| |
Main
Date: 03 Jan 2007 11:44:12
From: Dave (from the UK)
Subject: A useful feature, or just a waste of time ??
|
I've been working on extending the open-source database 'Scid'
http://scid.sourceforge.net/
I've set up a new project on Sourceforge. Someone else wanted to add a
feature to this. I have some serious concerns this is a bit of a toy and
of little practical use. He says it is "very useful" - my own opinion is
a bit different.
Just interested in what others think.
The basic plan is that you play chess against a chess engine of your own
rating. That engine (Phalanx) is designed to play weak, so makes the
occasional blunder.
In addition to the engine you are playing, a second chess engine (crafty
recommended), which is a lot stronger, will analyse the other engines
performance and report to you, whilst you are playing the engine, the
fact the engine has just blundered
To me, this is pointless. It almost encourages you not to think, ss you
will be told if your opponent blunders - you only have to find out where
This is what I wrote with the intention of putting it into the ChessDB
tutorial,
http://chessdb.sourceforge.net/tutorial/t_tool_play_with_coach.php
but later I decided against that, as I felt it was of little practical use
This is what the author of the patch wrote.
http://prolinux.free.fr/scid/
Note, this find blunders made by the chess engine - not you. That is
intensional apparently.
Interested in your views on this. Is this, as he claims a good aid to
teaching, or is it what I think - a total waste of time?
If you want, you can download his modified version of Scid with the patch.
--
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: 04 Jan 2007 08:49:41
From: Anders Thulin
Subject: Re: A useful feature, or just a waste of time ??
|
Dave (from the UK) wrote:
> The basic plan is that you play chess against a chess engine of your own
> rating. That engine (Phalanx) is designed to play weak, so makes the
> occasional blunder.
Doesn't that rather go outside the area of application for Scid?
It may be useful ... but it's not anything that clearly belongs in a chess
database. 'Feeping creaturism' I'd call it.
--
Anders Thulin ath*algonet.se http://www.algonet.se/~ath
|
| | |
Date: 04 Jan 2007 11:04:36
From: pascal
Subject: Re: A useful feature, or just a waste of time ??
|
Anders Thulin a �crit :
> Dave (from the UK) wrote:
>
>> The basic plan is that you play chess against a chess engine of your own
>> rating. That engine (Phalanx) is designed to play weak, so makes the
>> occasional blunder.
>
> Doesn't that rather go outside the area of application for Scid?
> It may be useful ... but it's not anything that clearly belongs in a chess
> database. 'Feeping creaturism' I'd call it.
You are right : those features are not of a database. On the other hand
Scid already had such features that don't belong to a chess database.
For example :
- launch an analysis engine : you have a "training" option for play
- in the tree window : you also have a "training" option
To add extra features, without breaking previous ones and usability is
sensible and harmless, I think.
Pascal Georges
|
| | | |
Date: 04 Jan 2007 13:37:00
From: David Richerby
Subject: Re: A useful feature, or just a waste of time ??
|
pascal <[email protected] > wrote:
> To add extra features, without breaking previous ones and usability
> is sensible and harmless, I think.
This extension may or may not be sensible; my personal feeling is that
it isn't sensible as SCID is a database and this just isn't database
functionality.
But adding spurious features is certainly not harmless. It makes the
software bigger than it needs to be and size brings a whole host of
problems. A few extra kb of code isn't really a problem when you're
talking about multi-megabyte databases but increasing the complexity
of the code increases the potential for bugs, especially when new code
is added later. It also makes the program appear to be harder to use:
even though there are plenty of people who know that they don't need
to understand the features they don't use, there are lots of people
who will see a big menu and run away.
Adding spurious features is plain bad software engineering. Why don't
you use your code to patch a user interface like xboard, rather than
putting it into a database? It would be much more appropriate there.
Dave.
--
David Richerby Swiss Permanent Clock (TM): it's like
www.chiark.greenend.org.uk/~davidr/ a clock but it'll be there for ever
and it's made in Switzerland!
|
| |
Date: 03 Jan 2007 14:47:18
From: CeeBee
Subject: Re: A useful feature, or just a waste of time ??
|
"Dave (from the UK)" <[email protected] >
wrote in rec.games.chess.analysis:
> In addition to the engine you are playing, a second chess engine (crafty
> recommended), which is a lot stronger, will analyse the other engines
> performance and report to you, whilst you are playing the engine, the
> fact the engine has just blundered
>
> To me, this is pointless. It almost encourages you not to think, ss you
> will be told if your opponent blunders - you only have to find out where
This idea is quite common in chess tutoring: find the blunder in a certain
position and exploit it with the right move sequence. You could see it as an
in-game training version of that principle.
The ChessBases's Fritz GUI's have such a provision as a training tool;
although it's the engine "blundering" for you itself and indicating the
tactical blunder with a flashing blue light.
Although I'm not familiar how many people actually use this feature in their
training, I prefer a more structured approach of a book or a program like
CT-Art etc., which categorizes the themes and - as far as the books are
concerned - explain the background to you as well.
OTOH one could claim that a weaker player could try to build up a position
in which tactical errors can occur from your opponent, and then your
opponent would make the mistake for you. In that case you could train on -
say - certain openings you know well (or better: well enough) and learn
which kind of tactics come wrong which kind of play and game. Such an
interactive approach is not easily replicated in a book.
--
CeeBee
*** entia non sunt multiplicanda praeter necessitatem ***
|
| |
Date: 03 Jan 2007 14:08:41
From: pascal
Subject: Re: A useful feature, or just a waste of time ??
|
Thanks for taking so much care of my time.
As I already told you, if a player wants to play against an engine
simulating an ELO rate without being disturbed by any indication, just
unckeck the three options in the dialog box (just read the manual
online, I put the sentence in bold just for you).
