Bob, can we have a substantive answer please ... ?
Steve Blatchford
Hi Bob,
Could we please have a rather more substantive answer to the intelligently
argued case put forward by Angel (copied below) ?
She is suggesting all the programmers and program ideas are dead-end street,
no ?
Steve
-----Original Message-----
From: Robert Hyatt <hy...@crafty.cis.uab.edu>
Newsgroups: rec.games.chess.computer
Date: 10 August, 1998 03:48
Subject: Re: Dr. Hyatt's Stubborn refusal to Admit the Strength of Chess
Computers
>AngelEvan7 <angel...@aol.com> wrote:
>
>: Hello, this is Angel.
>
>: Round and round we go with the CURIOUS MALE LOGIC.
>
>: Doh ! How to explain to the IDIOT PRETENDER ?
>
>: He we go, PMT and Sunshine Primrose Oil at the ready:
>
>: Chess strength is not on a continuum. It is stepped. Roughly 200 grading
points
>: per step.
>
>: You, Prof Hyatt, are on a low step. The same step as most of the other
>: pretend-knowledge-fools. You can tell who is on this step, becos they
say:
>: "positional is just long-range tactics". That is they think chess
strength is a
>: continuum. Faster, deeper, faster, deeper, faster, deeper, more, more,
gasp,
>: gasp, but Deep Blue came too soon. Leaving a great disatisfaction, not
>: alleviated by a good smoke.
>
>: Back to male logic: each step is a WAY OF THINKING. A new way of thinking
over
>: the step below. So each step is quite a leap. Each step requires a
particular
>: kind of brain/thought, and, not just that, but a particular kind of
personality
>: to.
>
>: Just like you need a particular kind of brain to know that 4565233983745
>: divided by 347623847 is 394853 (or whatever), calculated in 2 seconds
flat.
>: there are these kind of people. Usually hopeless at everything else. And
NOT
>: FEMALE.
>
>: Learn new language here: PARADIGM SHIFT. But u kno it alred.
>
>: Each 200 step needs paradigm-shift. Very difficult, not poss for many
(wrong
>: brain/personality). Throw away old, take on new. And take it on means
DO/LIVE
>: it. All very well to know positional concepts like 2 bishops; but another
thing
>: altogether to PLAY POSITIONAL. But u no kno this, you on lower step.
>
>: REF: Kotov Think Like a Grandmaster. You gotta lern to think difrent.
Gettit ?
>
>: U can kno stuff about step or two above your own level intellectually,
but not
>: in practice, not emotionally, not intuitively.
>
>: To know what a GM knows and your program doesn't (eg a whole massive
lot), and
>: to be able to explain it to someone else, and to be able to quantify it
>: somehow, and algorithmise it requires you to be there yourself. And you
are
>: not. You are way way off.
>
>: So stop pretending.
>
>: You wanna carry on ? Or give up now ? I got lots more.
>
>: Angel (no men allowed)
>
>
>
>Hi Rolf, glad you are back. Now if you would only post about computer
>chess, all would be well...
>
>:)
>
>
>--
>Robert Hyatt Computer and Information Sciences
>hy...@cis.uab.edu University of Alabama at Birmingham
>(205) 934-2213 115A Campbell Hall, UAB Station
>(205) 934-5473 FAX Birmingham, AL 35294-1170
Christopher R. Dorr
hardly intelligently argued; perhaps you were being sarcastic (I hope you
were). This argument is:
A. A blatant troll by the infamous Evans person (1 person, little Sean [or
perhaps Rolfie?!], for those who actually believe it is more than 1 person
posting)
B. Incoherent - grammatically and semantically meaningless
C. Filled by meaningless babble (Paradigm shift? Please...I thought only
management consultants prattled on about this).
On a related note, I can't believe the triple-digit-IQ crowd in this
newsgroup is spending time, effort, and bandwidth responding to this fool.
Don't you guys have a kill file?
Chris
Steve Blatchford wrote in message <6qph7u$ao2$1...@eros.clara.net>...
Robert Hyatt
: Hi Bob,
: Could we please have a rather more substantive answer to the intelligently
: argued case put forward by Angel (copied below) ?
: She is suggesting all the programmers and program ideas are dead-end street,
: no ?
: Steve
The below crapola is simply idle ramblings by someone that many of us
know by name and/or reputation. There is nothing new in computer chess
programs, just steady improvement. Don't put much stock in the stuff
below *until* you see someone produce a reasonable program based on
something "new".
However, when it comes to "trolls" like the junk below, I try to avoid
the stuff completely. If you are interested in where it comes from, check
the headers, and "think"... and at least one or two answers will instantly
jump to the front of the line. I "think" I know who is writing this stuff,
but I won't say anything until I am sure.
As far as a "dead-end street", "she" holds the license on that, no doubt.
:)
:>: Round and round we go with the CURIOUS MALE LOGIC.
:>
:>: Doh ! How to explain to the IDIOT PRETENDER ?
First, you'd have to be smarter than the "pretender" so this case is
over already... :)
:>
:>: Chess strength is not on a continuum. It is stepped. Roughly 200 grading
: points
:>: per step.
:>
It is *not* stepped, and this shows *total* ignorance of chess. Just
because rating classes within the USCF are broken into 200 point classes
says *nothing* about the game of chess. This is drivel.
:>: You, Prof Hyatt, are on a low step. The same step as most of the other
:>: pretend-knowledge-fools. You can tell who is on this step, becos they
: say:
:>: "positional is just long-range tactics". That is they think chess
: strength is a
:>: continuum. Faster, deeper, faster, deeper, faster, deeper, more, more,
: gasp,
:>: gasp, but Deep Blue came too soon. Leaving a great disatisfaction, not
:>: alleviated by a good smoke.
:>
:>: Back to male logic: each step is a WAY OF THINKING. A new way of thinking
: over
:>: the step below. So each step is quite a leap. Each step requires a
: particular
:>: kind of brain/thought, and, not just that, but a particular kind of
: personality
:>: to.
Again, drivel, not supported by *any* GM or IM player that I can think of.
Perhaps by someone that had partaken of a few too many illegal substances,
but not real chess players... :)
:>
:>: Just like you need a particular kind of brain to know that 4565233983745
:>: divided by 347623847 is 394853 (or whatever), calculated in 2 seconds
: flat.
:>: there are these kind of people. Usually hopeless at everything else. And
: NOT
:>: FEMALE.
:>
:>: Learn new language here: PARADIGM SHIFT. But u kno it alred.
:>
:>: Each 200 step needs paradigm-shift. Very difficult, not poss for many
: (wrong
:>: brain/personality). Throw away old, take on new. And take it on means
: DO/LIVE
:>: it. All very well to know positional concepts like 2 bishops; but another
: thing
:>: altogether to PLAY POSITIONAL. But u no kno this, you on lower step.
:>
:>: REF: Kotov Think Like a Grandmaster. You gotta lern to think difrent.
: Gettit ?
Nope... Otherwise I guess I have gone through at least 8 "paradigm shifts"
since my first program was under 1500, my current one is over 2300 (USCF),
at least. And my current program is not that different than my under 1500
program was, same search ideas, same sort of evaluation ideas (but more of
them with faster hardware of course) and so forth. As I said, this is really
more of an idle "troll" than a statement of anything that means anything...
which is typical of this poster most of the time... and the evidence on the
ID of this person is "mounting".. :)
:>
:>: U can kno stuff about step or two above your own level intellectually,
: but not
:>: in practice, not emotionally, not intuitively.
:>
:>: To know what a GM knows and your program doesn't (eg a whole massive
: lot), and
:>: to be able to explain it to someone else, and to be able to quantify it
:>: somehow, and algorithmise it requires you to be there yourself. And you
: are
:>: not. You are way way off.
:>
:>: So stop pretending.
:>
:>: You wanna carry on ? Or give up now ? I got lots more.
People with diarrhea generally do... :)
Steve Blatchford
Christopher R. Dorr wrote in message <35d05...@news.one.net>...
>How on earth can someone reply intelligently to this load of drivel? It is
>hardly intelligently argued; perhaps you were being sarcastic
No.
>(I hope you
>were). This argument is:
>
>A. A blatant troll by the infamous Evans person (1 person, little Sean [or
>perhaps Rolfie?!], for those who actually believe it is more than 1 person
>posting)
Yet again demonstrating the disastrous set of relationships within this
newgroup and wider.
a) the general assumption that who is saying something is more important
than what is being said.
b) relatedly, the only acceptable viewpoints being those wrapped up in the
'correct' type of wrapping paper.
c) relatedly again, the application by readers of their own choice of
wrapping paper (for condemnation or approval) when there is uncertaintly as
to origin.
d) relatedly again, the idea that every new poster must fall into one of the
various camps; and the rapid depositing of newbies into said camps.
e) the neat and easy solution of having a devil's pit into which to throw
any unacceptable viewpoint. "he's just like Rolf, innit?"
>
>B. Incoherent - grammatically and semantically meaningless
to some :-)
but why look for the real gold when you have the fool's go(l)d already ?
>
>C. Filled by meaningless babble (Paradigm shift? Please...I thought only
>management consultants prattled on about this).
perhaps I'ld better make an attempt at translating the angel-outpourings
into rgcc-acceptable prose ...........
what (s)he is trying to say (mixed up with various feminist and other bits)
is:
1. Chess knowledge/strength of play is stepped, not continuous. Players get
gradually better within each step, but require a great leap to move from
step to step. A great leap in ways of thought, usually requiring a throwing
out of the previous way, each step. Not everybody has the necessary thought
patterns/brain structure/personality to make it on and upwards.
2. It is possible, by reading/talking to strong players, to get some of the
ideas/constructs that strong players have. But that a proper understanding,
at the level (step) of a strong player, can only be achieved by
doing/performing. The difference between intellectual; and intuitive, truly
understanding knowledge. (if this is difficult for you, think of driving a
car. i don't know how to drive by knowing what and where are the pedals and
steering wheel; i only truly know by doing it, over and over until it
becomes second nature. Only after this stage could I dream of going onto the
next steps, being a car driving instructor, or a rally driver, or car
designer or whatever). And that, thus, getting two steps on, requires
mastery of the step below.
3. The crop of practising chess programmers are of low chess strength.
Patzers, says Angel. On the basis that the top grades on icc are 3100+, icc
grades appear inflated over fide grades by 500 points or more. Thus, for
example Hyatt would be 1950-500=1450, Moreland 1650-500=1250 and so on. Most
of the other programmers are similar, they get to 1300,1400,1500,1600 fide,
give up and start programming instead. Constant pattern.
4. Thus the crop of current chess programmers (not just Hyatt) are way way
below GM strength. At 200 elo per step, that's 8 steps give or take. The
quantum at which they don't understand that they don't know that they don't
know is just too large to even think about.
5. To know what a GM knows and these program/programmer combinations don't
(eg a whole massive lot), and to be able to explain it to someone else, and
to be able to quantify it somehow, and *algorithmise it* requires the
programmer to be there himself. And they are not. They are way way off.
6. The programmers know that the programs are lacking knowledge. Big time.
They also know that another program, running 1000x faster, allegedly with a
sophisticated eval function, a team of GMs help, the massive advantage of
free singular extensions (from the massive parallelisation), is arguably not
even as good as a super-gm. They know that they are 10-15-20 years from
reaching equivalent speeds. They know that nobody has any idea how to get
in/quantify/algorithmise the incredibly complex knowledge required to deal
with the problems shown up in Rebel-Anand RBBvRR material inbalance
situation amongst many others. Especially not they with their 1200-1700 fide
knowledge. So they know, in their hearts, that they can't get to fool's
paradise.
7. But this doesn't stop them pretending. Like an army that knows the war is
lost, they still have to forage for food, eat, retreat, continue. Keep that
university post, maintain the chess program company profitability, get that
PhD, sell those programs, make those lists.
>
>
>On a related note, I can't believe the triple-digit-IQ crowd
Where, where, where ? Show me them .... :-)
>in this
>newsgroup is spending time, effort, and bandwidth responding to this fool.
Who spends the time, effort and bandwidth chasing fool's go(l)d ? :-)
And who is the fool ? :-)
>Don't you guys have a kill file?
Yeah they do. It's called either the killfile, or the censored forum, or
shouting down, or rudeness or denial. The intelligent ones just quit, got
kicked out for being a nuisance, or ignored on the grounds they were Rolf.
Steve
>
>Chris
>
>
>Steve Blatchford wrote in message <6qph7u$ao2$1...@eros.clara.net>...
>>
>>
>>Hi Bob,
>>
>>Could we please have a rather more substantive answer to the intelligently
>>argued case put forward by Angel (copied below) ?
>>
>>She is suggesting all the programmers and program ideas are dead-end
>street,
>>no ?
>>
>>Steve
>>
>>
>>-----Original Message-----
>>From: Robert Hyatt <hy...@crafty.cis.uab.edu>
>>Newsgroups: rec.games.chess.computer
>>Date: 10 August, 1998 03:48
>>Subject: Re: Dr. Hyatt's Stubborn refusal to Admit the Strength of Chess
>>Computers
>>
>>
>>>AngelEvan7 <angel...@aol.com> wrote:
>>>
>>>: Hello, this is Angel.
>>>
>>>: Round and round we go with the CURIOUS MALE LOGIC.
>>>
>>>: Doh ! How to explain to the IDIOT PRETENDER ?
>>>
>>>: He we go, PMT and Sunshine Primrose Oil at the ready:
>>>
>>>: Chess strength is not on a continuum. It is stepped. Roughly 200
grading
>>points
>>>: per step.
