . . 3
   2 . . . . . . . . . # . . . . . . . . . 2
   1 . . . . . . . . . # . . . . . . . . . 1
     A B C D E F G H J K L M N O P Q R S T

(This has 23 stones, and I think what I proposed at the time had 25, with 
extra stones at J8 and L12.)

Now if Black plays mirror go, with every move on one side reflected
across the line of stones (so e.g. H2 is answered with M2) then Black
should win by at least two points.

I talked about this with Martin and we couldn't convince ourselves
that there's no way to defeat it with (super)ko tricks, but we
couldn't find such a way ourselves. 

It might even work to play the same mirror-go strategy with this
starting position (22 stones)

     A B C D E F G H J K L M N O P Q R S T
  19 . . . . . . . . . # . . . . . . . . . 19
  18 . . . . . . . . . # . . . . . . . . . 18
  17 . . . . . . . . . # . . . . . . . . . 17
  16 . . . . . . . . . # . . . . . . . . . 16
  15 . . . . . . . . . # . . . . . . . . . 15
  14 . . . . . . . . . # . . . . . . . . . 14
  13 . . . . . . . . . # . . . . . . . . . 13
  12 . . . . . . . . . # . . . . . . . . . 12
  11 . . . . . . . . # # # . . . . . . . . 11
  10 . . . . . . . . # . # . . . . . . . . 10
   9 . . . . . . . . . # . . . . . . . . . 9
   8 . . . . . . . . . # . . . . . . . . . 8
   7 . . . . . . . . . # . . . . . . . . . 7
   6 . . . . . . . . . # . . . . . . . . . 6
   5 . . . . . . . . . # . . . . . . . . . 5
   4 . . . . . . . . . # . . . . . . . . . 4
   3 . . . . . . . . . # . . . . . . . . . 3
   2 . . . . . . . . . # . . . . . . . . . 2
   1 . . . . . . . . . # . . . . . . . . . 1
     A B C D E F G H J K L M N O P Q R S T

or even this one (20 stones):

     A B C D E F G H J K L M N O P Q R S T
  19 . . . . . . . . . # . . . . . . . . . 19
  18 . . . . . . . . . # . . . . . . . . . 18
  17 . . . . . . . . . # . . . . . . . . . 17
  16 . . . . . . . . . # . . . . . . . . . 16
  15 . . . . . . . . . # . . . . . . . . . 15
  14 . . . . . . . . . # . . . . . . . . . 14
  13 . . . . . . . . . # . . . . . . . . . 13
  12 . . . . . . . . . # . . . . . . . . . 12
  11 . . . . . . . . . # . . . . . . . . . 11
  10 . . . . . . . . # . # . . . . . . . . 10
   9 . . . . . . . . . # . . . . . . . . . 9
   8 . . . . . . . . . # . . . . . . . . . 8
   7 . . . . . . . . . # . . . . . . . . . 7
   6 . . . . . . . . . # . . . . . . . . . 6
   5 . . . . . . . . . # . . . . . . . . . 5
   4 . . . . . . . . . # . . . . . . . . . 4
   3 . . . . . . . . . # . . . . . . . . . 3
   2 . . . . . . . . . # . . . . . . . . . 2
   1 . . . . . . . . . # . . . . . . . . . 1
     A B C D E F G H J K L M N O P Q R S T

Does anyone have a way to defeat this mirror-go strategy in any of
these starting positions?

-- 
William Harold Newman <william.newman@xxxxxxxxxxxxxxxxx>
software consultant
PGP key fingerprint 85 CE 1C BA 79 8D 51 8C  B9 25 FB EE E0 C3 E5 7C