| |
Main
Date: 28 Sep 2006 07:41:08
From: Patrick
Subject: Any tips for computer chess?
|
I seem to be a perpetual novice. I learned chess over forty years ago,
and I've studied books on it and admired the game all this time. But I
haven't practiced enough to become what I'd consider a decent player
(an intermediate-level player, I guess).
Practice is obviously what I need. But how to go about it exactly?
First off, I don't play against people. I know that may sound weird,
but I'm a noncompetitive kind of guy, and I have no interest in beating
people or being beaten by them. I do all my chess playing against a
computer AI.
A year or so ago, I bought Fritz 8, with the intention of using it as a
chess mentor--to practice and also to review famous games. I used it
for a while but didn't stick with it. Many of its features are too
advanced for me (and its database is enormous). I ended up ignoring
everything except "friend mode" and the handicap/just-for-fun games.
Just recently I bought Chessmaster 9000, because I heard that it has a
"chess academy" that's about as good as CM 10th. Indeed, I've been
enjoying working my way through the chess problems. I'm also impressed
with the way I can play a game, then have CM analyze it, and go back
through, reading the annotations. (Fritz can do that too, I guess, but
the one time I tried it, it took forever to finish the analysis--I ran
it overnight--and then the annotations didn't seem very helpful
anyway.)
I loaded a "famous game" from the CM library and read through it, but
the annotations are so sparse that I didn't learn much from it. I
guess I should try having CM analyze the game first, after which there
should be annotations for more of the moves.
Btw, I hate the CM graphic interface. The windows often can't be sized
to fit neatly, so I'm always having to do without something on the
screen. The small 3D boards fit better, but I hate using them; I
always go back to 2D.
I'm also studying my way through chess books, which of course offer
more in the way of explanation.
But I'm wondering what more I can do to get the most out of CM 9000 (or
Fritz 8).
Another big question right now is time controls. In tournament mode
(for rated games), it doesn't seem possible to have separate time
controls. I don't want to sit and wait a long time for the computer to
make its move, but I don't want to be under time pressure myself either
(I blunder enough without having to worry about that).
I also have a couple handheld chess computers left over from last time
I caught the chess bug. I told myself I'd take the game with me
wherever I went--and I did for a while. One is a pretty high quality
(and fairly expensive) Novag machine that's just been gathering dust.
I bought it because it's loaded with features, plays a strong game, and
yet has only an algebraic LCD display; I wanted to force myself to use
it along with a real chess set so I could enjoy the look and feel of a
standard-sized game.
Anyhow, I hate letting all this hardware and software go to waste.
Years ago, when CM3000 was new, I bought a book called "Mastering the
Chessmaster"--a user guide that told how to make the most of the
software. I could use something like that for today's Chessmaster (or
Fritz).
--Patrick
|
|
| |
Date: 29 Sep 2006 17:33:46
From:
Subject: Re: Any tips for computer chess?
|
[email protected] wrote:
> You're right, ! fits better than ?!. Crafty suggests 18.Be2 Nc2+ 19.Kf1
Or rather, ?! fits better than !. :)
-- likesforests
---
Become a Chess Expert
http://likesforests.blogspot.com/
|
| |
Date: 29 Sep 2006 17:30:33
From:
Subject: Re: Any tips for computer chess?
|
David Richerby wrote:
> <[email protected]> wrote:
> Neither side ever had an attack in the game you quote. And the whole
> point of playing with increments is to avoid running out of time.
Increments are great... I don't like winning via wood-pushing. Black
thought he had an attack. I was content to win material by defending.
>> 1.c4 e5 2.Nc3 Nf6 3.Nf3 Nc6 4.e3 Bb4 5.Qc2
Book: English Opening / Four Knights
>> 5...e4? 6.Nxe4
5...e4? is a mistake. The e4-square had two attackers and three
defenders.
>> 6...O-O 7.a3 Nxe4 8.axb4 Re8 9.b5 Nb4
The obvious recapture was 8.Qxe4 Re8 9.Qc2 Bc5. I decided on 8.axb4 Re8
9.b5 Nb4 because I get the bishop pair (imbalance) and Black has more
chances to go wrong. The doubled b-pawns didn't look like trouble here.
>>10.Qb3 c5 11. bxc6 a5
>
> Correct here is 11... Nxc6. It's a theme that crops up from time to
> time and is well worth knowing.