>>>
>>>: You, Prof Hyatt, are on a low step. The same step as most of the other
>>>: pretend-knowledge-fools. You can tell who is on this step, becos they
>>say:
>>>: "positional is just long-range tactics". That is they think chess
>>strength is a
>>>: continuum. Faster, deeper, faster, deeper, faster, deeper, more, more,
>>gasp,
>>>: gasp, but Deep Blue came too soon. Leaving a great disatisfaction, not
>>>: alleviated by a good smoke.
>>>
>>>: Back to male logic: each step is a WAY OF THINKING. A new way of
>thinking
>>over
>>>: the step below. So each step is quite a leap. Each step requires a
>>particular
>>>: kind of brain/thought, and, not just that, but a particular kind of
>>personality
>>>: to.
>>>
>>>: Just like you need a particular kind of brain to know that
4565233983745
>>>: divided by 347623847 is 394853 (or whatever), calculated in 2 seconds
>>flat.
>>>: there are these kind of people. Usually hopeless at everything else.
And
>>NOT
>>>: FEMALE.
>>>
>>>: Learn new language here: PARADIGM SHIFT. But u kno it alred.
>>>
>>>: Each 200 step needs paradigm-shift. Very difficult, not poss for many
>>(wrong
>>>: brain/personality). Throw away old, take on new. And take it on means
>>DO/LIVE
>>>: it. All very well to know positional concepts like 2 bishops; but
>another
>>thing
>>>: altogether to PLAY POSITIONAL. But u no kno this, you on lower step.
>>>
>>>: REF: Kotov Think Like a Grandmaster. You gotta lern to think difrent.
>>Gettit ?
>>>
>>>: U can kno stuff about step or two above your own level intellectually,
>>but not
>>>: in practice, not emotionally, not intuitively.
>>>
>>>: To know what a GM knows and your program doesn't (eg a whole massive
>>lot), and
>>>: to be able to explain it to someone else, and to be able to quantify it
>>>: somehow, and algorithmise it requires you to be there yourself. And you
>>are
>>>: not. You are way way off.
>>>
>>>: So stop pretending.
>>>
>>>: You wanna carry on ? Or give up now ? I got lots more.
>>>
>>>: Angel (no men allowed)
>>>
>>>
>>>
>>>Hi Rolf, glad you are back. Now if you would only post about computer
>>>chess, all would be well...
>>>
>>>:)
>>>
>>>
>>>--
Robert Hyatt
: perhaps I'ld better make an attempt at translating the angel-outpourings
: into rgcc-acceptable prose ...........
: what (s)he is trying to say (mixed up with various feminist and other bits)
: is:
: 1. Chess knowledge/strength of play is stepped, not continuous. Players get
: gradually better within each step, but require a great leap to move from
: step to step. A great leap in ways of thought, usually requiring a throwing
: out of the previous way, each step. Not everybody has the necessary thought
: patterns/brain structure/personality to make it on and upwards.
As I said previously, I don't buy "stepping"... because what do you do
with someone that is 2199 and someone that is 2201, or someone that is
2399 and someone that is 2400. There is no "stepping". Advancing 200
points certainly requires a lot of new acquired knowledge and experience,
but the overall rating system is quite continuous, which can bee seen if
you study how Elo's rating system actually works...
: 2. It is possible, by reading/talking to strong players, to get some of the
: ideas/constructs that strong players have. But that a proper understanding,
: at the level (step) of a strong player, can only be achieved by
: doing/performing. The difference between intellectual; and intuitive, truly
: understanding knowledge. (if this is difficult for you, think of driving a
: car. i don't know how to drive by knowing what and where are the pedals and
: steering wheel; i only truly know by doing it, over and over until it
: becomes second nature. Only after this stage could I dream of going onto the
: next steps, being a car driving instructor, or a rally driver, or car
: designer or whatever). And that, thus, getting two steps on, requires
: mastery of the step below.
: 3. The crop of practising chess programmers are of low chess strength.
: Patzers, says Angel. On the basis that the top grades on icc are 3100+, icc
: grades appear inflated over fide grades by 500 points or more. Thus, for
: example Hyatt would be 1950-500=1450, Moreland 1650-500=1250 and so on. Most
: of the other programmers are similar, they get to 1300,1400,1500,1600 fide,
: give up and start programming instead. Constant pattern.
This is erroneous... The very upper reaches of ICC are certainly over-
inflated... but once you drop down to the 2200-2400 range, things are
actually pretty sane... IE most IM players are in the 2300-2400 range,
with a few notable exceptions... Most GM players are in the 2500-2600
range with a very few notable exceptions...
I doubt I'm a "1400" player, for example, after having played for 40 years
and worked on a computer chess program for 30 years... There are many
others, such as Vincent Diepeveen (a legit IM), Marth Hirsch, a very strong
chess player, Dave Kittinger (ditto), Dave Slate (>2000 USCF) and so forth
and so on...
: 4. Thus the crop of current chess programmers (not just Hyatt) are way way
: below GM strength. At 200 elo per step, that's 8 steps give or take. The
: quantum at which they don't understand that they don't know that they don't
: know is just too large to even think about.
This I know is incorrect... And I can suggest a couple of GM players you
might talk to about "my" positional understanding of the game. I'm certainly
tactically rusty as the devil (if you look at my "1900 rating and the games I
have played, you can see what I mean here) but I'm hardly stupid to the ways
of positional chess... and when I don't understand something I certainly know
who to ask for clarification...
The difference between a 2200 and a 2500 is not *just* knowledge, it is
*experience* as well. The GM has seen so many positions already that he can
use this to guide his play right along with his positional understanding. I'd
not be surprised to find someone that had read about chess for years that
understood just as much as a GM about pawn structure and the like, but without
the experience base of that GM, he'd probably get killed, because the GM would
understand the dos and don'ts, but he'd also understand the *exceptions* due to
experience.
: 5. To know what a GM knows and these program/programmer combinations don't
: (eg a whole massive lot), and to be able to explain it to someone else, and
: to be able to quantify it somehow, and *algorithmise it* requires the
: programmer to be there himself. And they are not. They are way way off.
Maybe or maybe not. It might be possible that a program that is good enough
in some areas, can compete with a GM and simply "conceal" the things it doesn't
do very well by avoiding the positions where these "things" are likely to be
encountered. IE if we are having multiple physical-endurance type contests,
and I can't swim very well, I just never go near water. And if you can't
somehow drag me to it, or trick me into going near it on my own, you won't
ever see that weakness.
Whether this can happen or not I don't know. But current programs and a growing
number of evidenciary games suggest that we are getting away with this more than
I would have thought. And until the GM players discover that if they drag us
to the ocean they can drown us, they are going to have a hard row to hoe. And
many are not yet interested in isolating and exploiting weaknesses, they want
to play as they have always played before. And that can lead to dire consequences
against computers, if the GM doesn't exercise caution.
: 6. The programmers know that the programs are lacking knowledge. Big time.
: They also know that another program, running 1000x faster, allegedly with a
: sophisticated eval function, a team of GMs help, the massive advantage of
: free singular extensions (from the massive parallelisation), is arguably not
: even as good as a super-gm. They know that they are 10-15-20 years from
: reaching equivalent speeds. They know that nobody has any idea how to get
: in/quantify/algorithmise the incredibly complex knowledge required to deal
: with the problems shown up in Rebel-Anand RBBvRR material inbalance
: situation amongst many others. Especially not they with their 1200-1700 fide
: knowledge. So they know, in their hearts, that they can't get to fool's
: paradise.
Some of us "know"... but that (fortunately) doesn't prevent us from trying.
The "Wrights" proved that sometimes trying is enough, because enough monkeys in
a room full of typewriters will definitely produce poetry, given enough time. :)
: 7. But this doesn't stop them pretending. Like an army that knows the war is
: lost, they still have to forage for food, eat, retreat, continue. Keep that
: university post, maintain the chess program company profitability, get that
: PhD, sell those programs, make those lists.
Pretending is a marketing ploy. At least in my case, it doesn't fit, because
I have been quite honest and objective in my assessment of computer chess
programs over the years. But they are *definitely* stronger now than 10 years
ago. They will *definitely* be stronger in 10 years than they are now. Yet
the human is not evolving significantly, at least on the same time scale, so
the end may not quite be in sight, but it is near, nonetheless...
:>
:>
:>On a related note, I can't believe the triple-digit-IQ crowd
: Where, where, where ? Show me them .... :-)
They are around... they are just drowned out by a couple of single-digit
types... :)
:>in this
Christopher R. Dorr
Steve Blatchford wrote in message <6qpn5l$bj5$1...@eros.clara.net>...
>
>Christopher R. Dorr wrote in message <35d05...@news.one.net>...
>>How on earth can someone reply intelligently to this load of drivel? It is
>>hardly intelligently argued; perhaps you were being sarcastic
>
>No.
>
>>(I hope you
>>were). This argument is:
>>
>>A. A blatant troll by the infamous Evans person (1 person, little Sean [or
>>perhaps Rolfie?!], for those who actually believe it is more than 1
person
>>posting)
>
>Yet again demonstrating the disastrous set of relationships within this
>newgroup and wider.
>
>a) the general assumption that who is saying something is more important
>than what is being said.
When someone who has consistently demonstrated a complete lack of
sensibility and substance speaks, their opinions necessarily carry less
weight; i.e. when an idot has said ridiculous things 99 times in a row, the
100th thing tends to be ignored.
>
>b) relatedly, the only acceptable viewpoints being those wrapped up in the
>'correct' type of wrapping paper.
>
>c) relatedly again, the application by readers of their own choice of
>wrapping paper (for condemnation or approval) when there is uncertaintly as
>to origin.
>
>d) relatedly again, the idea that every new poster must fall into one of
the
>various camps; and the rapid depositing of newbies into said camps.
>
>e) the neat and easy solution of having a devil's pit into which to throw
>any unacceptable viewpoint. "he's just like Rolf, innit?"
>
>>
>>B. Incoherent - grammatically and semantically meaningless
>
>to some :-)
>
>but why look for the real gold when you have the fool's go(l)d already ?
What on earth does this mean?
>>
>>C. Filled by meaningless babble (Paradigm shift? Please...I thought only
>>management consultants prattled on about this).
>
>perhaps I'ld better make an attempt at translating the angel-outpourings
>into rgcc-acceptable prose ...........
You mean understandable english?
>what (s)he is trying to say (mixed up with various feminist and other bits)
>is:
>
>1. Chess knowledge/strength of play is stepped, not continuous. Players get
>gradually better within each step, but require a great leap to move from
>step to step. A great leap in ways of thought, usually requiring a throwing
>out of the previous way, each step. Not everybody has the necessary thought
>patterns/brain structure/personality to make it on and upwards.
>
This is the old 'Quantum theory of chess', postulated by Soltis among
others. In his case, he was describing chess knowledge, rather than a 'way
of thinking'. As a reasonably strong chess player (USCF Master), I'd tend
to agree with Soltis. I progressed from 1000 rated player (by steps) to USCF
2200 over several years. I increased my knowledge substantially, but I doubt
my 'way of thinking' (whatever that is) dramatically changed.
>2. It is possible, by reading/talking to strong players, to get some of the
>ideas/constructs that strong players have. But that a proper understanding,
>at the level (step) of a strong player, can only be achieved by
>doing/performing. The difference between intellectual; and intuitive, truly
>understanding knowledge. (if this is difficult for you, think of driving a
>car. i don't know how to drive by knowing what and where are the pedals and
>steering wheel; i only truly know by doing it, over and over until it
>becomes second nature. Only after this stage could I dream of going onto
the
>next steps, being a car driving instructor, or a rally driver, or car
>designer or whatever). And that, thus, getting two steps on, requires
>mastery of the step below.
Remember this argument for below...
>3. The crop of practising chess programmers are of low chess strength.
>Patzers, says Angel. On the basis that the top grades on icc are 3100+, icc
>grades appear inflated over fide grades by 500 points or more. Thus, for
>example Hyatt would be 1950-500=1450, Moreland 1650-500=1250 and so on.
Most
>of the other programmers are similar, they get to 1300,1400,1500,1600 fide,
>give up and start programming instead. Constant pattern.
Please cite evidence of these statements. Many GM's on ICC have FIDE ratings
+/- 200 points from their ICC ratings. I think it is very dangerous to
extrapolate a clear and linear link b/t FIDE and ICC ratings. Please cite
the
specific ratings of the programmers you are talking about. Constant pattern?
Because 'Angel [Sean] says' doesn't carry a whole lot of weight around here.
>4. Thus the crop of current chess programmers (not just Hyatt) are way way
>below GM strength. At 200 elo per step, that's 8 steps give or take. The
>quantum at which they don't understand that they don't know that they don't
>know is just too large to even think about.
>
As you said above, chess may well be one of those activities that one 'does'
rather than 'knows how to do'. If this is the case (and I'm not saying it
is) what benefit would come from being a GM programmer who could play GM
chess, and not describe (either in english or algorithmically) how it was
done?
>5. To know what a GM knows and these program/programmer combinations don't
>(eg a whole massive lot), and to be able to explain it to someone else, and
>to be able to quantify it somehow, and *algorithmise it* requires the
>programmer to be there himself. And they are not. They are way way off.
Again, if a GM can't explain it, how could he/she program it?
>6. The programmers know that the programs are lacking knowledge. Big time.
>They also know that another program, running 1000x faster, allegedly with a
>sophisticated eval function, a team of GMs help, the massive advantage of
>free singular extensions (from the massive parallelisation), is arguably
not
>even as good as a super-gm. They know that they are 10-15-20 years from
>reaching equivalent speeds. They know that nobody has any idea how to get
>in/quantify/algorithmise the incredibly complex knowledge required to deal
>with the problems shown up in Rebel-Anand RBBvRR material inbalance
>situation amongst many others. Especially not they with their 1200-1700
fide
>knowledge. So they know, in their hearts, that they can't get to fool's
>paradise.
