Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: The Need for Fischer Random Chess !

Author: Tord Romstad

Date: 14:58:49 06/05/04

Go up one level in this thread


On June 05, 2004 at 13:06:21, Sune Fischer wrote:

>On June 05, 2004 at 11:18:49, Anthony Cozzie wrote:
>>>
>>>But even if it is not a lot of work, why do it if nobody cares?   Jorge is the
>>>only person who has expressed any kind of interest in my FRC engine.
>
>If 1 in 50 are interested in FRC, isn't that still 1 in 50?

I think the ratio is much less than 1 in 50.  Let's make a really silly
estimate:

There are about 300 public chess engines, and about 10 public FRC endgames.
If we assume that the interest in both games were equally big, I would expect
to receive about 30 times as many e-mails from the FRC crowd than from the
chess crowd (because the chess crowd has no good reason to be very
interested in my particular engine when there are so many stronger engines
around).  But the truth is that I receive about 100 times more e-mails about
chess than about FRC.  This indicates that perhaps 1 in 3000 is interested in
FRC.

>Perhaps 1 in 30 are interested in an SMP engine, but SMP is 1000 times harder to
>implement so if you want to do something for the customers, why not begin with
>FRC.

Another reason why I think this comparison is flawed is that the interest in SMP
is likely to grow when more customers get dual-CPU computers.  At the moment
there is no reason to believe that the interest in FRC will grow dramatically.

>>>>>The truth is that there is almost zero interest in FRC.  From a commercial
>>>>>point of view, adding FRC support to Shredder or Hiarcs would be a complete
>>>>>waste of time.
>>>>
>>>>Well so is adding SMP support, and unlike FRC that's not a small hack at all.
>>>>:)
>>>
>>>SMP support is much more useful.  It makes it easier to win tournaments
>>>like the WCCC, which is nice for advertisement purposes.  I think more
>>>customers are interested in buying a program which is marketed as
>>>the "computer chess world champion" than a program which is marketed
>>>as being "stronger than Frenzee and Gothmog at FRC".
>>>:-)
>
>Perhaps, but what about the title of "computer FRC world champion"? :)

Right now, it wouldn't be a prestigious title at all.  If we organized an FRC
world
championship tomorrow, the chances that you or I will be the winner are quite
good.  I doubt that people will rush to buy our engines because of that.  :-)

Given a few weeks or months of preparation, the competition would almost
certainly be harder.  If some of the top commercials had any interest in
participating, they would of course beat us easily.  But I really doubt if
they would bother to come.

>Honestly I don't think winning WCCC matters, what matters is getting a match
>against Kasparov. WCCC doesn't seem to be opening that door.

Winning the WCCC is often what opens the door to getting a match against
Kasparov.  I cannot remember any engine which has never won a world
championship title which has played a match against a top 10 GM.

>I think WCCC is mostly a gettogether for programmers, the "title" can't be that
>important or we would see Tiger, The King, Rebel, Ruffian and the others
>fighting for it too.

With the exception of The King, I don't think any of these engines sell as much
as Fritz, Junior or Shredder.

>At first sight it may seem like computers should have a tougher time without
>books than humans, ie. humans will still understand the basic strategies on how
>to develop better.
>The question is how much of an ordeal it is for humans to face completely new
>patters already from the opening. Programs on the other hand I wouldn't expect
>to have trouble in that area, they don't do patterns very well anyway.

My guess is that it wouldn't matter much.  The relative strength of GMs and
computers would remain more or less the same in FRC.  But because we agree
that this isn't the point of FRC, there isn't much point in continuing that part
of the discussion.

>>I haven't spent very much time studying openings either.  I think you can get up
>>to maybe 2000 rated without a lot of openings, but at that point its an absolute
>>must.  And even if you don't spend the time studying the theory itself, you
>>still need to know the standard plans for each side in each opening.
>
>I don't really understand Tord's point of view here, I am sure you can play club
>level chess without needing to read a lot, however anyone that do read the books
>also seem to get advantages from that.
>
>I don't know how many times I've walked into opening traps against 1400 rated
>players :)
>As soon as they are out of the opening they start to throw away pieces to 2-3
>move combinations though.

