Comments/Ratings for a Single Item

~~Earlier~~ ⇧Reverse Order⇩ ~~Later~~

Rich Hutnik wrote on Wed, Apr 2, 2008 04:49 PM UTC:

I believe the framework of chess can be addressed now so that we never turn
chess into a solved game.  I personally believe there is part of the answer
in a game like Seirawan Chess, or a pocket version with reserves, but I
don't think they alone have the answer.  It also doesn't address the
framework issue either that gets chess stuck, and all the classic abstract
strategy game (stuck here means set on a path to being 'solved', without
a way to adjust before it does).  

My take on Godel's Incompleteness Theorm is that you don't solve the
systemic issues with a certain set of rules by creating more rules of the
same type.  If it is show, for example with chess, that a set
configuration of chess pieces on the board eventually produces something
that is solved, then changing the configuration of the pieces on the board
alone doesn't resolve it either (one time, fixed).  You can change the
their starting position (aka Chess960/Pick your Army/MetaChess or the V
and X versions of IAGO Chess), the layout of the board at start (and also
changing it during play, aka Beyond Chess), or when the pieces enter the
game (IAGO/Seirawan/Pocket Knight/Pocket Mutant), and help to push things
out further.  If you build into the framework by which you can do all of
the above, you buy more time.  What regular chess has now is not a way to
make chess get 'unstuck', allowing it to adjust over time.  I suggest
all of the above be considered and integrated, and the players settled on
what works best.   Eventually even this mix of everything leads to a
'stuck' position as the playing community may figure out what is
optimal.  By then, some other people will need to come up with another
layer of rules to insure things are unstuck.  

I can't say this for certain, but I do know unsticking chess by doing all
of the above should likely buy chess another 1000 years, using all of the
above methods described.  The key to having it get unstuck is to have it
done in a way that it is evolutionary, so the playing community can
migrate over time and get used to the changes.  Also added to the mix are
'mutators' which are meta-changes to how the game works that get added
during play.  PlunderChess, for example, is built on a mutator that is
active from the start, pieces fusing together.  Even these added can have
an impact, and force people to think more creatively, relying on
principles.  These changes act as weather, and another key element to
getting chess unstuck (and other abstract strategy games for that matter).
 All these elements help to battle to keep a game from getting stuck,
without the use of random element, or hidden information, which is the
standard method used to unstick a game.  Like, the case of backgammon,
luck prevents it from getting stuck for a long time.  Stratego uses hidden
information, and the bluff element causes players to play other players. 
In this you need to know your opponent more than the environment.  Because
of this, a game like poker can be played even 1000 years from now, because
you play the players, and luck also offsets (hidden information+luck). 
Magic: The Gathering, and also Cosmic Encounter also relate to this, which
has in its makeup things that continue to change the rules.  I believe such
mutators can be applied to a game like chess, but not in such a chaotic
manner.  In other words, you can have a game that is a pure abstract
strategy game, but where the rules do change during the course of a game,
if the players control when the rules come into effect and the potential
rules are fully known by all players in the game. 

Please feel free to comment here.

Rich Hutnik wrote on Wed, Apr 2, 2008 09:29 PM UTC:

In another thread, Singh wrote:

Not a member, so responding to 'Unsticking Chess' here.  Regarding:

 'doing all of the above should likely buy chess another 1000 years'

In my opinion, not even close.  As soon as someone designs a computer
smart enough to improve itself, processing power explodes exponentially. 
The future is going to be way, way different than anything we can imagine
using the current paradigm.
-----------------------------------------------------
This is why I was starting a discussion on what that paradigm can be.  I
personally believe that the open-source method would lend well towards
this, via community consensus.  This could start with the chess variant
crowd, and them coordinating. I am offer people a chance to discuss the
IAGO Chess System as a starting point, from the drops and gating, to the
classification system, to the attempt to get Capablanca pieces onto an 8x8
board.  This could perhaps lead to a new paradigm. 