OK. I'm a reasonable strong player. And a reasonable programmer. The
best chess program I've ever written was about USCF 1400 strength. Why?
Because *obviously* there are a great number of programming techniques,
ideas, and concepts that are *far* more important than the strength of the
programmer.
If the ability to write a strong program is correlated with the ability to
play good chess, then why did my program suck? Because I lacked the
necessary programming skill and knowledge to make it work well. Not the
chess knowledge...the programming ability.
What matters is the ability and knowledge of the programmmer. I don't know
how strong Bob Hyatt or Ed Schroeder is as a player; I do know how good they
are as programmers. In creating a good program, this is what counts.
If you have any evidence to the contrary, please cite it. Right now, Rebel,
Fritz and Crafty (running on very fast hardware) would *clearly* be among
the best 250 players in the world. This should make it clear that the
ability to write a very strong program is not the same as the ability to
play very strong chess oneself. Otherwise, it would be simply impossible for
these programs to be this strong. You seem to have refuted your own
argument.
>7. But this doesn't stop them pretending. Like an army that knows the war
is
>lost, they still have to forage for food, eat, retreat, continue. Keep that
>university post, maintain the chess program company profitability, get that
>PhD, sell those programs, make those lists.
>
>>
>>
>>On a related note, I can't believe the triple-digit-IQ crowd
>
>Where, where, where ? Show me them .... :-)
>
>>in this
>>newsgroup is spending time, effort, and bandwidth responding to this fool.
>
>Who spends the time, effort and bandwidth chasing fool's go(l)d ? :-)
>
>And who is the fool ? :-)
You mean besides those who treat other fools as if they had something
reasonable to say? :)
>
>>Don't you guys have a kill file?
>
>Yeah they do. It's called either the killfile, or the censored forum, or
>shouting down, or rudeness or denial. The intelligent ones just quit, got
>kicked out for being a nuisance, or ignored on the grounds they were Rolf.
>
Really? I thought the intelligent ones were Bob and others who contributed
something to this newsgroup about computer chess (you know...what this
newsgroup is nominally about?) rather than spew personal attacks (like
Rolfie) or pre-pubescent attention-seeking nonsense (like Sean and his
aliases).
>Steve
>
>
>>
>>Chris
>>
>>
>>Steve Blatchford wrote in message <6qph7u$ao2$1...@eros.clara.net>...
>>>
>>>
>>>Hi Bob,
>>>
>>>Could we please have a rather more substantive answer to the
intelligently
>>>argued case put forward by Angel (copied below) ?
>>>
>>>She is suggesting all the programmers and program ideas are dead-end
>>street,
>>>no ?
>>>
>>>Steve
Chris
Phil Innes
knowledge, it is
*experience* as well. The GM has seen so many positions already that he
can
use this to guide his play right along with his positional
understanding.
Ding an sich!
Mark Taimanov wrote the following this morning:
Against a background of rather colourless draws there was a brighter
looking duel between M.Kobalija - O.Korneev. In a sharp variation of the
"Four knights game", where
Black sacrifices a pawn for the initiative at an early stage, the
contenders up to the 17-th move were confidently following by game
Shirov-Short in Novgorod, but on the 17-th move Kobalija applied a
novelty ( instead of 17.0-0 - 17. Ng2!? ). The idea of the manoeuvre was
the consolidation of his forces and an aspiration to neutralize Black's
main attacking resource - the break-through f4.
Nevertheless O.Korneev had all the same to resort to this method even
if it meant material loss, for if White has time to castle, Black's
business would become simply joyless. There now emerged puzzling
complications, in which the White king tried to slip away from direct
threats, and Black had to put on the mantle of a new attack with
plentyful sacrifices. Eventually Kobalija managed to safely overcome all
danger and, when his king took cover in the silent harbour of the
Queen's side, Korneev was dished.
----
Mark is commenting on how GMs actually play each other under tournament
conditions. This one paragraph of comment is so far from the level of
commentary, often insisted upon, here, that he might well be talking
about a different game! There is a regretable tendency to validate the
computer, no matter what, and if the result doesn't make much chessic
sense, who is to know?
Combining a few recent posts with this idea it is clear that popular
opinion on what is sound, advisable, or can be attempted over the board
is missing a certain validity. Completely null ideas like "its okay if
you are 1700" are Kulturschande!
Chess playing engines perform at under 2000, as Expert B. They have not
made any contributions (relatively) to opening theory. It is the "stack"
of opening theory which allows them to achieve higher apparent ratings.
Without this advantage we have reached the end of what a single Turing
engine can do. The trend of evidence does not seem to indicate that
higher ratings can be achieved by faster processor or bus speeds.
I am really making a strong comment! There may be some future gains, but
nothing substantial. Improvement in chess-playing programs will not come
from this direction. If programming authors cannot incorporate (or even
understand) the different level of play as cited above, I cannot imagine
things changing!
This is not a simple subject, but probably central to real advances to
our area of interest here.
Cordially, Phil Innes
Robert Hyatt
: Bob Hyatt posted: The difference between a 2200 and a 2500 is not *just*
: knowledge, it is
: *experience* as well. The GM has seen so many positions already that he
: can
: use this to guide his play right along with his positional
: understanding.
: Ding an sich!
: Cordially, Phil Innes
I disagree, based on one principle that I hit in my AI class everytime I
teach it: there is *nothing* that mandates that a computer must "play chess
*like* a human".
IE, I ask my AI class "can someone name a flower that rhymes with "nose"?" and
everyone comes up with "rose". After a few more such examples "color that
rhymes with tack" and so forth, I ask them "how they do that?" And I point
out that *if* they start thinking about "linked lists" that they are on the
wrong track, because we'd have more pointers than data in our memory. And
we generally get off into "associative memory" for a bit here.
And everyone agrees that a computer can *never* do this sort of task. And
then I drop the "hammer"... and tell them "what if I have an electronic
dictionary containing the phonetic spelling/pronunciation of every word known
to man?" And they typically say "so what?" And I then point out that on a
good high-performance machine, I can search that *entire* dictionary in a
millisecond and *still* be able to name the flower that rhymes with nose, or
whatever.
A *vastly* different solution than what we do as humans, but does that
matter? Or does the correct result matter?
So, I say it again, I see *nothing* that says that we must analyze a game
as given above. Computers are fast, and getting faster, and programs are
getting better every year because of this. I don't believe for a second
that faster hardware doesn't make a program better, just look at the last
ten years for a convincing proof. And I don't believe that a computer has
to play chess using the *same methodology* as a human. It would be nice,
but it is not a necessary condition. With that background, I still firmly
believe that it is *possible*. Because computers are clearly stomping masters
at long and short games, and are beating a fair number of IM's at 40/2 as well.
And Deep Blue beat the best player in the world at 40/2, using this "inferior"
approach. Inferior to us, because we don't work like the computer. But maybe
not as inferior as we'd like to believe, because chess is an endeavor measured
by *results* not by *methodology*. If we can beat GM players at blitz today,
and that is absolutely a certainty supported by thousands of games on the chess
servers, then we can beat GM players at 40/2 tomorrow. How long before "tomorrow"
arrives? Not sure. But a *lot* sooner than "never"... Maybe 20 years for
micros, maybe not. 20 years sounds about right to me as an outer limit for
micro computers playing and beating GM's at 40/2 regularly. It might be
only 10. I'd probably choose 10 before I'd go 30 however..
Steve Blatchford
Robert Hyatt wrote in message <6qr399$89l$1...@juniper.cis.uab.edu>...
Here we go round the Mulberry bush, the Mulberry bush, the .......
>based on one principle that I hit in my AI class everytime I
>teach it: there is *nothing* that mandates that a computer must "play
chess
>*like* a human".
>
>IE, I ask my AI class "can someone name a flower that rhymes with "nose"?"
and
>everyone comes up with "rose". After a few more such examples "color that
>rhymes with tack" and so forth, I ask them "how they do that?" And I point
>out that *if* they start thinking about "linked lists" that they are on the
>wrong track, because we'd have more pointers than data in our memory. And
>we generally get off into "associative memory" for a bit here.
>
>And everyone agrees that a computer can *never* do this sort of task. And
>then I drop the "hammer"... and tell them "what if I have an electronic
>dictionary containing the phonetic spelling/pronunciation of every word
known
>to man?" And they typically say "so what?" And I then point out that on a
>good high-performance machine, I can search that *entire* dictionary in a
>millisecond and *still* be able to name the flower that rhymes with nose,
or
>whatever.
Wow ! That's a very clever teaching method ! Your class seems in total
thrall of your greatness ! Very impressive your majesty !
>
>A *vastly* different solution than what we do as humans, but does that
>matter? Or does the correct result matter?
>
>So, I say it again, I see *nothing* that says that we must analyze a game
>as given above. Computers are fast, and getting faster, and programs are
>getting better every year because of this.
Then why not write one program and leave it to be run on faster and faster
machines as time progresses ? Think of all that spare time you'ld have to
enthrall your class with !
>I don't believe for a second
>that faster hardware doesn't make a program better, just look at the last
>ten years for a convincing proof. And I don't believe that a computer has
>to play chess using the *same methodology* as a human.
Just as well. Since you and the other programmers can't play chess like a
strong human anyway !
>It would be nice,
>but it is not a necessary condition. With that background, I still firmly
>believe that it is *possible*. Because computers are clearly stomping
masters
>at long and short games, and are beating a fair number of IM's at 40/2 as
well.
>And Deep Blue beat the best player in the world at 40/2, using this
"inferior"
>approach. Inferior to us, because we don't work like the computer.
Ooh. I dunno about that. Some behave quite robotically at times !
>But maybe
>not as inferior as we'd like to believe, because chess is an endeavor
measured
>by *results* not by *methodology*.
Enter Thorsten, stage left muttering about bean-counters, evil-empires,
quality, mothersons, pyramids. etc. etc.
>If we
Identity crisis rears head (again) ....
>can beat GM players at blitz today,
>and that is absolutely a certainty supported by thousands of games on the
chess
>servers, then we
(and again)
>can beat GM players at 40/2 tomorrow. How long before "tomorrow"
>arrives? Not sure. But a *lot* sooner than "never"... Maybe 20 years for
>micros, maybe not. 20 years sounds about right to me as an outer limit for
>micro computers playing and beating GM's at 40/2 regularly. It might be
>only 10. I'd probably choose 10 before I'd go 30 however..
>
>
For other expositions of the faster, faster, deeper, deeper, more, more,
gasp, gasp ideology of chess programming please read rgcc Jan 1 1990, Jan 2,
Jan 3, ............ August 10 1998, August 11 1998 - reference Hyatt (hit me
with that rythm stick, sorry, baseball bat).
Still no answer to how to deal with the sort of problem demonstrated in
Anand-Rebel RRvRBB, all programs say losing side is +2.5 pawns winning,
because imbalanced material situation does not lend itself to assessment by
simplistic additive materialisic evaluation function as expoused by
gasp,gasp above. (Incedentally, Hyatt on ch 64, along with multitudes of
other programmers all were quite convinced that Rebel was winning - as in my
prog says +2.5, therefore I say +2.5, because I have no thought capability
of my own)
Therefore, demonstrating current approach is dead-end street. As ani anand
kno
Steve
PS for exposition of further nonsensical kibbitzing drivel tune in to ch 64
on day of any big match. One or two kno wot theys torking bout, but Hyatt
and the rest ? :))))))))))))))))))
Steve Blatchford
Robert Hyatt wrote in message <6qpkbe$orb$1...@juniper.cis.uab.edu>...
>Steve Blatchford <sblat...@clara.net> wrote:
>
>
>: Hi Bob,
>
>: Could we please have a rather more substantive answer to the
intelligently
>: argued case put forward by Angel (copied below) ?
>
>: She is suggesting all the programmers and program ideas are dead-end
street,
>: no ?
>
>: Steve
>
>know by name and/or reputation. There is nothing new in computer chess
>programs, just steady improvement. Don't put much stock in the stuff
>below *until* you see someone produce a reasonable program based on
>something "new".
>
>However, when it comes to "trolls" like the junk below, I try to avoid
>the stuff completely. If you are interested in where it comes from, check
>the headers, and "think"... and at least one or two answers will instantly
>jump to the front of the line. I "think" I know who is writing this stuff,
>but I won't say anything until I am sure.
>
>As far as a "dead-end street", "she" holds the license on that, no doubt.
>
Thus demonstrating all that is and has been wrong with rgcc for several
years. Hyatt. Mr Rude himself. Perhaps one should ask oneself what it is
about this man that causes all manner of other participants (Fong, Ed,
Chrisw, Rolf, Gillgasch to name but a few) to want to bring him down a peg
or two ?
The man is rude. He lies. His arguments are cheats. Much of what he says is
crapolla, big time. He endlessly pr's himself. As Mrs Evans says he'll do
whatever he has to to save his own skin.
>:)
>
>
>
>:>: Round and round we go with the CURIOUS MALE LOGIC.
>:>
>:>: Doh ! How to explain to the IDIOT PRETENDER ?
>
>First, you'd have to be smarter than the "pretender" so this case is
>over already... :)
Not difficult when dealing with Mr Thicko.
>
>:>
>:>: Chess strength is not on a continuum. It is stepped. Roughly 200
grading
>: points
>:>: per step.
>:>
>
>It is *not* stepped, and this shows *total* ignorance of chess. Just
>because rating classes within the USCF are broken into 200 point classes
>says *nothing* about the game of chess. This is drivel.
No. You talk drivel. It is a well established theory. Shows what you know.