I had some feedback since my own release of Scid, all positive.
So, don't waste your own time now.
Pascal Georges
Dave (from the UK) a �crit :
> The basic plan is that you play chess against a chess engine of your own
> rating. That engine (Phalanx) is designed to play weak, so makes the
> occasional blunder.
>
> In addition to the engine you are playing, a second chess engine (crafty
> recommended), which is a lot stronger, will analyse the other engines
> performance and report to you, whilst you are playing the engine, the
> fact the engine has just blundered
>
> To me, this is pointless. It almost encourages you not to think, ss you
> will be told if your opponent blunders - you only have to find out where
>
> This is what I wrote with the intention of putting it into the ChessDB
> tutorial,
>
> http://chessdb.sourceforge.net/tutorial/t_tool_play_with_coach.php
>
> but later I decided against that, as I felt it was of little practical use
>
> Interested in your views on this. Is this, as he claims a good aid to
> teaching, or is it what I think - a total waste of time?
>
> If you want, you can download his modified version of Scid with the patch.
|
| | |
Date: 03 Jan 2007 13:28:22
From: Dave (from the UK)
Subject: Re: A useful feature, or just a waste of time ??
|
pascal wrote:
> Thanks for taking so much care of my time.
>
> As I already told you, if a player wants to play against an engine
> simulating an ELO rate without being disturbed by any indication, just
> unckeck the three options in the dialog box (just read the manual
> online, I put the sentence in bold just for you).
>
> I had some feedback since my own release of Scid, all positive.
>
> So, don't waste your own time now.
>
> Pascal Georges
I already told you there is a point in having weaker engines, so the
fact you can configure it it useful. But I'm interested in the views of
others on your coach mode.
I realise you can switch off your coaching features, but I'm wondering
if they are any use at all really, or are they just a novelty.
Scid was never designed to *play* chess in. Even the most basic things
are missing.
1) No time controls, other than limit the engine to x seconds per move.
2) No clocks.
3) Can't offer a draw.
4) Can't see draw offers from the engine.
I can't help but feel if Scid or ChessDB are used for playing, those
issues are a far higher priority than coaching modes that display in red
if an engine makes a blunder.
Perhaps you can explain the *practical* benefit of knowing when the
engine has blundered during the game. I can see the point of knowing
later, but not during the game.
I like the ability to control the strength with a slider, but unless you
have anything to back up the numbers, I think a simple 0-100 would be
more accurate. Having 1200-2200 is confusing, if those ratings are far
from accurate.
--
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: 04 Jan 2007 00:55:15
From: Johnny T
Subject: Re: A useful feature, or just a waste of time ??
|
Dave (from the UK) wrote:
> Scid was never designed to *play* chess in. Even the most basic things
> are missing.
>
> 1) No time controls, other than limit the engine to x seconds per move.
> 2) No clocks.
> 3) Can't offer a draw.
> 4) Can't see draw offers from the engine.
>
> I can't help but feel if Scid or ChessDB are used for playing, those
> issues are a far higher priority than coaching modes that display in red
> if an engine makes a blunder.
Scid may never be the product for this. And that is a valid point.
There are a suite of products that are necessary for the chess player,
and there is an inherent push to put functionality into one product, if
for no other reason than you momentarily have the ear of the developer.
Especially if you do not have the talent or inclination to do it yourself.
> Perhaps you can explain the *practical* benefit of knowing when the
> engine has blundered during the game. I can see the point of knowing
> later, but not during the game.
But here is the practical point. It is difficult to play against an
engine, especially for a non-world class player. It is difficult for
the engine to fail appropriately, it is difficult for the human to get
up for it at the right times. (The engine in schizoid mode plays
partially like the beast that it is, sometimes as something that barely
makes legal moves).
What is nice about the blunder indicator, is that it is like a puzzle
generator for the player. That if I care about *this* move, I can look
deeper and find the solution. (and go back and see if I saw the same
thing over the board as the watcher saw).
This doesn't mimic real life, or anything else at all. So what? It
doesn't make it a novelty. It is yet another thing that makes my
study/gameplay/computer interactivity fun, instructive, interesting.
Maybe it makes me a better player. Maybe I use the program more. Maybe
it is fun in and of itself. All of these things seem reasonable.
Whether or not it belongs in Scid? Maybe not. But who cares if it is in
there or not? If all of the surroundings are there, and it could be
added, relatively cheaply, it seems to be an interesting thing to add.
It may be something that is added here because it could, then eventually
forked to its own thing.
As a non-world class player, I think the approach is an interesting one
in making computers enjoyable to play against. I would definitely try
it out. Whether it held my interest over time. I don't know. But it
might.
|
| | | | |
Date: 04 Jan 2007 11:10:11
From: pascal
Subject: Re: A useful feature, or just a waste of time ??
|
Johnny T a �crit :
> This doesn't mimic real life, or anything else at all. So what? It
> doesn't make it a novelty. It is yet another thing that makes my
> study/gameplay/computer interactivity fun, instructive, interesting.
> Maybe it makes me a better player. Maybe I use the program more. Maybe
> it is fun in and of itself. All of these things seem reasonable.
You got the motivations :-)
> Whether or not it belongs in Scid? Maybe not. But who cares if it is in
> there or not? If all of the surroundings are there, and it could be
> added, relatively cheaply, it seems to be an interesting thing to add.
> It may be something that is added here because it could, then eventually
> forked to its own thing.
Let me precise things : this is already implemented and available here :
http://prolinux.free.fr/scid/
I now plan to add after this "play tactical games" feature an "opening
trainer". This will require some work !
Pascal Georges
|
|