Yup, that would have saved him a pawn.
> > 12.cxb7 Bxb7 13.d3 Nc5 14.Qc3 Bc6 15.d4 Na4 16.Qd2 Be4?? 17.Rxa4 Bb1
With 16...Be4??, he must have been planning Nc2+. If Black prepared
with Nb6 first, his combination might have been effective.
> An odd-looking move.
17...Bb1 seems like a waiting move--he probably didn't know how to
continue. 17...Nc2+ would have meant no castling. 17...Rb8 or 17...Qb6
would have respectable moves bringing his army closer to the action.
> > 18.Rxb4!
>
> I don't think that's worth a ! -- I'd give it a ?! or, more likely,
> just a plain ?. You've sacrificed the exchange for a pawn and the
> right to castle: I don't see the value in that and nor does Fritz 8,
> which scores the position as +3.91 to 14 ply after 18.Rxb4 and
> +4.66/15 after 18.Be2: in other words, it sees no significant
> difference between the positions other than material.
I thought of the sacrifice as an immediate way to clarify the win.
18.Rxb4?! axb4 19.Bd3 Bxd3 20.Qxd3 Ra1 21.O-O. I sacrifice my rook for
his knight, ability to exchange bishops, ability to castle, and a few
easy follow-up moves worth 30-45 seconds on the clock.
At the end of the sequence I have extra time, a solid position, two
extra pawns, and two minor pieces for his rook which I could hold even
against a class B player.
You're right, ! fits better than ?!. Crafty suggests 18.Be2 Nc2+ 19.Kf1
Nb4 20.Na1 Na2 21.Bd3 Bxd3 22.Nxd3 Nxc1. As a result, White is two
pawns and a piece up.
> > 18... axb4 19.Bd3?!
>
> Surely, 19.Qxb4 ?
I figured I had enough material and wanted to simplify the position.
> > 19... b3?? 20.Bxb1 Ra1 21.Bd3 1-0
>
>
> Dave.
|
| | |
Date: 02 Oct 2006 10:26:25
From: David Richerby
Subject: Re: Any tips for computer chess?
|
<[email protected] > wrote:
> David Richerby wrote:
>> I don't think that's worth a ! -- I'd give it a ?! or, more likely,
>> just a plain ?. [...]
>
> I thought of the sacrifice as an immediate way to clarify the win.
> 18.Rxb4?! axb4 19.Bd3 Bxd3 20.Qxd3 Ra1 21.O-O. I sacrifice my rook
> for his knight, ability to exchange bishops, ability to castle, and
> a few easy follow-up moves worth 30-45 seconds on the clock.
The extra time on the clock could be pretty important in a 2-12 game,
yes. I'd forgotten to take that into account. And, as you say, the
position is still comfortably won.
Dave.
--
David Richerby Old-Fashioned Moistened Drink (TM):
www.chiark.greenend.org.uk/~davidr/ it's like a refreshing juice beverage
but it's moist and perfect for your
grandparents!
|
| |
Date: 29 Sep 2006 21:23:48
From: pulgao
Subject: Re: Any tips for computer chess?
|
On 28 Sep 2006 07:41:08 -0700, "Patrick" <[email protected] >
wrote:
>Anyhow, I hate letting all this hardware and software go to waste.
>Years ago, when CM3000 was new, I bought a book called "Mastering the
>Chessmaster"--a user guide that told how to make the most of the
>software. I could use something like that for today's Chessmaster (or
>Fritz).
See:
http://www.chesscentral.com/lopez/guide-computer-chess.htm
-- Steve Lopez
|
| |
Date: 29 Sep 2006 11:48:19
From:
Subject: Re: Any tips for computer chess?
|
[email protected] wrote:
> Patrick wrote:
> > I was impressed with Fritz's 250K-game database when I first installed
> > it. Then, after loading a couple games and looking at them, I suddenly
> > realized I had no use for all that.
>
> A game database is helpful for learning about an opening.
>
> For example, yesterday I played a game that went 1.c4 e5 2.Nc3 Nf6
> 3.Nf3 Nc6 4.e3 Bb4. This is the English Opening / Four Knights.
>
> I played 5.Qc2, but noted 5.Nd5 looked strong. After the game, I opened
> my database to check. I found that 5.Qc2 is the mainline and wins 50.5%
> while 5.Nd5 wins only 46.5%.