>
>:>: You, Prof Hyatt, are on a low step. The same step as most of the other
>:>: pretend-knowledge-fools. You can tell who is on this step, becos they
>: say:
>:>: "positional is just long-range tactics". That is they think chess
>: strength is a
>:>: continuum. Faster, deeper, faster, deeper, faster, deeper, more, more,
>: gasp,
>:>: gasp, but Deep Blue came too soon. Leaving a great disatisfaction, not
>:>: alleviated by a good smoke.
>:>
>:>: Back to male logic: each step is a WAY OF THINKING. A new way of
thinking
>: over
>:>: the step below. So each step is quite a leap. Each step requires a
>: particular
>:>: kind of brain/thought, and, not just that, but a particular kind of
>: personality
>:>: to.
>
>Again, drivel, not supported by *any* GM or IM player that I can think of.
Here we go round the Mulberry bush, the Mul .....
Democratisation of your argument with imaginary GM's won't wash. I know more
GM's than you, nah, nah, nah, nah-nah.
>Perhaps by someone that had partaken of a few too many illegal substances,
>but not real chess players... :)
Here we go round the attempted character-assassination Mulberry bush, the
Mul ....
And Hyatt is a REAL chess-player ? :)))))))))
>
>:>
>:>: Just like you need a particular kind of brain to know that
4565233983745
>:>: divided by 347623847 is 394853 (or whatever), calculated in 2 seconds
>: flat.
>:>: there are these kind of people. Usually hopeless at everything else.
And
>: NOT
>:>: FEMALE.
>:>
>:>: Learn new language here: PARADIGM SHIFT. But u kno it alred.
>:>
>:>: Each 200 step needs paradigm-shift. Very difficult, not poss for many
>: (wrong
>:>: brain/personality). Throw away old, take on new. And take it on means
>: DO/LIVE
>:>: it. All very well to know positional concepts like 2 bishops; but
another
>: thing
>:>: altogether to PLAY POSITIONAL. But u no kno this, you on lower step.
>:>
>:>: REF: Kotov Think Like a Grandmaster. You gotta lern to think difrent.
>: Gettit ?
>
>Nope... Otherwise I guess I have gone through at least 8 "paradigm shifts"
>since my first program was under 1500, my current one is over 2300 (USCF),
>at least.
Misses point completely. Human strength and program elo are two entirely
different things.
>And my current program is not that different than my under 1500
>program was, same search ideas, same sort of evaluation ideas
yes, we all know you've not had an original idea in 30 years.
>(but more of
>them with faster hardware of course) and so forth. As I said, this is
really
>more of an idle "troll" than a statement of anything that means anything...
>which is typical of this poster most of the time... and the evidence on
the
>ID of this person is "mounting".. :)
Sean, Rolf, Chris, any more guesses ? You'll be the last to get it (as
ever). Your poker is worse than your chess.
>
>:>
>:>: U can kno stuff about step or two above your own level intellectually,
>: but not
>:>: in practice, not emotionally, not intuitively.
>:>
>:>: To know what a GM knows and your program doesn't (eg a whole massive
>: lot), and
>:>: to be able to explain it to someone else, and to be able to quantify it
>:>: somehow, and algorithmise it requires you to be there yourself. And you
>: are
>:>: not. You are way way off.
>:>
>:>: So stop pretending.
>:>
>:>: You wanna carry on ? Or give up now ? I got lots more.
>
And finally, just to finish off in the usual style, the BOTTY reference
.....
>People with diarrhea generally do... :)
>
Steve Blatchford
Christopher R. Dorr wrote in message <35d07...@news.one.net>...
>>newgroup and wider.
>>
>>a) the general assumption that who is saying something is more important
>>than what is being said.
>
>When someone who has consistently demonstrated a complete lack of
>sensibility and substance speaks, their opinions necessarily carry less
>weight; i.e. when an idot has said ridiculous things 99 times in a row, the
>100th thing tends to be ignored.
>
In the extreme case this would be reasonable. Except that here (rgcc and
ccc, computer chess in general, as well as chess itself) most posts are
pre-assessed by readers before they are read. You'll note endless references
to the writer of a post being this that or the other negative, rather than
to a direct addressing of the content. I'm taking Hyatt to task for this
elsewhere. IMO he's a particularly strong exponent of the ad hominem
argumentative style.
Partly this comes from chess itself, where there is a very strong pecking
order in value of output, based on elo. icc even has a filter based on elo,
no ? This does have the advantage of keeping supposedly low-grade patzer
comments under control; but also creates unpleasant side effects. Here, for
example.
And partly it comes from a natural tendency to believe and accept the words
of 'experts'. Most here are casual readers, who do they believe ? Lacking
background technical knowledge in a very confused field, they tend to
respect programmers output, titled persons output; and, in an argument
situation, if they are going to take sides, its much safer to side with
those with the right sort of wrapping paper for their output.
It has been pointed out [danger, rgcc off-topic word coming] that this leads
to a kind of technical fascism, with masses and 'expert' leaders. Made all
the worse when the experts are wrong, or lie, or are playing games, or have
axes to grind. I think Rolf saw this, and in his own way, fought it. The
rest you know.
>>
>>b) relatedly, the only acceptable viewpoints being those wrapped up in the
>>'correct' type of wrapping paper.
>>
>>c) relatedly again, the application by readers of their own choice of
>>wrapping paper (for condemnation or approval) when there is uncertaintly
as
>>to origin.
>>
>>d) relatedly again, the idea that every new poster must fall into one of
>the
>>various camps; and the rapid depositing of newbies into said camps.
>>
>>e) the neat and easy solution of having a devil's pit into which to throw
>>any unacceptable viewpoint. "he's just like Rolf, innit?"
>>
>>>
>>>B. Incoherent - grammatically and semantically meaningless
>>
>>to some :-)
>>
>>but why look for the real gold when you have the fool's go(l)d already ?
>
>
>What on earth does this mean?
Hyatt is the fool's god. My opinion. The gems of wisdom are to found
elsewhere. Often amongst his detractors.
>
>>>
>>>C. Filled by meaningless babble (Paradigm shift? Please...I thought only
>>>management consultants prattled on about this).
>>
>>perhaps I'ld better make an attempt at translating the angel-outpourings
>>into rgcc-acceptable prose ...........
>
>
>You mean understandable english?
Thanks for the compliment :)
>
>>what (s)he is trying to say (mixed up with various feminist and other
bits)
>>is:
>>
>>1. Chess knowledge/strength of play is stepped, not continuous. Players
get
>>gradually better within each step, but require a great leap to move from
>>step to step. A great leap in ways of thought, usually requiring a
throwing
>>out of the previous way, each step. Not everybody has the necessary
thought
>>patterns/brain structure/personality to make it on and upwards.
>>
>
>This is the old 'Quantum theory of chess', postulated by Soltis among
>others. In his case, he was describing chess knowledge, rather than a 'way
>of thinking'. As a reasonably strong chess player (USCF Master), I'd tend
>to agree with Soltis. I progressed from 1000 rated player (by steps) to
USCF
>2200 over several years. I increased my knowledge substantially, but I
doubt
>my 'way of thinking' (whatever that is) dramatically changed.
We kind of agree then ?
For myself, I am quite sure this step, plateau, step, plateau development
takes place.
At one time, I was on the tactical plateau. I knew positional concepts, and
could take advantage of them. Two bishops, whatever. But a stronger player,
who I played against often, kept saying: 'play positional, Steve, play
positional'. I'ld say, but I am, I know this positional stuff; and he'ld
say, no you're not PLAYING it. And for a while I didn't understand what he
meant. But sudddenly it clicked, I was always looking out for tactics, when
all I needed was to get my advantage and sit on it. The win was then
technical. I was playing tactical with positional knowledge. The shift
forwards, next step, whatever, was play positional.
So I got better on this plateau. Could do just fine with strong players. But
GMs and strong IMs would slaughter me; and I had no idea why. I'ld have good
positions and lose. Couldn't make it go my way. I'ld listen to them say 'go
here, this, go here, and what can black do?'; and I'ld think, but black is
better (additive simplistic eval function). But black would lose. How ?
Black has better position. But the key was 'what can black do?'. It takes
some time to grasp that even in a 'better' position, you can still be lost;
because 'what can you do?' Get that idea, and it is another plateau, the
plateau that says, let the other guy get a 'better' position that he can do
nothing with and then loses (Anand-Rebel). The idea that all that positional
additive evaluation can be rendered meaningless.
My point ? Is that the guys on 'positional is long-range tactics' level of
understanding don't understand or know or have any idea of the higher
concepts (not claiming I do, btw) necessary to evaluate positions. They
think it is a smooth slope, when it isn't. They think more additive
knowledge, more depth is the solution. IMO it isn't. It is not how the thing
works.
Folk-wisdom, my own experiences and intuition. Plus Hyatt's icc grade.
Moreland's is even lower.
>
>
>>4. Thus the crop of current chess programmers (not just Hyatt) are way way
>>below GM strength. At 200 elo per step, that's 8 steps give or take. The
>>quantum at which they don't understand that they don't know that they
don't
>>know is just too large to even think about.
>>
>
>As you said above, chess may well be one of those activities that one
'does'
>rather than 'knows how to do'. If this is the case (and I'm not saying it
>is) what benefit would come from being a GM programmer who could play GM
>chess, and not describe (either in english or algorithmically) how it was
>done?
None at all.
a) benefit from low chess knowledge programmer - none.
b) benefit from high knowledge programmer who can't describe or
algorithmise - none.
c) benefit from high knowledge programmer who can describe and
algorithmise - large.
>
>>5. To know what a GM knows and these program/programmer combinations don't
>>(eg a whole massive lot), and to be able to explain it to someone else,
and
>>to be able to quantify it somehow, and *algorithmise it* requires the
>>programmer to be there himself. And they are not. They are way way off.
>
>
>Again, if a GM can't explain it, how could he/she program it?
Quite. You need the explaining capability as well. Some of them can.
Although many chess books are make-work and fill-space, several actually
contain useful info, which implies these people do exist.
>
>>6. The programmers know that the programs are lacking knowledge. Big time.
>>They also know that another program, running 1000x faster, allegedly with
a
>>sophisticated eval function, a team of GMs help, the massive advantage of
>>free singular extensions (from the massive parallelisation), is arguably
>not
>>even as good as a super-gm. They know that they are 10-15-20 years from
>>reaching equivalent speeds. They know that nobody has any idea how to get
>>in/quantify/algorithmise the incredibly complex knowledge required to deal
>>with the problems shown up in Rebel-Anand RBBvRR material inbalance
>>situation amongst many others. Especially not they with their 1200-1700
>fide
>>knowledge. So they know, in their hearts, that they can't get to fool's
>>paradise.
>
>
>OK. I'm a reasonable strong player. And a reasonable programmer. The
>best chess program I've ever written was about USCF 1400 strength. Why?
>Because *obviously* there are a great number of programming techniques,
>ideas, and concepts that are *far* more important than the strength of the
>programmer.
>
>If the ability to write a strong program is correlated with the ability to
>play good chess, then why did my program suck? Because I lacked the
>necessary programming skill and knowledge to make it work well. Not the
>chess knowledge...the programming ability.
Sure. But all the Lego is around now. The ideas (and there are not so many
of them) are known. Add a little magic, stir it up, fiddle around and you're
away ...
>
>What matters is the ability and knowledge of the programmmer. I don't know
>how strong Bob Hyatt or Ed Schroeder is as a player; I do know how good
they
>are as programmers. In creating a good program, this is what counts.
It count to get where 'we' are today. It counts for the current paradigm
(dead-end street IMO). But it doesn't solve the holes [repeats Anand-Rebel
RRvRBB example].
>
>If you have any evidence to the contrary, please cite it. Right now, Rebel,
>Fritz and Crafty (running on very fast hardware) would *clearly* be among
>the best 250 players in the world. This should make it clear that the
>ability to write a very strong program is not the same as the ability to
>play very strong chess oneself. Otherwise, it would be simply impossible
for
>these programs to be this strong. You seem to have refuted your own
>argument.
Additive evaluation function, alfa-beta, nullmove, fast hardware, depth etc.
can only get you so far. Fast, deep search is, IMO, a kind of unsatisfactory
cheat, which, in any case, can be got round by application of human
intelligence.
>
>
>
>>7. But this doesn't stop them pretending. Like an army that knows the war
>is
>>lost, they still have to forage for food, eat, retreat, continue. Keep
that
>>university post, maintain the chess program company profitability, get
that
>>PhD, sell those programs, make those lists.
>>
>>>
>>>
>>>On a related note, I can't believe the triple-digit-IQ crowd
>>
>>Where, where, where ? Show me them .... :-)
>>
>>>in this
>>>newsgroup is spending time, effort, and bandwidth responding to this
fool.
>>
>>Who spends the time, effort and bandwidth chasing fool's go(l)d ? :-)
>>
>>And who is the fool ? :-)
>
>
>You mean besides those who treat other fools as if they had something
>reasonable to say? :)
>
Wise men know that even a fool has something to say.
See ? Human knowledge comes in steps, too !
>
>>
>>>Don't you guys have a kill file?
>>
>>Yeah they do. It's called either the killfile, or the censored forum, or
>>shouting down, or rudeness or denial. The intelligent ones just quit, got
>>kicked out for being a nuisance, or ignored on the grounds they were Rolf.
>>
>
>
>Really? I thought the intelligent ones were Bob and others who contributed
>something to this newsgroup about computer chess (you know...what this
>newsgroup is nominally about?) rather than spew personal attacks (like
>Rolfie) or pre-pubescent attention-seeking nonsense (like Sean and his
>aliases).
