Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: The problem with big-O is one of definitions

Author: José Carlos

Date: 06:02:57 05/10/01

Go up one level in this thread


On May 09, 2001 at 20:39:50, Dann Corbit wrote:

>On May 09, 2001 at 20:28:29, Peter McKenzie wrote:
>
>>On May 09, 2001 at 20:05:24, Dann Corbit wrote:
>>
>>>On May 09, 2001 at 20:00:09, Peter McKenzie wrote:
>>>
>>>>On May 09, 2001 at 19:34:08, Dann Corbit wrote:
>>>>
>>>>>On May 09, 2001 at 19:31:32, Ricardo Gibert wrote:
>>>>>[snip]
>>>>>>>If someone pays you to give an algorithm analysis of chess will you really
>>>>>>>report that it is O(1)?
>>>>>>
>>>>>>
>>>>>>Yes and I will point to the access of Nalimov EGTBs as an example of such an
>>>>>>algorithm. I will observe that in principle 5-man EGTBs can be extended to
>>>>>>32-man EGTBS, though this has no practical significance.
>>>>>
>>>>>This is an incompetent assessment.  32 man EGTB's cannot even conceivably be
>>>>>attempted if half the universe were turned into computers and the other half
>>>>>computed madly until the power went out.
>>>>
>>>>Dan, it seems to me that Ricardo is presenting a logical argument here.  I don't
>>>>think the argument is refuted by you calling it incompetent!
>>>>
>>>>Similarly, I don't see why the practical difficulties of constructing 32 man
>>>>EGTBs should detract from their theoretical existance.
>>>
>>>Because if you can't make one, it won't ever exist.
>>
>>I don't think that matters in this context.  Two points:
>>
>>1) Just because we can't make one now doesn't mean we will never be able to make
>>one.  For 32 EGTBs, we may discover for example some incredible compression
>>scheme that will allow them to be built with the state of the art technology in
>>20 years time.
>>
>>2) In any case, we were talking about 'theory' weren't we?  Who cares if we can
>>built it or not.  It is more of a thought experiment.
>>
>>>
>>>Sort of like powering a rocket to the moon with a spoonful of baking soda and a
>>>tablespoon of vinegar.  "You can't get there from here."
>>>
>>>"Why should we let reality get in the way of a perfectly good discussion?"
>>>He asked himself aloud.
>>>
>>>Let me ask you (Peter) as someone who seems very sensible and 'detached from the
>>>heat of the battle' --
>>>
>>>Do you consider an algorithm like Alpha-Beta with PV search for the game of
>>>chess to be exponential or O(1) or something else?  Why do you so answer?
>>
>>I would consider THIS ALGORITHM to be O(exp(n)) where n is search depth between
>>1 and MAX_N, where MAX_N is some large N (deepest node required to solve chess
>>by search).
>>
>>But I would consider the game of chess to have a theoretical complexity of O(1)
>>since it can in theory be solved by a big lookup table.
>>
>>Sound reasonable?
>
>I will never use that definition.  It isn't useful for computation.
>I don't think that is the way it is done in the industry.
>
>It is frightening to think that I might have an algorithm analyzed, and get an
>answer "It is O(1)" when it is completely intractible.
>
>It seems to me that there should be a way to clearly define this in an method
>that is both unambiguous and useful.
>
>"Chess is O(1)" isn't it.

  I've read this thread quite a bit late, so I probably have missed some posts
somewhere else in the page, but I'll give my opinion (sorry if it's been already
discussed somewhere else):

  Complexity, AFAIK, don't apply to _problems_, but to _algorithms_. So,
different algorithms for solving chess have different complexities. A table
lookup algorithm for solving chess (if such a thing is possible) is O(1).
AlfaBeta with PVS is O(exp(n)), being n the depth.
  I don't see how this can be so confusing.

  José C.



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.