>
> So 5.Qc2 may be better, but why? I follow a few games. After 5.Nd5 e4!
> 6.Ng1 White has to undevelop his knight, which looks rather ugly. After
> 5.Qc2 Bxc3 Qxc3 development is equal... so now I know why 5.Qc2 is
> better and will play it in future games.
>
> Quickly playing through some GM-level games in those lines would also
> help me get a feel for the typical continuations, middlegame plans, and
> endgames in that line.
The actual game, at 2 minutes + 12 seconds time controls. At fast time
controls, humans sometimes press an attack too hard, hoping to exhaust
their opponent's clock.
1. c4 e5 2. Nc3 Nf6 3. Nf3 Nc6 4. e3 Bb4 5. Qc2 e4? 6. Nxe4 O-O 7. a3
Nxe4 8. axb4 Re8 9. b5 Nb4 10. Qb3 c5 11. bxc6 a5 12. cxb7 Bxb7 13. d3
Nc5 14. Qc3 Bc6 15. d4 Na4 16. Qd2 Be4? 17. Rxa4 Bb1 18. Rxb4! axb4 19.
Bd3 b3? 20. Bxb1 Ra1 21. Bd3 1-0
|
| | |
Date: 29 Sep 2006 20:37:47
From: David Richerby
Subject: Re: Any tips for computer chess?
|
<[email protected] > wrote:
> The actual game, at 2 minutes + 12 seconds time controls. At fast
> time controls, humans sometimes press an attack too hard, hoping to
> exhaust their opponent's clock.
Neither side ever had an attack in the game you quote. And the whole
point of playing with increments is to avoid running out of time.
> 1.c4 e5 2.Nc3 Nf6 3.Nf3 Nc6 4.e3 Bb4 5.Qc2 e4? 6.Nxe4 O-O 7.a3 Nxe4
> 8.axb4 Re8 9.b5 Nb4 10.Qb3 c5 11.bxc6 a5
Correct here is 11... Nxc6. It's a theme that crops up from time to
time and is well worth knowing.
> 12.cxb7 Bxb7 13.d3 Nc5 14.Qc3 Bc6 15.d4 Na4 16.Qd2 Be4? 17.Rxa4 Bb1
An odd-looking move.
> 18.Rxb4!
I don't think that's worth a ! -- I'd give it a ?! or, more likely,
just a plain ?. You've sacrificed the exchange for a pawn and the
right to castle: I don't see the value in that and nor does Fritz 8,
which scores the position as +3.91 to 14 ply after 18.Rxb4 and
+4.66/15 after 18.Be2: in other words, it sees no significant
difference between the positions other than material.
> 18... axb4 19.Bd3
Surely, 19.Qxb4 ?
> 19... b3? 20.Bxb1 Ra1 21.Bd3 1-0
Dave.
--
David Richerby Flammable Slimy Hi-Fi (TM): it's like
www.chiark.greenend.org.uk/~davidr/ a music system but it's covered in
goo and it burns really easily!
|
| |
Date: 29 Sep 2006 11:36:10
From:
Subject: Re: Any tips for computer chess?
|
Patrick wrote:
> I was impressed with Fritz's 250K-game database when I first installed
> it. Then, after loading a couple games and looking at them, I suddenly
> realized I had no use for all that.
A game database is helpful for learning about an opening.
For example, yesterday I played a game that went 1.c4 e5 2.Nc3 Nf6
3.Nf3 Nc6 4.e3 Bb4. This is the English Opening / Four Knights.
I played 5.Qc2, but noted 5.Nd5 looked strong. After the game, I opened
my database to check. I found that 5.Qc2 is the mainline and wins 50.5%
while 5.Nd5 wins only 46.5%.
So 5.Qc2 may be better, but why? I follow a few games. After 5.Nd5 e4!
6.Ng1 White has to undevelop his knight, which looks rather ugly. After
5.Qc2 Bxc3 Qxc3 development is equal... so now I know why 5.Qc2 is
better and will play it in future games.
Quickly playing through some GM-level games in those lines would also
help me get a feel for the typical continuations, middlegame plans, and
endgames in that line.
|
| |
Date: 29 Sep 2006 07:46:44
From: Patrick
Subject: Re: Any tips for computer chess?
|
Dave (from the UK) wrote:
> Patrick wrote:
>
> > First off, I don't play against people. I know that may sound weird,
> > but I'm a noncompetitive kind of guy, and I have no interest in beating
> > people or being beaten by them. I do all my chess playing against a
> > computer AI.
