Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Sorry but wishful thinking will NOT help Mr. Gulko ?!

Author: Uri Blass

Date: 07:12:20 03/09/02

Go up one level in this thread


On March 09, 2002 at 09:21:09, Terry McCracken wrote:

>On March 08, 2002 at 12:08:49, Robert Hyatt wrote:
>
>>On March 08, 2002 at 08:20:41, Jorge Pichard wrote:
>>
>>>On March 08, 2002 at 05:06:31, Chris Carson wrote:
>>>
>>>>On March 07, 2002 at 20:09:25, mike schoonover wrote:
>>>>
>>>>>i can see where this thread is going to turn into a real
>>>>>pissing contest.  :)
>>>>>i wish the best to gulko.he his a fine chess player.
>>>>
>>>>I do to, I have always enjoyed his games and respected his play.
>>>>
>>>>>he's the best sentimental favorite so far.
>>>>>go G!give them processers stack overflows.
>>>>
>>>>Give em hell, hope he wins and takes them down.
>>>>
>>>>>regards
>>>>>mike
>>>
>>>Sorry to say this but wishful thinking will NOT help Mr.Gulko, the only thing
>>>that would help him is if the organizer dicides to use a slow computer like a
>>>Athlon 700 Mhz. But at this time control even a mere 700 Mhz would tare him
>>>apart.
>>
>>
>>Maybe or maybe not.  I watched a GM "dismantle" a very fast AMD box last
>>week 6 games in a row...  and 8 out of 10 games overall.  And the time control
>>was a very slow 3 0...
>>
>>The GMs are dangerous.  _very_ dangerous.  To think otherwise is very foolish.
>>Chess is a _long_ way from being solved by computers...
>
>All GM's are dangerous, plain and simple.
>
>You're correct, most people here haven't a clue about top young and often
>relatively unkown GM's that play Bullet Chess and often beat comps. on ICC.
>Not to say older and even some old GM's are not impressive!
>Beating them at 3 0 is most impressive!

This is not my opinion

It is more easy to set a prepared a trap at home before the game for a 3 0 game
then to prepare a trap for a slow time control game.

I will be more impressed if they can beat some non deterministic good machine 8
times out of 10 at 3 0.

It is possible to do it by doing a random changes in the evaluation.
The program is not going to play significantly worse if you do small random
change in the weights of the evaluation(not more than 5%) but it is going to be
harder to repeat a game that the player prepared at home and I suspect that part
of the games that players win against machines are games that they prepared at
home based on their own copy of the same program(learning cannot help because it
is a different copy).

Uri



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.