Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: verbal commentary (learning game plans from Fritz)

Author: margolies,marc

Date: 00:07:45 06/12/03

Go up one level in this thread


dear stu,
I don't think your imagination is wandering. But I do think your explanation
went for a small walk.
Here's the rub: if a player cannot find stimulation from a chess position yet
requires the "crutch" of hortatory remarks from his computer opponent to finish
a game well, then while he may be thinking, yet not thinking "independently." So
it's not good for Chess character (at least in my opinion)
I shall never damn anyone for needing help of cousre. I need plenty too. The
strength I think comes in the honesty of reaconing directly with that. But to
say some one is more indepedent because they rely on a machine instead of humans
-- gosh I dunno. half a dozen of one or the other!
best to ya, marc

On June 11, 2003 at 20:21:05, stuart taylor wrote:

>On June 11, 2003 at 15:11:17, margolies,marc wrote:
>
>>stu,
>> on reflection, I recall that you can toggle vebal commentary-- what you call
>>smart chatter-- as an insertion into analysis when running the fritz analysis
>>module. there are about I guess 300 different text messages the prog can use
>>like 'strengthening square d5' I think this is what you are looking for.
>>the fritzy wont do this in real time while you play-- it's an analysis option.
>>
>>it should not become a playing mode i think,'coach' is enough! we dont need the
>>machine to do our thinking for us when we play that's too lazy for me!
>>
>>Also if you need to see how the machine might answer it's own moves there is
>>'shoot-out' mode to follow a variation.
>>
>>PS the main reason I will never use vebal commentary in engine analysis, when
>>the computer 'sounds too human' it becomes very credible -- and I know fritz
>>often misses important strategic plan when analyzing. This is clear to me
>>because I will show my games to strong IM here in New York (tournament games)
>>and when I have time to run Fritz overnight,say, on a game first ..sure it will
>>show me my blunders ,BUT if I miss an important PLAN (so necessary for LEARNING
>>how to play) I rarely see any insight from A chessbase analytical product.
>>
>>Maybe that's the recomendation CB needs??
>>bye, marc
>
>Obviously one problem might be that there ARE not many insights.
>If it could say thing things like "what you just did exchanges an important
>piece for a less important piece", and why that is, then I think that would be a
>high enough concept to get me extremely involved in a game. And things like
>"beware on the light squares" etc. Even things like "a King side attack is/isn't
>now a wise/viable consideration" (with reasons too!).
> DURING a game? Well, if it mentions strategic aims, and it is for you to work
>out how to do them, that might be quite engaging. Later on in analysing you can
>see if and how the program itself does what it preached.
>If Fritz sounds too human, you can learn to argue with it by trying to see if
>they are wrong etc. You can then develope great independence in thinking! (Am I
>imagining?)
>Also, at any time that the machine changes its mind after seeing deeper (e.g.
>after 10 seconds, 30 seconds, 4 minutes.......) the chatter has to say so, and
>to say what it thinks now. So you will also be used to imagining that it might
>change its mind anyway.
>S.Taylor



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.