Bob is both the problem and the solution. On the one side he held (holds)
the news group together for years, while other came, went, gave up, got
bored. In this sense, he is the news group. He gets much credit from for the
free Crafty and his academic title. I think he abuses this credit.
On the other side he is an arrogant fool, spouting stupid opinions on
everything from dismemberment of convicts to child-rearing. He makes out he
knows everything about everything, PR's himself endlessly, any mention of
Rebel (for example) will result in pages of stuff about how Crafty does this
that and the other - by prolification of output he almosts creates a
situation where no other program/programmer can get a look in. He certainly
lies. He engages in untruthful character assassinations when the going gets
tough. He invokes in others a desire to argue with him, to bring him down a
few pegs, such that you can see the history of the newsgroup as a stream of
arguments between Bob and a series of participants, all with the same urge.
Steve
Robert Hyatt
: Robert Hyatt wrote in message <6qr399$89l$1...@juniper.cis.uab.edu>...
:>I disagree,
: Here we go round the Mulberry bush, the Mulberry bush, the .......
:>based on one principle that I hit in my AI class everytime I
:>teach it: there is *nothing* that mandates that a computer must "play
: chess
:>*like* a human".
:>
:>IE, I ask my AI class "can someone name a flower that rhymes with "nose"?"
: and
:>everyone comes up with "rose". After a few more such examples "color that
:>rhymes with tack" and so forth, I ask them "how they do that?" And I point
:>out that *if* they start thinking about "linked lists" that they are on the
:>wrong track, because we'd have more pointers than data in our memory. And
:>we generally get off into "associative memory" for a bit here.
:>
:>And everyone agrees that a computer can *never* do this sort of task. And
:>then I drop the "hammer"... and tell them "what if I have an electronic
:>dictionary containing the phonetic spelling/pronunciation of every word
: known
:>to man?" And they typically say "so what?" And I then point out that on a
:>good high-performance machine, I can search that *entire* dictionary in a
:>millisecond and *still* be able to name the flower that rhymes with nose,
: or
:>whatever.
: Wow ! That's a very clever teaching method ! Your class seems in total
: thrall of your greatness ! Very impressive your majesty !
So the way to discuss something is "if you can't offer any reasonable
facts or opinions, toss out insults"?? very good then...
:>
:>A *vastly* different solution than what we do as humans, but does that
:>matter? Or does the correct result matter?
:>
:>So, I say it again, I see *nothing* that says that we must analyze a game
:>as given above. Computers are fast, and getting faster, and programs are
:>getting better every year because of this.
: Then why not write one program and leave it to be run on faster and faster
: machines as time progresses ? Think of all that spare time you'ld have to
: enthrall your class with !
Er, maybe because I'd rather not wait? IE I happen to believe that eventually
depth alone will be enough. But, as I have said many times, "smarts" in
conjunction with depth seems to work better, if the "smarts" is done reasonably.
That's the direction I and everyone else I know of (that are writing chess
programs) are going...
:>I don't believe for a second
:>that faster hardware doesn't make a program better, just look at the last
:>ten years for a convincing proof. And I don't believe that a computer has
:>to play chess using the *same methodology* as a human.
: Just as well. Since you and the other programmers can't play chess like a
: strong human anyway !
You might be surprised. I'd be willing to find out one night if you want to
log on to ICC and play a couple of games...
:>It would be nice,
:>but it is not a necessary condition. With that background, I still firmly
:>believe that it is *possible*. Because computers are clearly stomping
: masters
:>at long and short games, and are beating a fair number of IM's at 40/2 as
: well.
:>And Deep Blue beat the best player in the world at 40/2, using this
: "inferior"
:>approach. Inferior to us, because we don't work like the computer.
: Ooh. I dunno about that. Some behave quite robotically at times !
:>But maybe
:>not as inferior as we'd like to believe, because chess is an endeavor
: measured
:>by *results* not by *methodology*.
: Enter Thorsten, stage left muttering about bean-counters, evil-empires,
: quality, mothersons, pyramids. etc. etc.
Doesn't matter to me. I'd be just as happy beating Kasparov 6 0 with
horrible-looking chess moves (but which worked) as I would beating him
with chess moves that result in a shower of gold coins from the gallery.
You might be surprised. I'd be willing to find out one night if you want to
log on to ICC and play a couple of games...
:>If we
: Identity crisis rears head (again) ....
:>can beat GM players at blitz today,
:>and that is absolutely a certainty supported by thousands of games on the
: chess
:>servers, then we
: (and again)
:>can beat GM players at 40/2 tomorrow. How long before "tomorrow"
:>arrives? Not sure. But a *lot* sooner than "never"... Maybe 20 years for
:>micros, maybe not. 20 years sounds about right to me as an outer limit for
:>micro computers playing and beating GM's at 40/2 regularly. It might be
:>only 10. I'd probably choose 10 before I'd go 30 however..
:>
:>
: For other expositions of the faster, faster, deeper, deeper, more, more,
: gasp, gasp ideology of chess programming please read rgcc Jan 1 1990, Jan 2,
: Jan 3, ............ August 10 1998, August 11 1998 - reference Hyatt (hit me
: with that rythm stick, sorry, baseball bat).
If you want to be a wise guy, I'll simply point to Deep Blue vs Kasparov
and say "Q.E.D." Tough to refute.
: Still no answer to how to deal with the sort of problem demonstrated in
: Anand-Rebel RRvRBB, all programs say losing side is +2.5 pawns winning,
: because imbalanced material situation does not lend itself to assessment by
: simplistic additive materialisic evaluation function as expoused by
: gasp,gasp above. (Incedentally, Hyatt on ch 64, along with multitudes of
: other programmers all were quite convinced that Rebel was winning - as in my
: prog says +2.5, therefore I say +2.5, because I have no thought capability
: of my own)
I believe that is *my* idea you are using. Maybe you should come up with
your own? If that is possible. Had Rebel gone a few plies deeper it would
not have been +2.5, because he resigned at -2.something, and it was only a
few moves later. So it is solvable by depth, or it is solvable by new search
extensions, or it is solvable by new "knowledge" in the evaluation. But it is
certainly solvable...
: Therefore, demonstrating current approach is dead-end street. As ani anand
: kno
So dead end that it would have no chance at all, right? Except against
Kasparov? And even in the Anand games, Anand was in an inferior position in
both games, IMHO. In the first he was quite lucky to draw. In the second,
he might have been lucky to win.
: Steve
: PS for exposition of further nonsensical kibbitzing drivel tune in to ch 64
: on day of any big match. One or two kno wot theys torking bout, but Hyatt
: and the rest ? :))))))))))))))))))
You don't want to discuss computer chess, that's fine by me. I'll make this
my last response to you. You may have the last words, however worthless and
exaggerated they may be...
Steve Blatchford
A related posting from the world of computer go.
================================================
During a commentary at the EGC in Romania Saijo 8p claimed
that computers can never understand go because there be
too many variations and shortcuts by means of intuition be
inconceivable. After that I had a short discussion with
him about professional and computer thinking.
Saijo wondered why Westerners do always ask for reasons
instead of simply applying feelings. I replied this be
essential for programming and inquired about his thinking.
But he could just state he have no analytical mind and
absolutely cannot tell reasoning where there is intuition,
he be barely a professional go player. And I, of course,
was merely a go programmer:)
Now which approach will beat which when it comes to
programming: logic or pseudo-intuition? Or to be a little
more provocative: Have Asian professionals any chance to
create programs of professional level?
================================================
Always asking for reasons instead of simply applying feelings.
Human knowledge comes in steps :)
Steve
EvaNova1
>any unacceptable viewpoint. "he's just like Rolf, innit?"
Steve, I'm new here in this group. I still haven't decided which program to
buy. Fritz or Crafty. Now you talk about Rolf. Could you or someone explain the
main differences of these programs?
Kind regards
Eva
Robert Hyatt
:>e) the neat and easy solution of having a devil's pit into which to throw
:>any unacceptable viewpoint. "he's just like Rolf, innit?"
: Steve, I'm new here in this group. I still haven't decided which program to
: buy. Fritz or Crafty. Now you talk about Rolf. Could you or someone explain the
: main differences of these programs?
: Kind regards
: Eva
First, you can't buy Crafty as it is free. Second, Fritz is a stronger
program with a more polished front-end. Third, you wouldn't want a "Rolf"
if it were given to you as part of a 100 million dollar lottery.
Papalenvoy
>From: evan...@aol.com (EvaNova1)
>Date: 12/8/98 03:18 GMT
>
>>e) the neat and easy solution of having a devil's pit into which to throw
>>any unacceptable viewpoint. "he's just like Rolf, innit?"
>
>Steve, I'm new here in this group. I still haven't decided which program to
>buy. Fritz or Crafty. Now you talk about Rolf. Could you or someone explain
>the
>main differences of these programs?
>
>Kind regards
>
>Eva
>
Hallo there Eva my child, here is Father O'Flatterty again.
They tell me Fritz is from the Great Satan, so you'll not be wanting that one.
Crafty comes from the Great Know-it-All, but they tell me really he knows very
little.
And Rolf. Well, he's a Crafty Fritz, so there you have the best of both worlds
!
The Work of the Lord be done today.
Father O'Flatterty
>
Cameron Hayne
<sblat...@clara.net> wrote:
>Additive evaluation function, alfa-beta, nullmove, fast hardware, depth etc.
>can only get you so far. Fast, deep search is, IMO, a kind of unsatisfactory
>cheat, which, in any case, can be got round by application of human
>intelligence.
I can see that it might be called "unsatisfactory" since it is not not yet
(with current machines) fast or deep enough to produce consistent wins
against the best human players.
But in calling it a "cheat" I think you betray your true feelings
- you seem to have something against this arguably mechanistic method
just because it is mecanistic. A "satisfactory" chess playing computer
for you would seem to require using similar methods to those used by
human chess players. That's fine if that's what interests you but
I think that a chess playing program should be measured by its results
(i.e. wins, losses, draws) not by the way it obtains those results.
--
Cameron Hayne (ha...@crim.ca)
Centre de recherche informatique de Montreal
EvaNova1
>First, you can't buy Crafty as it is free. Second, Fritz is a stronger
>program with a more polished front-end. Third, you wouldn't want a "Rolf"
>if it were given to you as part of a 100 million dollar lottery.
>:)
>--
>Robert Hyatt Computer and Information Sciences
>hy...@cis.uab.edu University of Alabama at Birmingham
>(205) 934-2213 115A Campbell Hall, UAB Station
>(205) 934-5473 FAX Birmingham, AL 35294-1170
Thank you Mr. Hyatt!
I see that you can't use simple logic to answer a question. I asked what
program to buy. Fritz, Crafty or Rolf. I asked for the differences. Which one
is better? A free Crafty or a stronger Fritz? And Rolf is too strong for me?
Computerchess seems to be a strange field for us women.
I play chess and need a strong training. What I don't need is male chauvinism
of the Father and you, Mr. Hyatt.
Kind regards
Eva
EvaNova1
>Hallo there Eva my child, here is Father O'Flatterty again.
>They tell me Fritz is from the Great Satan, so you'll not be wanting that one.
>Crafty comes from the Great Know-it-All, but they tell me really he knows very
>little.
>And Rolf. Well, he's a Crafty Fritz, so there you have the best of both worlds
>!
>The Work of the Lord be done today.
>Father O'Flatterty
>>
I think I'm in the wrong newsgroup. I thought this one was about computerchess.
As a chessplayer I need a good training.
Father, you seem to be part of the male macho world. Do you really think that
you've answered my question for the best program?
I'll better leave this group.
I came with love.
What I got is male aggressiveness.
---
Kind regards
Eva
Robert Hyatt
: In article <6qs4se$qr5$1...@eros.clara.net>, "Steve Blatchford"
: <sblat...@clara.net> wrote:
:>Additive evaluation function, alfa-beta, nullmove, fast hardware, depth etc.
:>can only get you so far. Fast, deep search is, IMO, a kind of unsatisfactory
:>cheat, which, in any case, can be got round by application of human
:>intelligence.
: I can see that it might be called "unsatisfactory" since it is not not yet
: (with current machines) fast or deep enough to produce consistent wins
: against the best human players.
: But in calling it a "cheat" I think you betray your true feelings
: - you seem to have something against this arguably mechanistic method
: just because it is mecanistic. A "satisfactory" chess playing computer
: for you would seem to require using similar methods to those used by
: human chess players. That's fine if that's what interests you but
: I think that a chess playing program should be measured by its results
: (i.e. wins, losses, draws) not by the way it obtains those results.
It "betrays" a lot more than that, if you *really* think about it. Most
of us use "cheat" as a verb, with one notable exception from times past.
When you think about it, you might conclude that this post was written
by someone you know, then the grammer was polished by someone *else* that
you probably know, before being deposited here. I will tell you my
suspicious off-line if you are interested. But it does "fit" well... :)
: --
: Cameron Hayne (ha...@crim.ca)
: Centre de recherche informatique de Montreal
--
Robert Hyatt
: PS:
: *** BIG HINT ***
: There seems to be a presumption that you want to really get to Hyatt...
: These multi-ID tinklings do *nothing*, and IMO help his reputation with
: newbies and drifters, and do nothing to upset his relationships with
: regulars.
Reputation doesn't really matter here. I choose to ignore faceless,
IDless folks that only want to play the game of "name calling." It
matters not whether someone likes crafty, or hates it, it matters not
whether they like my coding style or hate it, and it matters not whether
they think I should be selling it, giving it away or doing neither. In
some things, I "follow my heart" and do what I want to do. End of that
story. If I choose to quite reading/posting on r.g.c.c, it will be
because *I* choose to do so, not because someone wants to pull out his
insult-gun daily and point it my way..