>
> I don't know about others, but I feel that playing people gives far more
> diversity in games. If you don't want to sit over a board, I think
> playing online chess might be a decent compromise. Also, some opponents
> will agree to analyse a game afterwards. That is helpful to both you and
> your opponent.
>
> I personally tend to blunder more when playing against a computer. I
> concentrate far less than if playing a person.
I guess I tend to blunder more too. But for me, paradoxically, this is
part of the reason I so far play exclusively against the computer AI.
My track record in chess is pretty dismal. I learned the game at age
12 (in 1967), and ever since then I've wished I could be a decent
player (i.e., an intermediate-level player). Yet I played very rarely
back then, and when I got to high school I refused to join the chess
club (it was very uncool from my teen POV). Instead I bought books and
studied the game by myself at home, practicing with a book called
"Solitaire Chess."
When I did play a game, what typically happened is that I'd get way
ahead positionally in the beginning (because unlike my opponents, I had
read chess books and understood some things about opening strategy) but
then make one or more tactical blunders and lose.
When the home computer came along (for us it was the late 1980s), I
rushed out and bought Chessmaster 2100, thinking that finally I'd be
able to practice enough to become a decent player. For a while I did
practice, but it seemed there was a huge gap between Novice and
Intermediate levels. I could play a comfortable game and win often
enough on Novice level, but I'd get crushed on Intermediate level
almost every time unless I concentrated *very* hard and took my time.
Being forced to concentrate very hard soon meant (to me) that chess was
all work and no play. All the fun went right out of it, and I got
discouraged.
Yet, my admiration of chess persisted, and periodically my enthusiasm
would return. I bought CM3000 and repeated the above experience. I
bought CM4000 and CM5500 when they were new--and again I had
experiences very similar to the above (though in CM5500 I enjoyed doing
the chess problems).
When I sit down to a game of chess, sometimes I'm up for a serious
challenge. Other times I'm drained after a long day, and I just want
to play at it for fun. At times like those, I much prefer to play a
lower-rated opponent so I can just be amused by his blunders, casually
take advantage of them, and go on to win without having to strain my
brain too much.
I read an article online yesterday (by Dan Heisman, IIRC) which said
that about 65 percent of one's games should be against slightly
higher-rated players, the rest against slightly lower-rated players.
The latter kind of game is meant to give one a break from being on the
cutting edge; it's a reminder to the player that he really is making
progress and has become strong enough to normally win against players
of a lower caliber.
I don't want to get in the habit of just trouncing weak players all the
time, never stepping up to an honest challenge. But OTOH, I do want
to--and *need* to--keep chess fun. As I mentioned above, I have a poor
track record: I keep getting discouraged and losing interest whenever
chess becomes too much work.
And this is part of the reason I don't play against people. In "real
life," I don't know anyone who plays chess anyway (I did have an IM
acquaintance for a while, but he died a few years ago). And my
experience with games I've played online is that they can be *very*
stressful. I haven't played much chess online, but I've played in
backgammon and cribbage tournaments (and even managed first place in
one of those)--and my blood pressure was up and my nerves really on
edge. It was so much more like work than play that I don't relish the
thought of going back to that.
In short, chess is just a hobby for me--a casual pastime; it's
something to kick back and enjoy at the end of a long day's work, or to
spend a few hours on during a weekend. It will be a much more
rewarding hobby if I can make some gradual progress and maybe go from
being a perpetual novice to an intermediate-level player. But above
all, I have to do whatever it takes to keep it fun. If I take it too
seriously and push myself beyond a certain point, I'm bound to once
again leave in a huff and turn to less rigorous games instead.
> > Another big question right now is time controls. In tournament mode
> > (for rated games), it doesn't seem possible to have separate time
> > controls. I don't want to sit and wait a long time for the computer to
> > make its move, but I don't want to be under time pressure myself either
> > (I blunder enough without having to worry about that).
>
> I've tried this, by forcing the computer to make moves quickly. I'm not
> convinced it is a good idea though. Since you can make use of your
> opponents time. For any given time control (say 30 minutes for all
> moves), you will play better if you give the computer 30 minutes too,
> since you will be thinking during its move.