I would rather this be an opening for a discussion, rather than saying it
is merely beyond what we can imagine, so why bother.

Finally, I would say that all this is for more than just 'oh it lies beyond the computer'.  It is for the purpose of serving the fullness of the chess community.

George Duke wrote on Sat, Apr 5, 2008 03:54 PM UTC:

1) The classification system 'a-e' for Gating-drops is incomplete and
could easily become 'a-z' without much effort. That is, for anyone
accepting the prolificist ethos.
2) What if OrthoChess were 7x7 or 6x6 (closer to 1950's Los Alamos)? Then we would all suspect that, even with all Hutnik's suggested idiosyncratic
methods, Chess would become 'stuck' again in 10 or 80 years. 8x8 seems
to be right on the threshold, or cusp, or interface, of inevitable
difficulty, whatever is tried in later-added  Drops, randomized openings,
or introduced exotic pieces etc. However there is no proof of that.
3) Is Chess a formal axiomatic system, after David Hilbert and Kurt Godel? No, because Chess, except allegorically, does not purport to represent reality like Mathematics.  So long as all lines of play are not solved, then some lines are unsolved, probably most when considering higher board sizes; and probability as to best lines of attack have already entered players' algorithms for play, or heurisitic equations, without the need for more structuring of hidden information.

Rich Hutnik wrote on Sat, Apr 5, 2008 04:39 PM UTC:

George, I would like to make several comments here:
1. Gating is supposed to be a definition, and a subset of drops.  I am
sure there are lots of way to do this.  The purpose of it is to give
people an idea as to how it differs from a standard drop.  It also
involves the relocating pieces on the board.  I am sure that people can
come up with more.  I believe the key is to have a stable definition, and
then list some major examples.  I am of the belief it is an important term
to consider, debate, and reach an agreement over.  The end and final shape
isn't as important as what it is.	
2. The issue of the 8x8 board is that it is now a convention, and a
starting point, for testing, because it is what is readily available, and
has an established chess game being played with millions of people play it
(that being FIDE Chess).  What was suggested is ONE approach to this, as a
possible way.  And no, I disagree with you on the less than 100 years
approach.  What is suggested is to use ALL the possible variant
conventions as a way to expand chess here.  This means reserves, it means
mutators, it means different board condition.  And with the reserves, it
means changing the mix of pieces.  It also means more that this.  But what
does matter is there is a common foundation this is all to fit into. 
Chess960 isn't going to get stuck in 100 years, why do you think a larger
system will?  If you suggest that it will get stuck again in 100 years,
well then this site is doomed to be stuck within 100 years.  

As for the IAGO Chess System classes, well it is taking what is seen today
as chess and variants, and expanding it, as a way to think about it.  You
have standard stuff (A-Class).  Then it is suggested that there be an
evolutionary design, that has a B-Class migration to it.  C and M Class
represent the slower fixed one, and the M-Class as the version where a
chess game can migrate to.  In the B-Class I am proposing that the piece
mix map to the rules (so we don't have an 8 pawn promote to queens
problem, which breaks when you add any more pieces).  Then with the
variants, I propose that you have a V-Class for accepted variants that
work, along with mutators, and pieces.  And an X-Class where things can be
experimented with.  This is meant as a starting point of discussion.

3. Anything that is a set of rules is axiomatic, as the definition of
axiomatic is rules.  So game rules would apply also.  What Godel's
incompleteness theorem says that no system of rules can be both complete
and non-contradictory.  In other words, every set of rules will end up
producing more rules.  In other words, rules keep evolving.  This is valid
here.  And if you think that games have nothing to do with math, I am sure
that the game theory people will be surprised.  And Combinatorial Game
Theorists (this is the foundation abstract strategy games are built on)
would be shocked.

4 comments displayed

~~Earlier~~ ⇧Reverse Order⇩ ~~Later~~

Permalink to the exact comments currently displayed.