: Instead... just write a program stronger than crafty. Using Knowledge.
: BTW, in doing so your's will be stronger than Chess System Tal too.
:)
: Should be easy, Steve, no?
: One author of a strong program gives you his whole source code to start
: with (and Tom K gives the source for his starter), should you want it.
: Oh yes, it was Bob Hyatt, who you insinuate tries to exclude others from
: the top of the pile.. Strange to give away your source code then....
: Does not compute.
Never did compute. There's more going on than meets the eye, just that
I don't have the time, energy or interest to figure out what it is. For
those that don't like crafty, don't use it. For those that don't like
me personally, ignore me. For those that think my opinions are all wrong,
don't read them (I do try to carefully indicate what is opinion, vs what I
have found to be "observed fact."
I fail to see the need to respond to those that you don't like, don't
trust, or whatever. In every newsgroup I find a few that I fit into that
category, and I simply choose to not respond to their posts, and we never
even have a chance to "cross-swords" in a newsgroup as a result, because I
formed an opinion about them, and made a conscious decision to simply "stay
away and not get sucked into an argument that leads nowhere."
The "bean counter" analogy is one example. If you don't like "bean counting"
programs, I don't have a problem with that. If you want to explain how your
program plays well without counting beans, that's ok too. If you say it often
enough, expect to be challenged to produce results rather than posts, however.
I can take my lumps OTB. Anyone with that strong of an opinion ought to be
prepared to do the same... ie "don't throw rocks in glass houses." is good
advice.
: Chop out the brute force bits if you want, just use the piece bitmaps.
: Put your knowledge based stuff in.
: But just come up with a program that scores >45% vs crafty, containing
: enough of "you" to distinguish it in the eyes of programmers from
: crafty.
: Now I believe Bob would actually be pleased or indifferent if you did
: this. But by your logic he would be rattled.
There are many "variants" of Crafty running around. Tom's program is a
much better starting point, because I'd rather deal with 1,000 lines of
code than 40,000 lines of code as a beginner. My only interest in giving
the source away was these two points: (1) to prevent everyone from trying
the same ideas over and over and over before they find what works and what
doesn't work. It has been a common problem with every new chess programmer
that I have talked to via email. This is one way to at least show what is
working now, and by reading the "history" in main.c, you can find out what
was tried and what worked and what failed, all the way back to day 1 when it
played its first game, something that has been totally unavailable in the
past. (2) to get the occasional idea via feed-back that makes something
better. I still get fixes here and there, new ideas here and there, and
lots of suggestions and feedback. None of which I would get if I had done
as I did with Cray Blitz...
: So just do it, "Steve". Beat crafty. I'll be waiting for you ;-)
: Or else keep helping to build up Bob's position.
: Fine by me, either way ;-)
: As always, I am only trying to help.
I only wish such "folks" would try being "constructive" rather than
"destructive" here. But knowing who they are ( at least reasonably
sure who they are ) this is not very likely to happen. Based on
past experience.
We will never be free of such behavior, unfortunately, with as many readers
as we have. We have only to figure out how to work around the static, to get
to the real "signal" and survive as best we can...
Robert Hyatt
: papal...@aol.com (Papalenvoy) wrote:
:>Father O'Flatterty
:>>
: I came with love.
No Sean, you got what you wanted...
Robert Hyatt
: Robert Hyatt <hy...@crafty.cis.uab.edu> wrote:
:>First, you can't buy Crafty as it is free. Second, Fritz is a stronger
:>program with a more polished front-end. Third, you wouldn't want a "Rolf"
:>if it were given to you as part of a 100 million dollar lottery.
:>:)
:>--
:>Robert Hyatt Computer and Information Sciences
:>hy...@cis.uab.edu University of Alabama at Birmingham
:>(205) 934-2213 115A Campbell Hall, UAB Station
:>(205) 934-5473 FAX Birmingham, AL 35294-1170
: Thank you Mr. Hyatt!
: I see that you can't use simple logic to answer a question. I asked what
: program to buy. Fritz, Crafty or Rolf. I asked for the differences. Which one
: is better? A free Crafty or a stronger Fritz? And Rolf is too strong for me?
: Computerchess seems to be a strange field for us women.
I believe I answered your question. You can't buy crafty, it is free. I
also said Fritz was better with a much nicer gui with lots of features and
is for sale. "rolf" is *not* a chess program. It is some sort of lifeform
that behaves in strange and odd ways.
: I play chess and need a strong training. What I don't need is male chauvinism
: of the Father and you, Mr. Hyatt.
There was no cheuvinism on my part, only a lack of reading on your part.
Re-read my answer carefully. It is clear. And non-gender based.
: Kind regards
: Eva
SHUT UP, SEAN!
and annoying!
EvaNova1 <evan...@aol.com> wrote in article
<199808121810...@ladder01.news.aol.com>...
> papal...@aol.com (Papalenvoy) wrote:
>
> >Hallo there Eva my child, here is Father O'Flatterty again.
>
> >They tell me Fritz is from the Great Satan, so you'll not be wanting
that one.
>
> >Crafty comes from the Great Know-it-All, but they tell me really he
knows very
> >little.
>
> >And Rolf. Well, he's a Crafty Fritz, so there you have the best of both
worlds
> >!
>
> >The Work of the Lord be done today.
>
> >Father O'Flatterty
>
> >>
>
> I think I'm in the wrong newsgroup. I thought this one was about
computerchess.
> As a chessplayer I need a good training.
>
> Father, you seem to be part of the male macho world. Do you really think
that
> you've answered my question for the best program?
>
> I'll better leave this group.
>
> I came with love.
>
> What I got is male aggressiveness.
>
>
>
> ---
> Kind regards
>
> Eva
>
>
>
>
>
mclane
>Computerchess seems to be a strange field for us women.
isn't this chauvinism ? IMO women play better chess than male.
Only that there are less women playing makes it look vice versa.
The girls I knew were able to learn much faster than males.
>I play chess and need a strong training. What I don't need is male chauvinism
>of the Father and you, Mr. Hyatt.
Rolf is no program. He is a person.
You cannot play with him...although he likes to play.
If you want a strong program, and you want to learn about chess, buy
rebel, mchess or virtual-chess, hiarcs or fritz.
Why do you insult Bob Hyatt to be chauvinist ?! I think this is not a
very good beginning here...
>Kind regards
>Eva
best wishes
mclane
mclane
>I think I'm in the wrong newsgroup. I thought this one was about computerchess.
this group is sometimes about computerchess. Most often it is about
males trying to mark their area with showing their superiority.
>As a chessplayer I need a good training.
than buy a knowledge based chess program...
mchess or rebel or hiarcs.
>I'll better leave this group.
do what you want to do.
>I came with love.
Calling somebody a chauvinist is not coming with love...
>What I got is male aggressiveness.
look in the mirror. YOU insulted somebody...
best wishes
mclane
Phil Innes
my post :
I would say three things. This extension of computer
performance may not be linear.
There is no evidence that simple brute force approaches will
solve pattern oriented problems.
A second point is that this view does not seem to be shared
(perhaps naturally) by the GM community.
Thirdly - and in my long initial post to this subject I
perhaps burried it - is the following: We agree, do we not,
that current playing strength is sub-2000. (Human players are
not allowed instantaneous access to massive opening libraries,
so lets make it even). I know from the human level of playing
that the conditions that support human level play at 19xx have
nothing to do with the kind of knowledge necessary to play at
2300.
I do not want to say too much, or be too spectacular. Am
enjoying the conversation. Be glad to hear your response.
Cordially, phil.
Christopher R. Dorr
Steve Blatchford wrote in message <6qs4se$qr5$1...@eros.clara.net>...
>>>Yet again demonstrating the disastrous set of relationships within this
>>>newgroup and wider.
>>>
>>>a) the general assumption that who is saying something is more important
>>>than what is being said.
>>
>>When someone who has consistently demonstrated a complete lack of
>>sensibility and substance speaks, their opinions necessarily carry less
>>weight; i.e. when an idot has said ridiculous things 99 times in a row,
the
>>100th thing tends to be ignored.
>>
>
>In the extreme case this would be reasonable. Except that here (rgcc and
>ccc, computer chess in general, as well as chess itself) most posts are
>pre-assessed by readers before they are read. You'll note endless
references
>to the writer of a post being this that or the other negative, rather than
>to a direct addressing of the content. I'm taking Hyatt to task for this
>elsewhere. IMO he's a particularly strong exponent of the ad hominem
>argumentative style.
I've been around this newgroup for four or five years...the only times I
have seen Bob react in an ad hominem manner was when he had been previously
personally attacked by the poster. In general, his style has appeared (to
me) to be very reasonable. Are you stating that he should not 'give as good
as he gets?'. That would appear to be a difficult to maintain position.
>Partly this comes from chess itself, where there is a very strong pecking
>order in value of output, based on elo. icc even has a filter based on elo,
>no ? This does have the advantage of keeping supposedly low-grade patzer
>comments under control; but also creates unpleasant side effects. Here, for
>example.
>
>And partly it comes from a natural tendency to believe and accept the words
>of 'experts'. Most here are casual readers, who do they believe ? Lacking
>background technical knowledge in a very confused field, they tend to
>respect programmers output, titled persons output; and, in an argument
>situation, if they are going to take sides, its much safer to side with
>those with the right sort of wrapping paper for their output.
But how did they get the 'right sort of wrapping' in the first place? Bob is
*obviously* one of the best qualified posters in this group. He wrote
Crafty, a clearly world-class program. He has a PhD (which does mean
something, Rolf's and Seans protestations notwithstanding), and he is a
professor. To me, these are 'credentials', not 'wrapping'. When you get on
an airplane, I assume that you assume that your pilot has the right
credentials. Would you be as trusting if the pilot didn't have the right
'wrapping'? It doesn't mean that he can't fly, it's just that if he has the
creds, then he probably can.
>It has been pointed out [danger, rgcc off-topic word coming] that this
leads
>to a kind of technical fascism, with masses and 'expert' leaders. Made all
>the worse when the experts are wrong, or lie, or are playing games, or have
>axes to grind. I think Rolf saw this, and in his own way, fought it. The
>rest you know.
>
Rolf rarely talked about technical issues. He accused Ed of stealing, and
Bob of being a fascist. How are these accusations related to a technical
disagreement with one of the experts. You mentioned 'ad hominem' attacks
before? Rolf was *the* major progenitor of them.
>>>
>>>b) relatedly, the only acceptable viewpoints being those wrapped up in
the
>>>'correct' type of wrapping paper.
>>>
>>>c) relatedly again, the application by readers of their own choice of
>>>wrapping paper (for condemnation or approval) when there is uncertaintly
>as
>>>to origin.
>>>
>>>d) relatedly again, the idea that every new poster must fall into one of
>>the
>>>
>>>to some :-)
>>>
>>>but why look for the real gold when you have the fool's go(l)d already ?
>>
>>
>>What on earth does this mean?
>
>Hyatt is the fool's god. My opinion. The gems of wisdom are to found
>elsewhere. Often amongst his detractors.
I disagree (as I believe the sizeable majority of newsgroup members would).
Bob's ideas and explanations have greatly added to my understanding of
computer chess. He has indeed passed many 'gems of wisdom'. Please find me
something of a remotely similar value ever posted by Rolf or Sean.
Kind of, but I feel that this is only for humans. I have had several student
who would progress in fits and starts. 1200 for a year, then *boom* 1450.
1450 for 18 months, then *boom* 1700. This happened to me. But I believe
that it is based on knowledge, and how humans collect and synthesize this
knowledge. I believe that chess programs proceed in a more linear manner.
This extension adds 15 points. This heuristic, perhaps 30. I believe that it
points to a fundamental fact: programs and people play chess in vastly
different manners, even though the results may be similar. This is one
reason why I believe that the competence of the programmer is of vastly
greater importance than the strength of his or her play. Bob Hyatt
understands very well how computers play chess. Anand understands very well
how humans play chess (not stylistically...I mean the process). I doubt that
greater insight into how a person thinks about chess will translate into a
better program. If I had to try to improve a chess program, I'd rather have
Bob at my side than Vishy 100 times out of 100.
Again, I believe this is almost irrelevent to their ability to improve their
already strong programs. If what you are postulating is correct, then there
should be a linear relationship between the chess strength of the
programmer, and the strength of his program. I believe that there is a great
deal of evidence that is contrary to this.
>>>4. Thus the crop of current chess programmers (not just Hyatt) are way
way
>>>below GM strength. At 200 elo per step, that's 8 steps give or take. The
>>>quantum at which they don't understand that they don't know that they
>don't
>>>know is just too large to even think about.
>>>
>>
>>As you said above, chess may well be one of those activities that one
>'does'
>>rather than 'knows how to do'. If this is the case (and I'm not saying it
>>is) what benefit would come from being a GM programmer who could play GM
>>chess, and not describe (either in english or algorithmically) how it was
>>done?
>
>None at all.
>
>a) benefit from low chess knowledge programmer - none.
>
>b) benefit from high knowledge programmer who can't describe or
>algorithmise - none.
>
>c) benefit from high knowledge programmer who can describe and
>algorithmise - large.
Again, our GM/programmer can algorithmise (Is this really a word?! :) ) how
*he* plays chess. WIll this benefit his program? Doubtfuls, as we have
already decided that the fundamental way each of them approaches and plays
chess is very different.
>
>>
>>>5. To know what a GM knows and these program/programmer combinations
don't
>>>(eg a whole massive lot), and to be able to explain it to someone else,
>and
>>>to be able to quantify it somehow, and *algorithmise it* requires the
>>>programmer to be there himself. And they are not. They are way way off.