OK. A few people have said this now, and it's sinking in. With older
chess software, you almost had to adjust the time controls. (In
Zillions of Games, I can't even do that. I set the AI level to
"pushover," and it still rips me to shreds.) Now it sounds like the
program is designed to dumb itself down to behave more like real people
of various ratings.
I appreciate that. I've never cared about time controls at all,
though--never played chess with a clock before. So that'll be a new
part of the experience for me, I guess.
An article I read yesterday suggested playing a mix of slow games and
fast games, so maybe I'll experiment with that.
> I believe you would improve more if you played human opponents and
> discussed the game afterwards with them. Even if you don't want to
> discuss it with them, I still think you should play humans. It is pretty
> anonymous over the internet, which may or may not be acceptable to you.
I may get to that point someday. Right now my main goal is to just
stick with chess, keep it fun, and not get discouraged and quit again.
Improvement is a secondary concern (though also important to me).
And I'm not one of those people who considers social interaction a huge
motivator. Quite the opposite, in fact. For instance, my wife and I
are both into physical exercise; but she won't do much of it unless she
joins a gym or has a workout partner or at least follows along with a
TV exercise show--she needs other people involved. In contrast, I've
been doing a daily workout for years--enjoying it and adding variety to
it and improving at it--completely on my own. I usually find the need
to socialize or compromise with other people to be a distraction that
slows me down. I generally prefer to keep goal-oriented activities
separate from my social life. When I'm interacting with people, I like
it to be more casual and unstructured.
--Patrick
|
| |
Date: 29 Sep 2006 11:55:22
From: Dave (from the UK)
Subject: Re: Any tips for computer chess?
|
Patrick wrote:
> First off, I don't play against people. I know that may sound weird,
> but I'm a noncompetitive kind of guy, and I have no interest in beating
> people or being beaten by them. I do all my chess playing against a
> computer AI.
I don't know about others, but I feel that playing people gives far more
diversity in games. If you don't want to sit over a board, I think
playing online chess might be a decent compromise. Also, some opponents
will agree to analyse a game afterwards. That is helpful to both you and
your opponent.
> Another big question right now is time controls. In tournament mode
> (for rated games), it doesn't seem possible to have separate time
> controls. I don't want to sit and wait a long time for the computer to
> make its move, but I don't want to be under time pressure myself either
> (I blunder enough without having to worry about that).
I've tried this, by forcing the computer to make moves quickly. I'm not
convinced it is a good idea though. Since you can make use of your
opponents time. For any given time control (say 30 minutes for all
moves), you will play better if you give the computer 30 minutes too,
since you will be thinking during its move.
I personally tend to blunder more when playing against a computer. I
concentrate far less than if playing a person.
I believe you would improve more if you played human opponents and
discussed the game afterwards with them. Even if you don't want to
discuss it with them, I still think you should play humans. It is pretty
anonymous over the internet, which may or may not be acceptable to you.
Computers are good to practice endgames if you have tablebases. Playing
to queen pawns, or defend against pawns being queen is well worth doing.
Without table bases though, chess computers can tend to be pretty
useless in the endgame.
--
Dave (from the UK)
Please note my email address changes periodically to avoid spam.
It is always of the form: [email protected]
Hitting reply will work for a few months only - later set it manually.
http://witm.sourceforge.net/ (Web based Mathematica front end)
|
| |
Date: 28 Sep 2006 20:59:14
From:
Subject: Re: Any tips for computer chess?
|
Patrick wrote:
> > > But I'm wondering what more I can do to get the most out of CM 9000
> >
> > Well, it offers a variety of computerized opponents and many tutorials
> > / quizzes. What else are you looking for?
>
> Not sure. Maybe nothing really. I guess I was kinda hoping for a
> computerized "mentor" to take me by the hand and walk me step by step
> through what all I need to do to play better and enjoy the game even
> more. Or maybe some advice on how to best take advantage of the
> database of famous games. I mean, if they're so poorly annotated,
> what's the point of running through them? Half the time I have no idea
> why a certain move is being made.
First let me start off by saying that the first people to reply to your
post have very good things to say.
