Re: computer-go: perfect play

[Don Dailey <drd@xxxxxxxxxxxxxxxxxxxxx>]

   From: "Vlad Dumitrescu" <vladdu@xxxxxxxxxxxxxxxxx>
   Date: Tue, 26 Sep 2000 08:46:58 +0200
   Reply-To: computer-go@xxxxxxxxxxxxxxxxx

   *** I don't want to get into a dispute, but I really wonder about one of
   Vincent's latest affirmations:
       >In chess you can go beyond that. It has happened that Kasparov had
   prepared games up to 30 moves deep.
       >He's not doing that on his own. He has a whole team to do it with him.

   Does that say anything about Kasparov's strength? With a good enough team,
   one would need only to have good memory in order to play well... (of course,
   the opponent should not know that and not play strange moves to create
   confusion)

   ---------


Actually, Vincent is right (for a change  :-) Teams of grandmasters do
work with the very best players,  especially the world champion.  They
try to find what  is call "TN's" or  "theoretical novelties", holes or
missing analysis in current theory.  Also, they can go very deep, even
into the endgame as Vincent says.

In a way, I find it kind of disgusting myself.  The best player in the
world, also gets the further advantage of  private analysis!  However,
it's hard to  argue this on  any kind of moral  grounds so I won't!  I
think we all have  the right to  procure knowledge, even hiring people
if we choose to obtain  this knowledge.  The knowledge sometimes get's
dispensed  to the  public  by means  of   games where  the novelty  is
introduced to the public.

What often happens is that the opening are prepared specifically for a
given  opponent, knowing  in advance  what his  favorite openings are.
Sometimes,  you must  wait  (I've done this myself   too to beat  much
stronger players) for your chance because you may play that opponent a
few times before you have the right color and get the right opening.

I have argued that this is just more evidence we are not that close to
playing  at ChessGod level.   If we  were, it  wouldn't be possible to
find theoretical novelties.

However  there are basically  2 kinds of  these.  One is an attempt to
refute  outright  a  position.  Imagine  that,  it's possible  to find
commonly  played openings that are LOST  positions and can be refuted!
It  turns  out  that computers have   also  discovered many of  these,
despite  the fact   that computers   are still weaker   than  the best
players.   It's by virtue of  their  very thorough tactical  analysis,
they can catch moves often missed by humans.

The other kind of TN is to simply  surprise your opponent and make him
waste   time   thinking.  Since you  have   the   advantage of offline
analysis,  you  can take the  time  to understand all  the most likely
replies and  their intricacies.  You may  not  have a true refutation,
but you suddenly have a big practical advantage.

Don