>>
>>
>>Again, if a GM can't explain it, how could he/she program it?
>
>Quite. You need the explaining capability as well. Some of them can.
>Although many chess books are make-work and fill-space, several actually
>contain useful info, which implies these people do exist.
>
But even if the can explain it, how does it help their program?
But these computer-programming ideas may be fundamentally incompatible with
the way people play chess. I believe that our chess knowledge may not truly
be able to be synthesized with the techniques that have *proven* time and
time again that they produce the best computer-played chess.
>>
>>What matters is the ability and knowledge of the programmmer. I don't know
>>how strong Bob Hyatt or Ed Schroeder is as a player; I do know how good
>they
>>are as programmers. In creating a good program, this is what counts.
>
>It count to get where 'we' are today. It counts for the current paradigm
>(dead-end street IMO). But it doesn't solve the holes [repeats Anand-Rebel
>RRvRBB example].
>
>>
>>If you have any evidence to the contrary, please cite it. Right now,
Rebel,
>>Fritz and Crafty (running on very fast hardware) would *clearly* be among
>>the best 250 players in the world. This should make it clear that the
>>ability to write a very strong program is not the same as the ability to
>>play very strong chess oneself. Otherwise, it would be simply impossible
>for
>>these programs to be this strong. You seem to have refuted your own
>>argument.
>
>Additive evaluation function, alfa-beta, nullmove, fast hardware, depth
etc.
>can only get you so far. Fast, deep search is, IMO, a kind of
unsatisfactory
>cheat, which, in any case, can be got round by application of human
>intelligence.
This seems to be an argument with the discipline of 'Artificial
Intelligence' in general, rather than chess programming in particular. Youir
point seems to come down to 'In order to mimic intelligent performance on
some task deemed to represent 'intelligence', should we mimic human thought
and behaviour, or simply pay attention to the result, and tailor the process
to the strengths of the computer?'
This is a fundamental AI question, and has been around for a very long time.
I'm not an AI expert, but it is my understanding that in many AI disciplines
(knowledge-based system, imaging and pattern recognition, and certainly in
chess) that the results from tailoring the techniques to the computer's
strengths have been better than those when human-like thought and behaviour
is attempted.
As I said before, I have never seen Bob act like an 'arrogant fool'. He
doesn't seem particularly self-promoting to me. The reason the talk usually
focus on Crafty is because the source and ideas are readily available. We
can say 'Look in thisfile.c at line 120, and see how this workd. Can't do
that with the commercial programs.
Personally, I have a great deal of respect for Bob, and a great deal of
appreciation for the knowledge that hes has given freely in this forum.
If Rolfie or Sean would like to contribute anything of meaning, perhaps our
respect for them would increase. But for now, it seems that they are
content to attack people personally, and try to garner attention for
themselves by acting out.
When their actions are deserving of respect, perhaps they will earn some.
Until then....
Christopher
Rolf Tueschen
Let me jump in this thread for a moment. I understand why you are upset.
It's true that nobody really answered your questions.
As you can see from my adress I am Rolf the one you were talking about.
Obviously misleaden by the typical drivel here of a very few folks who do
nothing but defaming me. Read in Dejanews.com what the wanted to do with me
in the past.
I'm a human being and NOT a computer program.
Let me answer your questions. It would help if you could give your approx.
strength. I could then give you better hints for a new program.
As Rober Hyatt already mentioned, FRITZ is probably the best commercial
program. It had different levels down to beginner but you are surely way
better yet. If you want to study chess seriously you really might decide to
buy FRITZ. And the reason is clear. You also can deal with thoussands of
deeply commented GM games, that means the games of the best players of all
time. Only with FRITZ you have an internationally valuable format for the
games collection. Other programs also have games but their format is NOT to
use worldwide. And these programs have no commentary or subvariations in
the games. But as a serious student you surely want to read what good
players thought about their own moves, no?
You didn't write too much about your other hobbies or education. Eva, if
you have only an average understanding of the internet, how to download
stuff and then how to deal with it in directories, you also could search
for the mentioned Crafty. Because as said it's freeware and in its strength
it's surely sufficiently strong for you if you're not a real WIM or WGM
actualley.
Then you try to play a few training games. If you want to download Crafty
you better ashould ask again because others might give you pages of
instructions how to download and then to build up this program. So all
depends on your preferences.
Write again if you have experiences yet with programs. Then special hints
could be interesting. Because different programs have different highlights.
The probably best member here for such questions is a guy called KK. That's
a canadian who collects all sort of newbie hints. Perhaps he's on vacation
at the moment because normally he could answer your questions better than
Mr. Hyatt. Who's simply a bit to high away from your interests. But he's
the author of the free program.
If you want to take a look also into res.games.chess.misc and
rec.games.chess.analysis. Both groups are much more peaceful than this
techno group ...
many greetings from germany
Rolf Tueschen
=================================================================
evan...@aol.com (EvaNova1) wrote:
>Robert Hyatt <hy...@crafty.cis.uab.edu> wrote:
>>First, you can't buy Crafty as it is free. Second, Fritz is a stronger
>>program with a more polished front-end. Third, you wouldn't want a "Rolf"
>>if it were given to you as part of a 100 million dollar lottery.
>>:)
>>--
>>Robert Hyatt Computer and Information Sciences
>>hy...@cis.uab.edu University of Alabama at Birmingham
>>(205) 934-2213 115A Campbell Hall, UAB Station
>>(205) 934-5473 FAX Birmingham, AL 35294-1170
>Thank you Mr. Hyatt!
>I see that you can't use simple logic to answer a question. I asked what
>program to buy. Fritz, Crafty or Rolf. I asked for the differences. Which one
>is better? A free Crafty or a stronger Fritz? And Rolf is too strong for me?
>Computerchess seems to be a strange field for us women.
>I play chess and need a strong training. What I don't need is male chauvinism
>of the Father and you, Mr. Hyatt.
>Kind regards
>Eva
Phil Innes
length in this same thread. I should like to say that we have (so far)
been able to disagree on a subject without calling each other names or
disparaging the other's point of view. Lets return to our conversation
at length... neither of us need to make some urgent point in the next
ten minutes.
Is this a fair characterisation of where we leave it (or you say it),
for your point of view:
"Brute force machine efficiencies and programming finesse will continue
to develop chess engines, as they have for the past ten years until they
achieve parity with GMs on a regular basis."
and mine: "there is a limit (as for human players) in linear
development, which we might now be nearing, and, GMs behave in a
different way than 1900 players, and are invested in a radically
different pattern-dynamic to achieve this differential."
Cordially, Phil
Phil Innes
"How does a strong human really play chess?" is the question that Steve
wants an algorithmic answer to!
Might as well ask "How does a smart person actually think?"
This has only philosophical interest - admittedly deep - for a very,
very long time it won't help with chess. We have no clues.
Vo bist du? A minor amendment is perhaps that there is no
involvement in chess-computing with said subject. Initiatives
tend to come from the other direction; psychology and involve
chess. Howard Gardner, writing in a book about various kinds
of intelligence, cites, in particlular, a spatial
intelligence. He goes on for about 12 pages on high level
chess playing abilities as being particularly non-linear.
Neural networking work on pattern-recognition isn't step 1, it is barely
step zero.
Again the right starting point, IMO. Gardner speaks of GMs
knowing 50,000 patterns. (Clear from his context, patterns not
'positions.')
So, massively parrallel it is. But what is it? LOL.
Phil
Robert Hyatt
: To continue with part of a conversation with Bob Hyatt, who reponded to
: my post :
: So, I say it again, I see *nothing* that says that we must analyze a
: arrives? Not sure. But a *lot* sooner than "never"..
: I would say three things. This extension of computer
: performance may not be linear.
I wouldn't argue that point at all. However, by the same token, there
is nothing that suggests that it isn't linear either. It could go either
way.
: There is no evidence that simple brute force approaches will
: solve pattern oriented problems.
same point. No evidence that it will nor won't. But there's also nothing
to say that chess has to be solved by pattern-oriented methodololgies.
That's been the problem with A.I. for 30 years now... not knowing how the
human "works" there is little chance that we are going to be able to emulate
that "approach"...
: A second point is that this view does not seem to be shared
: (perhaps naturally) by the GM community.
I'm not sure what to make of that. Perhaps a little poor-mouthing to
help "save-face" when the loss happens. IE, from experience, the ACM
chess tournaments were a *lot* more fun before 1982 or so, when we
were not expected to win. Once we were expected to win, a win meant
nothing, but a loss meant everything. I think GM's have this same
mental outlook and would prefer to be the "underdog" in case they do
get beat, but hope they look like geniuses when they win easily...
: Thirdly - and in my long initial post to this subject I
: perhaps burried it - is the following: We agree, do we not,
: that current playing strength is sub-2000. (Human players are
: not allowed instantaneous access to massive opening libraries,
: so lets make it even). I know from the human level of playing
: that the conditions that support human level play at 19xx have
: nothing to do with the kind of knowledge necessary to play at
: 2300.
I don't buy that low a rating, no. I might agree that chess programs
have holes in their "knowledge" that a good 2000 player doesn't. But to
offset those holes, they have tactical skills that a 2000 player can only
dream about.
It is difficult to try to "calculate" a rating for a program, when, as a
GM will tell you, they have the positional understanding of X but the
tactical accumen of Y (X and Y are two approximate Elo ratings). What is
more interesting is what is the effective performance level of such a
program? And the answer is a wildly variable number. IE I have personally
watched Crafty outplay two GM's in what should be drawn endgames, at long
time controls, and it looked like a 2600 player itself. And then I have
watched the *same* program fall for a long-term kingside attack because it
simply didn't realize, until too late, what was happening. But in a single
game, I would *never* bet anything important against the computer, because
that may well be the game where it shines like a 2600 player. That's why
I use the 2400 number repeatedly as a good estimate of programs... you will
see 2600 games and 2200 games. Against a GM, you won't see anywhere near that
dramatic type of fluctuation in playing skill..
: I do not want to say too much, or be too spectacular. Am
: enjoying the conversation. Be glad to hear your response.
: Cordially, phil.
The discussion is fun, so long as it stays fun. I enjoy them, and only
hope that others will join in, without starting the "mine is bigger than
yours" type of arguments when comparing GMs to computers. The main thing
I can offer is a mixture of 30 years of programming computers to play chess
along with several years of watching Crafty (and other programs like Ferret,
etc) vs GM players on ICC. It is illuminating...
Robert Hyatt
: Phil Innes wrote:
:> I would say three things. This extension of computer
:> performance may not be linear.
: Definitely agreed. I think it was Monty Newborn (I'm away from my books
: at present) who yonks ago published a prediction of the diminishing
: returns of brute force gains. I remember a Belle (?) chart comparing
: Belle at ply X+1 playing against Belle at ply X, as X increases. That
: prog
: vs same prog with slower hardware / less time mathematically tends to
: produce exaggeratedly good results for the first one (because of
: guessing efficiency if perm brain is on, also other reasons) is not
: relevant here as the differences
: get pretty small as X got reasonably big.
Note that he published a paper later that was more interesting in
that we ran some deep tests with crafty, and found that even going
from 13 to 14 plies deep, Crafty would change its final move in
over 15% of the cases we tested. Which would suggest that the
"diminishing returns" case might happen, but at a shallower slope
on the decrease than some might believe.
:> There is no evidence that simple brute force approaches will
:> solve pattern oriented problems.
: I probably agree (while brute force solves everything eventually,
: resources are very finite and solutions that take 10^(10^(10^(10^10)))
: years are not solutions).
: Could you give an example, please, of such a problem so I understand
: what "pattern based" means here. I use "pattern-based" as in weightings
: given in one program I am familiar with (maybe the technique is used
: commonly) to positions with certain pawn patterns, or even pawn-piece
: patterns. Clearly you refer to something else.
: "Theme" based I understand - try:
: W: Kd6, Ne4, a2, c5, f6, h6
: B: Kg8, Rh8, Bg2, c6, f7, h7
: with BTM - of course an easy win for black.
: BUT not for comps. See how long a computer takes to realise ...Bxe4
: sucks!
: {thanks to Mr Reid for this position}
This is not a good example. It shows something that is a problem, but I
have a solution for this: I would never play into this position in the
first place. IE a case where some smoke and errors would hide such a
problem by simply avoiding such positions where the king is trapped
with the rook trapped. If I don't let my program reach such positions,
then it is not really necessary that I make it understand them as well.
: This is a really excellent problem...
:> A second point is that this view does not seem to be shared
:> (perhaps naturally) by the GM community.
: :-)
: On the chess servers, at 5 0, (human) GMs still rule despite automated
: 533MHz Ferret and Quad-P6 crafty.
This I don't see. I took every player (GM) that crafty has played,
and ran a summary, and there is not a single one that is winning more
than one of every 4 games. "kc" came the closest if you factor out
bullet, but if you take all fast games, he was in the same percentage.
:> Thirdly - and in my long initial post to this subject I
:> perhaps burried it - is the following: We agree, do we not,
:> that current playing strength is sub-2000. (Human players are
:> not allowed instantaneous access to massive opening libraries,
:> so lets make it even).
: also elsewhere
:> Chess playing engines perform at under 2000, as Expert B.
: They have not made any contributions (relatively) to
: opening theory. It is the "stack" of opening theory which
: allows them to achieve higher apparent ratings.
: Please clarify. Are you saying that playing with no opening book,
: currently the better programs playing on good hardware play sub-2000?
: That is the way most will read it, and may indicate why it has not been
: responded to.
