Main
Date: 15 Apr 2006 23:40:39
From: Buller
Subject: Improving using chess computers
What are people's ideas on the best way to improve using chess computers or
programs? I have a machine with a set of beginner and intermediate levels
that are fairly close together in strength (plus a bunch of other
levels/time controls), and I've been playing a series of games against those
levels. When I win get 3 points in the last 4 games, I advance to the next
level, and when I get only 1 point in the last 4 games, I go back to the
previous level. So I'm always playing opponents in a narrow range of
strength. I wonder if it would be better to frequently or occasionally play
a much higher level and get beaten soundly, in order to experience some
harsher lessons. I know that either way, I should be analyzing my wins and
losses. Any ideas? Thanks.

Darryl






 
Date: 17 Apr 2006 10:25:26
From: David Richerby
Subject: Re: Improving using chess computers
Buller <[email protected] > wrote:
> What are people's ideas on the best way to improve using chess
> computers or programs?

The usual recommendation is to play mainly against people and use the
computer to help you analyze your games afterwards. The best way to
do that is analyse yourself and just use the computer to check that
you didn't miss anything. A GM recently advised me, `Fritz will help
you to find best moves (often but not always) but will not teach you
to understand chess!'


> I have a machine with a set of beginner and intermediate levels that
> are fairly close together in strength (plus a bunch of other levels/
> time controls), and I've been playing a series of games against
> those levels. When I win get 3 points in the last 4 games, I
> advance to the next level, and when I get only 1 point in the last 4
> games, I go back to the previous level.

I'd recommend sticking with the stronger opponent for more than four
games even if things aren't going well. You learn much more from
losing than from winning. I've been playing practice games against
Fritz and bumping its strength up by fifty points once I reckon I'm
about equal with it. I usually lose ten or fifteen straight games
after that but it gives me a better idea of what I'm doing wrong.


> So I'm always playing opponents in a narrow range of strength. I
> wonder if it would be better to frequently or occasionally play a
> much higher level and get beaten soundly, in order to experience
> some harsher lessons.

Probably not. That'll just tell you that you're crap at the opening,
middlegame and endgame, which isn't too helpful. Better to progress
in small steps and not know how many hundreds of yards the GMs are
ahead of you. :-) It can be useful to play the occasional game
against a weaker opponent, though, to make sure you can win won
positions.


> I know that either way, I should be analyzing my wins and
> losses.

Definitely. Check that you and your opponent didn't miss anything
important in your wins but spend more time on the losses.


Dave.

--
David Richerby Sadistic Homicidal Hat (TM): it's like
www.chiark.greenend.org.uk/~davidr/ a hat but it wants to kill you and it
wants to hurt you!


  
Date: 18 Apr 2006 00:17:27
From: Simon Waters
Subject: Re: Improving using chess computers
On Mon, 17 Apr 2006 10:25:26 +0100, David Richerby wrote:

>
> You learn much more from losing than from winning.

This season I'm discovering you can learn just as much from the games you
win, as long as you win them inspite of what you did early on, rather than
because of it.

This approach is much easier on ones grade, but does require some
co-operation from your opponents ;)

--
http://simonwaters.technocool.net/



   
Date: 18 Apr 2006 10:23:07
From: David Richerby
Subject: Re: Improving using chess computers
Simon Waters <[email protected] > wrote:
> On Mon, 17 Apr 2006 10:25:26 +0100, David Richerby wrote:
>> You learn much more from losing than from winning.
>
> This season I'm discovering you can learn just as much from the
> games you win, as long as you win them inspite of what you did early
> on, rather than because of it.

It's well-known that a game of chess is won by the person who makes
the next-to-last mistake. So long as you ensure that the last mistake
is made by your opponent, that means you're free to make as many
educational mistakes before then as you want. :-)


Dave.

--
David Richerby Hilarious Newspaper (TM): it's like
www.chiark.greenend.org.uk/~davidr/ a daily broadsheet but it's a bundle
of laughs!


  
Date: 17 Apr 2006 22:58:56
From: Buller
Subject: Re: Improving using chess computers
David,

That was excellent help. Thanks.