As for the annotated games in Chessmaster, it's true that many of them
are sparsely annotated. For the first several versions of CM, the
annotations are not particularly useful. It is only starting with the
games that were added for CM8000 that the annotations start to become
more frequent and less cryptic. However, those games would only account
for about 15% of all of the Classic Games, so there might be a lot of
hunting and pecking. Generally, I don't go to the Classic Games unless
somebody elsewhere mentions a "classic game", and then I look to see if
it is in the Library.
> > > Another big question right now is time controls. In tournament mode
> > > (for rated games), it doesn't seem possible to have separate time
> > > controls. I don't want to sit and wait a long time for the computer to
> > > make its move, but I don't want to be under time pressure myself either
> > > (I blunder enough without having to worry about that).
> >
> > If you want different (unfair) time controls, you can play in the game
> > room instead of the tournament room. Those won't affect your rating.
> >
> > You say you "sit and wait" while the computer makes it move. Imagine
> > that the computer's time is extra time for you to consider the
> > position, just like in a real tournament game.
This last point is VITAL, and it is the reason why we did not allow
separate time controls for Rated Games like we did in the Game Room.
You definitely need to learn how to think on your opponent's time.
> Guess I just haven't played around enough with the controls--and I've
> been away from computer chess for a while. What I remember from old
> chess computers is that if you set them to infinite time, you can
> figure on letting it sit overnight (or longer) while the computer
> calculates its move. And if you give yourself and the computer the
> same amount of time, you'll be mercilessly clobbered every game.
Not necessarily true. See next paragraph.
> But I suppose that with all these rated AI players in CM9000, the
> program dumbs itself down enough that a 1300-rated AI player does in
> five minutes about what a 1300-rated human player would do in five
> minutes. Is that right? IOW, I don't need to monkey with the time
> controls in order to balance the game?
The CM9000 FAQ (which, unfortunately, no longer seems to be available
online) talks a fair bit about how the personalities were rated. Here
are the pertinent entries:
Q. How are the ratings for the computer personalities calculated?
A. The rating for the Chessmaster personality is based on several
factors: the Swedish Rating List (SSDF) for the engine in Chessmaster
6000 on a Pentium 90 processor, an estimation of the strength of the
NEW version of The King, based on the SSDF's rating of the OLD version,
and several thousand computer vs. computer and human vs. computer
games.
The SSDF rating was converted from the Swedish scale to the U.S. scale
and an adjustment was made for later improvements to the chess engine.
Then a series of computer vs. human and computer vs. computer
tournaments were played (a total of over 15,000 games, about 3,000 of
which were human vs. computer) at a time control of Fischer 5/10 (five
minutes, plus 10 seconds for each move). The results of these games
were fed into a ratings calculator.
Each of the personalities has a base rating as measured on a Pentium
II-450 (the minimum spec processor for Chessmaster 9000). When you run
Chessmaster 9000 on your system, the program performs a CPU benchk
measurement and adjusts the ratings accordingly. That formula is [Base
Rating] + 50 * ln([Your CPU Speed] / [Base CPU Speed]) / ln(2). A
reasonable rule of thumb to use is that each doubling of CPU speed
results in a rating increase of 50 points.
Q. How accurate are the ratings for the personalities?
A. The ratings are mathematically precise, but they are mostly based on
computer vs. computer games. While the human vs. computer games added
much more "real-world" accuracy to the ratings in Chessmaster 9000,
many of the Chessmaster personalities have quirks in their playing
styles that a human will be able to exploit but a computer opponent
will not be able to "see". Discovering and then taking advantage of
these weaknesses is part of the fun of Chessmaster, but you might find
yourself able to defeat certain personalities that you don't think you
should be able to based on their ratings.
jm
|
| |
Date: 28 Sep 2006 18:33:25
From: Patrick
Subject: Re: Any tips for computer chess?
|
[email protected] wrote:
> Patrick wrote:
>
> Watching grandmaster games is entertaining and instructive, but a
> computer should not be your priy source for annotations. Computers
> are tactical by nature--they easily spot hanging pieces and knight
> forks, which decide many beginner-level games.
>
> However, they are clueless at evaluating positional characteristics
> such as a lead in space or development, pawn structures, etc. If you
> want good annotated games, I recommend Chernev's Logical Chess: Move by
> Move or Nunn's Understanding Chess: Move by Move.
That rings true. Interestingly, Chernev's book was one of the first I
ever bought--almost forty years ago. I'm surprised to see it still
frequently recommended.
