"Chris Mattern" wrote in message
...
In article p4eeh.6029$sM2.4804@trndny05, Chess One wrote:

"Chris Mattern" wrote in message
...

...

level chess by simply defining how they do so as somehow "not fair".

That's not true. Its 'not legal' which is the term that causes offense.

What is "not legal"?

Looking up moves and making notes.

The computer is not using any external source to look things up
or make notes.

This gets a bit boring, really! After answering the question, a new term is
introduced. Whether the look-up is internal - or external - and indeed the
db its looks up could be physically remote and 'wireless', is //beside the
point// that it looks-up.

This is an axiom - if the computer really had memory it would not need to
look anything up!

Just because everything programmed into a computer can be called data, is
simply to escape into an abstract collective pronoun, since everything in
the universe can be described as data!

We are discussing an aspect of consulting particular kinds of data.

snip

What else is there? The human has a memory, the computer has a memory.

I'm sorry - this is not to understand anything, and is nothing to do with
computer or people, but an understanding of what memory is. It is strange
that this misperception is insisted upon by programmers, who apparently
cannot achieve the logic leap of diffentiation between reading a book and
remembering what you read in a book.

The difference is that the book is external to you. The remembrance is
internal to you. If there's some other difference, you're going to have
to explain it to me.

I don't think so! I don't think I have to explain again that if a player
drags his entire library into the tournament room and looks up his moves
that this is illegal activity.

Its not memory when a person or anyone else consults data in this way. So,
please stop using the term 'memory' as such an obviously false analogy.

Phil Innes



Therefore it is not cheating if the computer
remembers something using its internal resources.

If I consult an opening book during play, is that memory?

It insults my or your intelligence by answering that question, 'yes'.

No, it doesn't. It's memory, it's just not *your* memory.

Until we define some basic terms, and provide some logical operators, we
really can't use words which have /analogous/ meanings, since if we
replaced
the word 'memory' by a phrase, it would render logical nonsense.

Memory is the storage of information. Chess matches allow you to use
*your* memory, but not someone else's.

Programmers should remember that most computer terms are anthropomorphic,
and only analogous, and not confuse the map for the territory.

Otherwise it is argued that when I consult something in a book I never
read
before then this is an act of memory. Whereas I say it is precisely not

Yes, it is an act of memory. It's the *book's* memory, however, not
yours.

memory, since otherwise I wouldn't need to consult it.

If the book did not contain, or remember, the information you needed,
there
would be no point in consulting it.

snip


You seem a little obsessed by this term "legal". I still do not
understand
why a computer with a memory is not "legal" but one without one is.

Obsessed or insistent?

You can have your 'memory' and your Ferrari too. Who cares about the
rules,
winning is all!

Memory is the storage of information. Chess rules allow you to use
your memory, but not another's. I still do not see why people are
allowed their memory but computers should not be.

--
Christopher Mattern

"Which one you figure tracked us?"
"The ugly one, sir."
"...Could you be more specific?"

In article uiVeh.498$it5.361@trndny06, Chess One wrote:

"Chris Mattern" wrote in message
...
In article p4eeh.6029$sM2.4804@trndny05, Chess One wrote:

"Chris Mattern" wrote in message
...

...

level chess by simply defining how they do so as somehow "not fair".

That's not true. Its 'not legal' which is the term that causes offense.

What is "not legal"?

Looking up moves and making notes.

The computer is not using any external source to look things up
or make notes.

This gets a bit boring, really! After answering the question, a new term is
introduced. Whether the look-up is internal - or external - and indeed the
db its looks up could be physically remote and 'wireless', is //beside the
point// that it looks-up.

No, the fact that it "looks up" cannot be the point, because a human
looks things up in his own memory and that is not disallowed. If you
are to going to define looking things up as using a mechanical device
and remembering things using a brain as not looking things up, and
then go on to say that looking things up is not allowed, then computers
cannot be allowed to play chess as all, as they must "look things up"
to function. If that is your position, well and good. It is consistent
and logical, and I will admit that it actually eliminates the need for
judgment calls as to what is or is not external information storage.
But I would see computers playing chess; I find it both interesting and
useful. Therefore I choose my definition of "looking up" or "reference
materials" to be information storage external to the subject, and
accept the difficulties and consequences that come with it. That is
the central problem, as I see it. You can have a consistent set of
definitions and rules that allows the computer to store and retrieve
lines of play as part of its operation. You can have a consistent set
of definitions and rules that does not allow the computer to play. I
cannot see a consistent set of definitions and rules that allows the
computer to play but does not allow it store and retrieve lines of play.