Darryl

"David Richerby" <[email protected] > wrote in message
news:AUy*[email protected]...
> Buller <[email protected]> wrote:
>> What are people's ideas on the best way to improve using chess
>> computers or programs?
>
> The usual recommendation is to play mainly against people and use the
> computer to help you analyze your games afterwards. The best way to
> do that is analyse yourself and just use the computer to check that
> you didn't miss anything. A GM recently advised me, `Fritz will help
> you to find best moves (often but not always) but will not teach you
> to understand chess!'
>
>
>> I have a machine with a set of beginner and intermediate levels that
>> are fairly close together in strength (plus a bunch of other levels/
>> time controls), and I've been playing a series of games against
>> those levels. When I win get 3 points in the last 4 games, I
>> advance to the next level, and when I get only 1 point in the last 4
>> games, I go back to the previous level.
>
> I'd recommend sticking with the stronger opponent for more than four
> games even if things aren't going well. You learn much more from
> losing than from winning. I've been playing practice games against
> Fritz and bumping its strength up by fifty points once I reckon I'm
> about equal with it. I usually lose ten or fifteen straight games
> after that but it gives me a better idea of what I'm doing wrong.
>
>
>> So I'm always playing opponents in a narrow range of strength. I
>> wonder if it would be better to frequently or occasionally play a
>> much higher level and get beaten soundly, in order to experience
>> some harsher lessons.
>
> Probably not. That'll just tell you that you're crap at the opening,
> middlegame and endgame, which isn't too helpful. Better to progress
> in small steps and not know how many hundreds of yards the GMs are
> ahead of you. :-) It can be useful to play the occasional game
> against a weaker opponent, though, to make sure you can win won
> positions.
>
>
>> I know that either way, I should be analyzing my wins and
>> losses.
>
> Definitely. Check that you and your opponent didn't miss anything
> important in your wins but spend more time on the losses.
>
>
> Dave.
>
> --
> David Richerby Sadistic Homicidal Hat (TM): it's
> like
> www.chiark.greenend.org.uk/~davidr/ a hat but it wants to kill you and
> it
> wants to hurt you!




 
Date: 16 Apr 2006 01:21:31
From: marcosnpesic
Subject: Re: Improving using chess computers
Try playing real persons...

(FICS) http://www.freechess.org/ (real-time play)

(QueenAlice) http://www.queenalice.com/ (postal "timed" games)

Then use your favorite program to analise yor games...
So you wii be able to see "what=B4s the problem"...
Are you outplayed on the oppening ? Falling on traps on the
beggining... Study Oppenings... and tatics...
This Happens on midle game as well... -- > Study tatics...
You achieve a "equal" material endagame, but are lossing more than
winning -- > study end games...
You are playings always with "material balance"... and so... and so...
and then "boom" you sundely lose a piece... or receive a huge atack...
-- > study strategy...

The problem with playing "computers" is that they are very predictable
when using "personalites" always going in same line...

I think that you have to use computer as a "mentor" to give lesons, for
some praticing and evaluation... But for overral training or just for
fun, we all should play "humans"...

XD



  
Date: 18 Apr 2006 00:47:35
From: Wilma
Subject: Re: Improving using chess computers
That really works. I used CM9k to go over my games, and I came up with a
check list of how I lost. When I use the check list, I get much higher
scores in rated tournaments, and analysis shows I play better games.
That increases my enjoyment and my self-esteem at the same time. :)

Wilma


"cosnpesic" <[email protected] > wrote in
news:[email protected]:

> Try playing real persons...
>
> (FICS) http://www.freechess.org/ (real-time play)
>
> (QueenAlice) http://www.queenalice.com/ (postal "timed" games)
>
> Then use your favorite program to analise yor games...
> So you wii be able to see "what�s the problem"...
> Are you outplayed on the oppening ? Falling on traps on the
> beggining... Study Oppenings... and tatics...
> This Happens on midle game as well... --> Study tatics...
> You achieve a "equal" material endagame, but are lossing more than
> winning --> study end games...
> You are playings always with "material balance"... and so... and so...
> and then "boom" you sundely lose a piece... or receive a huge atack...
> --> study strategy...
>
> The problem with playing "computers" is that they are very predictable
> when using "personalites" always going in same line...
>
> I think that you have to use computer as a "mentor" to give lesons,
for
> some praticing and evaluation... But for overral training or just for
> fun, we all should play "humans"...
>
> XD
>
>