> > But I'm wondering what more I can do to get the most out of CM 9000
>
> Well, it offers a variety of computerized opponents and many tutorials
> / quizzes. What else are you looking for?
Not sure. Maybe nothing really. I guess I was kinda hoping for a
computerized "mentor" to take me by the hand and walk me step by step
through what all I need to do to play better and enjoy the game even
more. Or maybe some advice on how to best take advantage of the
database of famous games. I mean, if they're so poorly annotated,
what's the point of running through them? Half the time I have no idea
why a certain move is being made.
I was impressed with Fritz's 250K-game database when I first installed
it. Then, after loading a couple games and looking at them, I suddenly
realized I had no use for all that.
Anyhow, I popped into the Chess Cafe earlier today and also read an
article at Chessville. Between the two, I think I'm now oriented to
what I need to do.
Taking the advice of the Chessville article, I ordered three books:
Polgar's chess problems, Everybody's Second Chess Book, and
Pandolfini's endgame course. I figure the problems can be daily
exercises and the other two books can help ensure that I've got the
basics down.
Right now I'm re-studying Seirawan's "Play Winning Chess," which I
finished a year or two ago. I was in the middle of his "Tactics" book
when my interest in chess petered out last time. Figure that will be
good to get back to after the Pandolfini book.
> > Another big question right now is time controls. In tournament mode
> > (for rated games), it doesn't seem possible to have separate time
> > controls. I don't want to sit and wait a long time for the computer to
> > make its move, but I don't want to be under time pressure myself either
> > (I blunder enough without having to worry about that).
>
> If you want different (unfair) time controls, you can play in the game
> room instead of the tournament room. Those won't affect your rating.
>
> You say you "sit and wait" while the computer makes it move. Imagine
> that the computer's time is extra time for you to consider the
> position, just like in a real tournament game.
Guess I just haven't played around enough with the controls--and I've
been away from computer chess for a while. What I remember from old
chess computers is that if you set them to infinite time, you can
figure on letting it sit overnight (or longer) while the computer
calculates its move. And if you give yourself and the computer the
same amount of time, you'll be mercilessly clobbered every game.
But I suppose that with all these rated AI players in CM9000, the
program dumbs itself down enough that a 1300-rated AI player does in
five minutes about what a 1300-rated human player would do in five
minutes. Is that right? IOW, I don't need to monkey with the time
controls in order to balance the game?
Thanks for the replies. Guess I'm on my way again.
--Patrick
|
| | |
Date: 29 Sep 2006 10:29:22
From: David Richerby
Subject: Re: Any tips for computer chess?
|
Patrick <[email protected] > wrote:
> I guess I was kinda hoping for a computerized "mentor" to take me by
> the hand and walk me step by step through what all I need to do to
> play better and enjoy the game even more.
Computers can't do that for you: that's what you get from a human
coach.
> Or maybe some advice on how to best take advantage of the database
> of famous games. I mean, if they're so poorly annotated, what's the
> point of running through them?
They can be useful to find out what moves strong players play in the
opening and what kind of middle-games and endgames these openings tend
to lead to. If I want to get a feel for an opening, I'll often whizz
through ten or twenty GM games (only a couple of seconds a move) just
to see what happens in that kind of game.
> Half the time I have no idea why a certain move is being made.
Again, computers can't explain that, except in the really basic cases.
(Dude, he's just taken your queen with his. If you don't recapture,
you're gonna lose.) To develop your understanding of why moves are
made, you need a book of well-annotated master games and/or the
guidance of a human coach. Another approach is to ask here
(rec.games.chess.analysis would be more appropriate for that kind of
question.)
> And if you give yourself and the computer the same amount of time,
> you'll be mercilessly clobbered every game.
>
> But I suppose that with all these rated AI players in CM9000, the
> program dumbs itself down enough that a 1300-rated AI player does in
> five minutes about what a 1300-rated human player would do in five
> minutes. Is that right?
Not quite. What you'll probably find is that a 1300-rated AI player
does in five seconds what a 1300-rated human player would do in five
minutes and then sits around doing nothing for four minutes and
fifty-five seconds to make it look more even. The effect is the same,
of course. (At least, that's what Fritz does.)
> IOW, I don't need to monkey with the time controls in order to
> balance the game?
Correct.
Dave.