This is an axiom - if the computer really had memory it would not need to
look anything up!

So you are defining "memory" as "human memory". A computer using a
hard disk is not "memory" but "looking things up". Very well. The
logical consequence of this position is that a computer cannot be
allowed to play chess at all, as it cannot run without "looking things
up" on its hard drive.

Just because everything programmed into a computer can be called data, is
simply to escape into an abstract collective pronoun, since everything in
the universe can be described as data!

If I am understanding you here, you are complaining that I am refusing
to differentiate between what type of data the computer is storing in
determining what it is allowed to store. Just so. The human brain is
theoretically capable of storing any type of data, and we allow the
chess player to bring to the table any data he can store in his head.
He is not allowed to use any materials to consult with, regardless
of what information those materials contain. Therefore we cannot
arbitrate the data the computer is allowed to store on the basis of
what the data represents, because we do not limit human players so.
The decision must be made on the basis of where and how the information
is stored.

We are discussing an aspect of consulting particular kinds of data.

No, we aren't, or at least I'm not. What data a human player is
allowed to access in a chess match does not depend in slightest
on what kind of data it is, only where and how it is stored.
Therefore, the discussion needs to be about aspects of consulting
*any* kind of data.

snip

What else is there? The human has a memory, the computer has a memory.

I'm sorry - this is not to understand anything, and is nothing to do with
computer or people, but an understanding of what memory is. It is strange
that this misperception is insisted upon by programmers, who apparently
cannot achieve the logic leap of diffentiation between reading a book and
remembering what you read in a book.

The difference is that the book is external to you. The remembrance is
internal to you. If there's some other difference, you're going to have
to explain it to me.

I don't think so! I don't think I have to explain again that if a player
drags his entire library into the tournament room and looks up his moves
that this is illegal activity.

Yes, it is. Because, as I said, the library is external to the player.

Its not memory when a person or anyone else consults data in this way. So,
please stop using the term 'memory' as such an obviously false analogy.

So you maintain the position that a mechanical data storage device is not
"memory" and therefore cannot be allowed to be used in a chess match.
If that is your position, you need to accept that the consequence of this
view is that computers cannot participate in chess matches. And given that
is your position, then we will simply have to agree to disagree. Any
logical position must start with defintions and axioms and then reason
to the conclusion. I would see computers play in chess matches and have
chosen the basis that would allow that. I admit that freely. Other
viewpoints are possible, and I can respect that. I simply cannot see
a consistent viewpoint that allows computers to play in matches but does
not allow them to have a reference of opening lines.


--
Christopher Mattern

"Which one you figure tracked us?"
"The ugly one, sir."
"...Could you be more specific?"

"Martin Brown" wrote in message
ups.com...

Ange1o DePa1ma wrote:

"Chess One" wrote

I would only wish that better means would be deployed for evaluation
purposes, but the weakest part of the engine is glossed by the opening
book, so even observing those evaluation weaknesses is difficult, if
not
impossible. I think that no-one has wanted to grasp the nettle, and
take a
hard look at evolving a legal chess engine, since it would immediately
drop the program's rating to less than commercial viability.

Less than 100 ELO points difference if the Hydra benchmarks on the
University of Paderborne CompSci site are to be believed.

http://wwwcs.uni-paderborn.de/cs/flu...n_springer.pdf

Interestingly, I found this comment on page 3 or 4: "In another game, Junior
played against Shredder, and Shredder stayed in the opening database until
the final draw!"
As for the value of the book, I read through all 10 pages and didn't see
references to non-book play, especially not with human opponents. Instead
there was a contrary statement about current efforts and a strong
concentration on optimising use of the book by external statistical
approaches, and move options assigned an ELO.

That is only somewhat like the research conducted by quantitative analysis
by David Levy, Kevin O'Conne;; and David Watt [1981 Imprint Capablanca] and
I hvae two books from that series. Instead of Elo-rated results, Levy chose
percentage played and percentage WDL for each move.

But evidence of computer v human testing with book=off is scarce, or even
completely absent.