Then you should just take them out of book immediately.  The last serious game
I played, a team event about 5 years after I stopped playing chess voluntarily,
was
against a player rated around 1500-1600.  I really didn't want a theoretical
struggle, because I hadn't pushed a pawn or opened a chess book for years.
After some thought, I played 1. Nc3, and won easily.  Of course 1. Nc3 doesn't
offer quite the same chances of fighting for an opening advantage as 1. e4 or
1. d4, but it is safe to bet that a 1500 player isn't able to refute it.

>>But I agree, FRC is one of the most boring chess variants out there.
>
>If FRC is broing then Chess is boring, because it was never intended to be a
>variant as such. FRC is what you end up with, when you take "chess minus opening
>theory". The whole idea is to change as little as possible so it reamains the
>_same_ game, only one "problem" removed.

In my eyes, the "problem" you remove is the single big advantage classical chess
has compared to the countless other chess variants.  Chess has a rich culture,
tradition and history.  It is very fascinating to study chess games during the
last
150 years and see how the opening ideas have developed, and how the top players
constantly manage to churn out new ideas in openings which has been analysed
for decades.  Watching positions familiar from classic old games by Tal or
Botvinnik gives a feeling of history which you just don't get if all the games
are
played from randomly chosen positions.  I find it more intersting to let the
players choose the battleground than to roll a dice.

Another asset of playing the games from an old, well-known starting position
is that some extremely complicated and interesting positions are analyzed to
an enormous depth, which would otherwise never have been achieved.  Have
you read "Fire on Board", Alexei Shirov's game collection?  If not, I would
strongly recommend it.  The games and analysis is fabulous, and it is clearly
the best and most entertaining chess book I have read in recent years.  One
of my favorite parts of the book is the chapter about the Botvinnik variation
in the Semi-Slav, where Shirov has annotated all his games in this wild and
spectacular opening.  If we had all played FRC instead of classical chess,
something similar this chapter wouldn't have been possible.  No doubt
games would be played with positions just as beautiful as the Botvinnik
variation, but the positions would be quickly forgotten and never analyzed
deeply or understood very well.

To a certain extent, the reason for our disagreement might be a difference
in perspective.  I no longer play chess, and I probably never will.  I follow
the chess news, and I love to study game collections and endgame books,
and to analyze endgames and complex tactical endgames.

>Most chess players like the game and are not interested in strange variants, yet
>they might still be bored with all the theory, those are the ones who should be
>looking towards FRC for a solution.

Another solution is to play 1. Nc3, or some other sound but unorthodox
move a few moves later.  As I have stated earlier, most of what is known as
"opening theory" isn't really theory at all, but just a mix of tradition and
vague educated guesses.  Nobody forces you to follow the theory.  You
might lose games because you follow the theory without remembering
it or understand it well enough, but you won't often lose games because
you don't play the thoroughly analyzed lines at all.

>I tried a few games of Gothic, in the middle of the opening I was suddenly mated
>in 1. I never saw it comming, I just couldn't picture the attacked squares of
>the new pieces. I don't know how long it takes to adjust to that, but if feels
>like a whole new game to me, quite far from chess.

I think it's really cool. :-)

All the well-known tactical motifs and endgame rules of classical chess are
still there, but in addition you discover lots of entirely new tactics, mating
patterns and endgames which you had never dreamed of before.  It takes
a lot of time to get used to it, but to me the game gets richer and more
interesting.  I don't mind feeling like a beginner again.

Hexagonal chess is even more interesting, because the more complicated
geometry of the board adds new possibilities even for the familiar old
chess pieces.  Rooks, bishops and knights can triangulate, and kings can
sometimes catch up with passed pawns from behind by moving in a
diagonal zigzag pattern.  The endgame in particular becomes much more
interesting than in any rectangular form of chess.

Tord



This page took 0 seconds to execute

Last modified: Thu, 15 Apr 21 08:11:13 -0700

Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.