--
David Richerby Frozen Erotic Spoon (TM): it's like
www.chiark.greenend.org.uk/~davidr/ a piece of cutlery but it's genuinely
erotic and frozen in a block of ice!
|
| |
Date: 28 Sep 2006 13:18:25
From:
Subject: Re: Any tips for computer chess?
|
Patrick wrote:
> I do all my chess playing against a computer AI.
I prefer playing humans. A simple "Hello, good luck!", "Good game.",
and a chat about the game afterwards makes it more entertaining. To
each their own.
> Just recently I bought Chessmaster 9000, because I heard that it has a
> "chess academy" that's about as good as CM 10th. Indeed, I've been
A wise choice.
> I loaded a "famous game" from the CM library and read through it, but
> the annotations are so sparse that I didn't learn much from it.
> guess I should try having CM analyze the game first, after which there
> should be annotations for more of the moves.
Watching grandmaster games is entertaining and instructive, but a
computer should not be your priy source for annotations. Computers
are tactical by nature--they easily spot hanging pieces and knight
forks, which decide many beginner-level games.
However, they are clueless at evaluating positional characteristics
such as a lead in space or development, pawn structures, etc. If you
want good annotated games, I recommend Chernev's Logical Chess: Move by
Move or Nunn's Understanding Chess: Move by Move.
Reading through Chernin-Van Der Wiel, 1997 was funny. Nunn tells me
it's completely won, the computer says it's even! A few moves later,
the computer comes to its senses.
> always go back to 2D.
On the computer, I also prefer a simple 2D board.
> But I'm wondering what more I can do to get the most out of CM 9000
Well, it offers a variety of computerized opponents and many tutorials
/ quizzes. What else are you looking for?
> Another big question right now is time controls. In tournament mode
> (for rated games), it doesn't seem possible to have separate time
> controls. I don't want to sit and wait a long time for the computer to
> make its move, but I don't want to be under time pressure myself either
> (I blunder enough without having to worry about that).
If you want different (unfair) time controls, you can play in the game
room instead of the tournament room. Those won't affect your rating.
You say you "sit and wait" while the computer makes it move. Imagine
that the computer's time is extra time for you to consider the
position, just like in a real tournament game.
I prefer 15 minutes + 10 second increment. It's a good compromise for
the busy between playing blitz and playing long games.
> --Patrick
|
| |
Date: 28 Sep 2006 18:26:51
From: David Richerby
Subject: Re: Any tips for computer chess?
|
Patrick <[email protected] > wrote:
> Just recently I bought Chessmaster 9000, because I heard that it has
> a "chess academy" that's about as good as CM 10th. Indeed, I've
> been enjoying working my way through the chess problems. I'm also
> impressed with the way I can play a game, then have CM analyze it,
> and go back through, reading the annotations. (Fritz can do that
> too, I guess, but the one time I tried it, it took forever to finish
> the analysis--I ran it overnight--and then the annotations didn't
> seem very helpful anyway.)
You can adjust the amount of time Fritz will take for its analysis
but...
> I loaded a "famous game" from the CM library and read through it,
> but the annotations are so sparse that I didn't learn much from it.
> I guess I should try having CM analyze the game first, after which
> there should be annotations for more of the moves.
... you won't learn chess by using computers to annotate master games.
Computer analysis is only really useful for spotting tactical errors:
this makes them great for analyzing your own games but poor for
analyzing master games, where there are very few tactical mistakes.
If you want to learn from master games (which is an excellent way to
learn), well-annotated books are the way forward. You can always use
the computer instead of a board! It's hard to tell what level you're
at from your post (almost everyone is a novice compared to Chessmaster
and Fritz!) but Dan Heisman's latest article at chesscafe.com should
give you some suggestions as to books that would suit you.
http://www.chesscafe.com/text/heisman68.pdf
(No, I have no clue why their archives are in pdf.) You want the
section `Read as many master collections as possible' but the other
sections (and the rest of his archive) will also be of interest.
> Another big question right now is time controls. In tournament mode
> (for rated games), it doesn't seem possible to have separate time
> controls. I don't want to sit and wait a long time for the computer
> to make its move, but I don't want to be under time pressure myself
> either (I blunder enough without having to worry about that).
In Fritz, you can get asymmetric time controls for blitz games. Now,
blitz games are usually fast but you can tell Fritz that you want to
play a ten-hour `blitz' game if you want: really, it just uses the
term to mean a game with only one time control. So, File
|
|