Cordially, Phil Innes


Chess engines are presently gaining strength at about 30 ELO points per
year. So using the opening book represents an increase in raw engine
strength equivalent to about 3 years of additional development work.

Has anyone studied:

1. how much weaker an engine would play with no opening book? With a
limited
opening book?

I stumbled upon this article about opening books and learning
algorithms and their influence on match outcomes. The summary results
we

Hydra 1.09 (no book) vs Shredder8 (default book) : +27, -25, =20 ELO
+9
Hydra 1.09 (autobook) vs Shredder8(default book): +37, -20, =15 ELO
+83
Hydra 1.09 (autobook) vs Shredder8(learning book): +33, -21, =18 ELO
+58

A human that specialised in anti-computer play might make a bit more of
a dint on an engine without an opening book. But it is also a double
edged sword, the computer can potentially take the game into territory
that is not well explored in the theory books.

Hydra appears to have a design that plays somewhat anti-human chess.

I'm also baffled, 20+ years into commercial computer chess, why engine
chess
ratings are not more closely harmonized with ELO ratings. With all the
titled players in the world, especially in Europe, it should be
relatively
easy to obtain a very large database of human vs. computer games.

I think the problem is that the profusion of engine vs engine matches
tends to dominate since they do not generally require paying to play
chess. Also it is becoming clear that the souped up private openings
books for top engines in these matches are increasingly sophisticated
to the extent that they are a sort of proxy match between the human
book authors setting traps for rival engines. Notably centred on
castling bonus and other quirks.

Regards,
Martin Brown

Martin Brown wrote:

Robt Hyatt once told me that Crafty would never play the Ruy since 3. Bb5
looks like a very poor move.

Ruy Lopez wasn't all that convinced by his own named opening either.

Maybe this opening is inferior, though it fooled many for
centuries. "You can fool some of the people, all of the time."

I don't doubt your word that Dr Hyatt said that. But it seems he was
wrong. Crafty 19.01 with 256MB cache sometimes gets 3. Bb5 as the
preferred line rather than 3. Bc4 in searches to plys around 16 (after
about 12 hours on a 3GHz machine). Currently at ply 17 it has 3. Bc4
0.09 and 3. Bb5 0.01 (at ply 16 they were 0.01 and 0.1 - ie roughly
swapped around).

And if you allow for say 5 centipawns of Gaussian noise dither

A clever improvement! Now Rybka can think without being
bothered by spectator noise, and the grinding of teeth which
begins anew after each crushing move. Does anyone sell
tooth protectors for this?

on the
raw evaluation function it might still sometimes play the nominaly
"inferior" line.

Engines like Shredder with better pawn structure and positional
awareness get 3. Bb5 top for a fair proportion of the time (and at
depth it prefers Bb5 for plys 19 through 22, with 23 reverting to Bc4,
and 24 still running). At ply 20 [ 23 ] the top 5 ranking moves a

3. Bb5 0.42 [ 0.24 ]
3. d4 0.31 [ 0.19 ]
3. Bc4 0.30 [ 0.31 ]
3. Nc3 0.23 [ 0.17 ]
3. Be2 0.23 [ 0.03 ]
(everything else negative)

Wow. I had no idea chess programs were already up so high
in search depth. I can remember back to when 8 or 9 plys
were considered impressive.

But I think it is a very poor idea to try and use what humans
merely *believe* is a superior move as the test for computers.
Better to come up with something purely objective, like say,
checkmate or snatching a free Queen.

The fact(?) that humans have not *yet* found a satisfactory
defense in the Ruy Lopez in no way proves that one does not
exist, nonetheless. A long time ago, after 1.P-K4 the response
....c5 was scoffed at as unorthodox (at best); now the move has
become acceptable -- even preferred.

After my next computer upgrade, I may buy Rybka and have
a look at its best line of play to see *why* it doesn't like 3.B-b5
much. Maybe there is a perfectly good reason. Just guessing,
I would say the programmers taught the program that Bishops
are better than Knights, and if the exchange is therefore scored
as a bad thing, the Bishop is a target on that square, and it is
possible that the scoring of pawn weakness which ensues
after ...a6 and ...b5 is insufficient to fully offset other factors.

In fact, it seems likely that IM Innes' complaint relates more
toward the programmers than it does toward computers
themselves. It is the programmers who have, indirectly, made
it unlikely for their machines to choose IM Innes' pet move.

-- help bot