Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: GM Smirin vs 4 comps - Match Predictions

Author: Chris Carson

Date: 10:28:28 04/17/02

Go up one level in this thread


On April 17, 2002 at 13:23:11, Robert Hyatt wrote:

>On April 17, 2002 at 12:19:15, Chris Carson wrote:
>
>>On April 17, 2002 at 10:58:13, Terry McCracken wrote:
>>
>>>On April 17, 2002 at 09:07:04, Chris Carson wrote:
>>>
>>>>On April 17, 2002 at 03:33:12, Uri Blass wrote:
>>>>
>>>>>On April 17, 2002 at 03:04:52, Terry McCracken wrote:
>>>>>
>>>>>>On April 17, 2002 at 01:31:51, Joe Little wrote:
>>>>>>
>>>>>>>On April 16, 2002 at 18:28:36, Otello Gnaramori wrote:
>>>>>>>
>>>>>>>>On April 16, 2002 at 17:19:13, martin fierz wrote:
>>>>>>>>>a program which plays a game like shredder vs.
>>>>>>>>>smirin is just not GM strength. it is 3000+ in tactics and 2000- in positional
>>>>>>>>>play.
>>>>>>>>
>>>>>>>>I think that chess is made over 90% of tactics... so 2700+ is not an optimistic
>>>>>>>>evaluation.
>>>>>>>>
>>>>>>>>Regards,
>>>>>>>>Otello
>>>>>>>
>>>>>>>
>>>>>>> I agree, seems pretty obvious to me but who am I?
>>>>>>
>>>>>>Yes Chess is 90% tactics at least, maybe even 95%! But that would still only
>>>>>>mean that programmes play around 2700 level in tactics only, not in positional
>>>>>>play and planning, which is _fundamental_ and till a programme aquires this
>>>>>>skill in won't be a _complete_ Grandmaster.
>>>>>>
>>>>>>Planning is many years away, positional play is advanced a long ways but still
>>>>>>needs improvement.
>>>>>>
>>>>>>Computers will play 2800+ in tactics long before it can actually manage deep
>>>>>>positional play, let alone planning.
>>>>>
>>>>>I think that computers are 2800+ in tactic even today and it is not something
>>>>>about the future.
>>>>>
>>>>>tactics is not only long combinations but mainly short combinations.
>>>>>No human can see every short combination that programs has no chance to miss.
>>>>>
>>>>>Uri
>>>>
>>>>I agree with you Uri.  I would add that computers still have some problems with
>>>>chess knowledge, however they make up for it with tactics.  I would also add
>>>>that Planning and Preparation are the keys (or just plain luck) to beating the
>>>>comps.  Do you have any additional ideas on beating the comps?
>>>>
>>>>I also agree with your evaluation on opening preparation.  I see no ethical
>>>>reason why a program must use a known book.  GM's get to use any
>>>>book/preparation they choose, it is only fair that the comps get to change their
>>>>books. (This discussion was in a different part of this thread).
>>>
>>>
>>>What's your rating Chris as you need to be 2800 to know whether or not
>>>Comps./Programmes have attained this level of sophistication.
>>
>>My rating is in my profile.  It is nowhere near 2800.  You have posted your
>>opinion on short vs long tactics in this and other threads, I wonder if you
>>follow the "need to be 2800 to know" rule for yourself?  My opinion is just
>>that.  If you do not like it, I can live with that.
>>
>>>
>>>Kasparov and Kramnik say todays' programmes on top PC's are not 2800+ in the
>>>tactical arena.
>>
>>Well, Kramnik recently said that Fritz 7 on 8-processors "definitely be over
>>2800 in its Elo performance".  Last time I checked, that was ELO 2800+ and I
>>believe it to be higher in tactics than in positional play.
>>See: http://www.chessbase.com/newsdetail.asp?newsid=255
>>
>>I doubt you can support your statement with a more recent Kramnik quote.
>>
>
>Just as surely as I believe that the statement by Kramnik is pure pre-match
>hyperbole, having nothing to do with facts at all.  As Kasparov has stated
>although again, I would take _his_ statements with a grain of salt as well as
>everybody has an agenda in this circumstance...
>
>But forget the 2800+ stuff.  It ain't gonna happen...

I agree with you.  GM Kramnik will win easily.  I was responding to the person
above, not sure why you answered?

>
>
>
>
>>>
>>>They miss too much in long range tactics due to the "Horizon Effect".
>>>
>>>Terry



This page took 0.01 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.