: I would _strongly_ take issue with such an assertion (if that is your
: assertion). Its falsity or truth could be easily established by playing
: with books off. Starting from middle or end game positions could also
: be tried (Nunn-type testing). I would guess that books-off has *at
: most* a 200 point rating effect pretty much across the range of playing
: speeds. My best guess would be 120 ELO. Less at very fast chess.
Or by examining a lot of games where an IM plays a3 and h3 as the first
two moves. They generally lose the first few such games, and only play
this to get a repeatable game since the book and its randomness is not
available. Then they try to improvise a plan that can win, taking many
games to perfect this, and they have success against some programs. Crafty
and Ferret both have "position learning" that tends to make this not work,
but it is a plan. Removing the opening book is not nearly so big a problem
as some might guess. It saves times for sure. But many versions of Crafty
will play the Ruy right down a main line, with no book at all. Others will
play the Guico, or whatever, depending on black's move.
: So, please clarify. :-)
:> I know from the human level of playing
:> that the conditions that support human level play at 19xx have
:> nothing to do with the kind of knowledge necessary to play at
:> 2300.
: Agreed (and assuming you are speaking as a >2200 player here).
: However, the jump from application of theory to humans, to its
: application to computers, is not really possible. To play well, even
: superbly, and to keep getting better indefinitely, computers *need not*
: learn a thing from human or GM play. That is not to say that they could
: not learn useful things, or improve even further - just that they do not
: need to. As an extreme, IMO a black box environment could be developed
: - fully automated, just involving comp vs comp games and auto-tuning,
: which kept on delivering advantages (small ones, but measurable, and
: every few months or years). The resulting chess may not seem like
: strong human play - but in absolute terms it would be stronger. Though
: far far from perfect, of course, always.
: Kind regards
: fca
Robert Hyatt
: Bob, I can see that you have several other posters going on at some
: length in this same thread. I should like to say that we have (so far)
: been able to disagree on a subject without calling each other names or
: disparaging the other's point of view. Lets return to our conversation
: at length... neither of us need to make some urgent point in the next
: ten minutes.
: Is this a fair characterisation of where we leave it (or you say it),
: for your point of view:
: "Brute force machine efficiencies and programming finesse will continue
: to develop chess engines, as they have for the past ten years until they
: achieve parity with GMs on a regular basis."
I would only add "there is nothing that says this won't happen." I would
not say that it *will* happen... subtle difference, but it's an important
issue to be because I *don't know* what will happen... I only hope...
: and mine: "there is a limit (as for human players) in linear
: development, which we might now be nearing, and, GMs behave in a
: different way than 1900 players, and are invested in a radically
: different pattern-dynamic to achieve this differential."
: Cordially, Phil
--
I agree with your statement as well... and for those two points, don't
see where they are mutually exclusive...
Robert Hyatt
: Robert Hyatt wrote:
:>
:> : Phil Innes wrote:
:>
:> :> I would say three things. This extension of computer
:> :> performance may not be linear.
:>
:> : Definitely agreed. I think it was Monty Newborn (I'm away from my books
:> : at present) who yonks ago published a prediction of the diminishing
:> : returns of brute force gains. I remember a Belle (?) chart comparing
:> : Belle at ply X+1 playing against Belle at ply X, as X increases. That
:> : prog
:> : vs same prog with slower hardware / less time mathematically tends to
:> : produce exaggeratedly good results for the first one (because of
:> : guessing efficiency if perm brain is on, also other reasons) is not
:> : relevant here as the differences
:> : get pretty small as X got reasonably big.
:>
:> Note that he published a paper later that was more interesting in
:> that we ran some deep tests with crafty, and found that even going
:> from 13 to 14 plies deep, Crafty would change its final move in
:> over 15% of the cases we tested.
: That is high. Was the series monotonic decreasing, or were there blips
: (if so, was this a blip)?
I am at home, paper is at the office, but 15% is still pretty high for
13->14 plies... I don't remember the highest percent change, offhand,
however... but I suspect someone reading has the JICCA article handy
and coule post the data...
: "Change move" of course is interesting, but:
: (a) if the positions used were crafty v crafty, they might be the sort
: where long tactics abounded (hence the change) - also, what game stages
: did the positions involve? (Basically, what was the data set, please?
: You know how I started KK69 ;-) ).
The positions came from Kasparov vs Deep Blue, and were taken from
all the games played by DB vs Kasparov and others (in the newborn
book on the match), every 7th move, 347 total positions tested...
: (b) "change" did not mean "get better" 100% of the time. IMO more
: conclusions can be drawn from the "who won" test (but then I am a
: Fischer-materialist)
No. Change didn't mean "get better" or "convert even to won game" or
anything like that.. just that if you give it more time, it will choose
a different move. Most positions were searched to depth=15 and many took
over 24 hours to search. The paper cut off at 14 because all were not
searched to 15 plies...
:> Which would suggest that the
:> "diminishing returns" case might happen, but at a shallower slope
:> on the decrease than some might believe.
:>
:> :> There is no evidence that simple brute force approaches will
:> :> solve pattern oriented problems.
:>
:> : I probably agree (while brute force solves everything eventually,
:> : resources are very finite and solutions that take 10^(10^(10^(10^10)))
:> : years are not solutions).
:>
:> : Could you give an example, please, of such a problem so I understand
:> : what "pattern based" means here. I use "pattern-based" as in weightings
:> : given in one program I am familiar with (maybe the technique is used
:> : commonly) to positions with certain pawn patterns, or even pawn-piece
:> : patterns. Clearly you refer to something else.
:>
:> : "Theme" based I understand - try:
:>
:> : W: Kd6, Ne4, a2, c5, f6, h6
:> : B: Kg8, Rh8, Bg2, c6, f7, h7
:> : with BTM - of course an easy win for black.
:>
:> : BUT not for comps. See how long a computer takes to realise ...Bxe4
:> : sucks!
:> : {thanks to Mr Reid for this position}
:>
:> This is not a good example.
: Not good for comps, sure ;-))
: The theme here is "Don't get your pieces trapped forever" which most
: beginners grasp real quick.
: The point is "Forever takes computers forever to grasp", or "Never is
: never grasped", take your pick.
: I suggest most players over say 1200 will not make the blunder, given a
: few seconds to think about things. But AFAIK ***NO*** programs solves
: this correctly, period. Please enlighten if I am wrong!
You are right... but the point is that my evaluation code would not allow
the position to occur in the first place. It would probably sac a pawn to
not get trapped into that position... giving the choice of getting trapped
or not winning a piece is an interesting test... but I don't believe it is
possible to force Crafty to accept such a position, ever...
:> It shows something that is a
: B-I-G
Here we differ. I consider something big if it causes problems. In 500
games, I don't see crafty reach a position where its rook is at h8 and it
moves the king to f8... maybe 1 time in 500... and it works like the devil
to move the king to let the rook out, becoming quite single-minded. I see
humans have to play kf1 with the rook at h1 more than crafty... Ergo, what
I don't see happen, is not a big problem.
I can show you positions where a human has no chance... pieces hanging all
over the board, overloaded pieces, pinned pieces, multiple threats, etc. And
a human would utterly die in such a position. Is that a big problem? Only
if the human gets pushed into such a position. I don't know of any that would
allow it to happen, they would start exchanging to keep it managable... and
then it is not a problem at all...
:> problem, but I
:> have a solution for this: I would never play into this position in the
:> first place. IE a case where some smoke and errors would hide such a
:> problem by simply avoiding such positions where the king is trapped
:> with the rook trapped.
: Pattern-based?
In a sense... it is handled two ways... a pattern with the rook at
h1, king at g1 or f1, or the rook at h2, king at f1,g1 or h1 (with the
pawn at h4 and g3 still there of course). But also by the rook having no
mobility, not being on an open file, etc...
:> If I don't let my program reach such positions,
:> then it is not really necessary that I make it understand them as well.
: You could only prevent it (as opposed to tweaking the probability
: slightly downwards) by hitting its eval significantly. You then get the
: usual problems - hit the eval enough, and it starts rejecting lines
: giving this in favour of other ones even when it shouldn't, or it avoids
: it even in good positions where it was irrelevant.
This may happen... can't comment until I see it affect a game, however,
and can't really imagine where a rook locked out of the game would not be
important and would be ignorable... unless you are already lost, which I
naturally hope doesn't happen...
:
:> :> A second point is that this view does not seem to be shared
:> :> (perhaps naturally) by the GM community.
:>
:> : :-)
:>
:> : On the chess servers, at 5 0, (human) GMs still rule despite automated
:> : 533MHz Ferret and Quad-P6 crafty.
:>
:> This I don't see. I took every player (GM) that crafty has played,
:> and ran a summary, and there is not a single one that is winning more
:> than one of every 4 games. "kc" came the closest if you factor out
:> bullet, but if you take all fast games, he was in the same percentage.
: I never said all or most GMs were better. I said GMs rule, i.e. it is
: GMs who are RIGHT at the top. And these GMs do not play crafty or
: Ferret often or at all - the set you examined is not the set of all
: players. They "rule" nonetheless. Their ICC grading says it all.
: Consider the 5 0 record of Vadik and Yotam and a couple of others
: (Yasser of late maybe). Do you think that (assuming they were playing
: seriously - Comps play seriously all the time) either crafty or ferret
: would score 50% or more against "them" ( ;-) ) at 5 0?
Yes I do... because I know too many GM players... and, at least against
the computers, I don't know of a one that takes it "casually". They are
about as serious and blood-thirsty a group as I have seen.
An example: several months ago I was talking with Roman about ways to
remain "anonymous" on ICC so he can play without getting bombarded with
questions, comments, etc. I suggested knocking off with the 3000+ ratings
because everyone "knew". He said "good idea"... he played a lower-rated
IM, and said "watch this." He blew a piece intentionally. But as I watched,
something "changed". And suddenly the GM started playing chess, and won the
game anyway. I asked "what in the world was that about?" He said something
like "I got caught up in the game, and forgot to lose." I see that with all
of them... so based on my personal conversations, I don't think any take it
"casually" at all...
Both Bruce and I have gotten comments from the GM players like "at bullet
this is totally impossible..." "at blitz it is simply too strong to play
anymore" and so forth. Those are private comments, so I don't think it is
any attempt to save face or anything...
I could be wrong... but haven't seen anything that would suggest it based
on direct conversations...
Would be nice to have a 5 0 tournament with Crafty/Ferret/WchessX/ZarkovX
as a team, and each GM player plays 4 games against each, and see how these
programs do against serious blitz players. I think it would be a total
blow-out myself, since we have all done terribly well against GM players
even at game/30 action chess...
:> :> Thirdly - and in my long initial post to this subject I
: One could say "position learning" involves long-term storage of big
: datasets (just like with an opening book) and should therefore be ruled
: out of a "chess engine only" type discussion. I'd go along with that.
I don't buy that, and don't want to start that argument again. The computer
is a processor, memory, cache, disk drive, network interface, etc, all rolled
into one. To lop off the computer's disk drive, I'd agree to, but we have to
lop off the human's head. They are not equivalent "entities". They will
*never* be equal entities. And any attempt to make the contest "equal" is
doomed to failure because there is *no* common ground between a man and
machine, other than the 64 squares and the 32 pieces that occupy those squares,
and the rules that define the game...
: I would permit you to draw conclusions from your datasets and put
: "proper" algorithms intp the code tried to avoid pifalls. By "proper" I
: EXCLUDE specific conditional avoidance of very large scale patterns (in
: the crudest sense, IF posn == <matrix>, play X). Not that you would put
: such "cheats" in anyway, because you'd have loads of them percolating
: into inner loops and the effect on speed would negate any conceivable
: advantage.
This would be pointless. As a human I know many opening lines 30 moves
deep. How to take that from *me* if you take it from the *computer*?
that's a pointless argument...
: The old Scisys Chess Champion Mk V and VI event went a little way down
: the Ruy with book off, provided (if I remember right) you gave them 1 e4
: to start with (else 1. Nf3 would be chosen - "pieces before pawns" ;-))
: ). And it had no position learning, of course!
:
: So, Phil, please clarify.
Phil Innes
understand
what "pattern based" means here. I use "pattern-based" as in weightings
given in one program I am familiar with (maybe the technique is used
commonly) to positions with certain pawn patterns, or even pawn-piece
patterns. Clearly you refer to something else.
I think I could. But you are not applying to a very valuable
source. There is a bloke in St. Petersburg who is a professor
of computational physics who would be better if you wished to
talk physics. Otherwise there are interesting patterns in
geometrical mathematics which could be discussed in plain
language.
But this particular patterning in my ref. is "yet something
other" (sounds evasive doesn't it, it is! the subject is damn
complicated, and good to start from some mutually understood
base). I am also parking this stuff for a while, Bob is
engaged. I am digesting Rolf. (phil)
Here is another view of the subject:
The white pieces and the black pieces
seemed to represent Manichean divisions
between light and dark, good and evil,
in the very spirit of man himself.
G. Kasparov.
Daniel Kang
>And partly it comes from a natural tendency to believe and accept the words
>of 'experts'. Most here are casual readers, who do they believe ? Lacking
>background technical knowledge in a very confused field, they tend to
>respect programmers output, titled persons output; and, in an argument
>situation, if they are going to take sides, its much safer to side with
>those with the right sort of wrapping paper for their output.
It's obvious that people with more experience tend to be more correct.
And when you're on the losing side of the argument, it's natural for
you to assume that you're not getting support because "they are
perceived as experts and I'm not" rather than just that you're WRONG.
>It has been pointed out [danger, rgcc off-topic word coming] that this leads
>to a kind of technical fascism, with masses and 'expert' leaders. Made all
>the worse when the experts are wrong, or lie, or are playing games, or have
>axes to grind. I think Rolf saw this, and in his own way, fought it. The
>rest you know.
Of course we all should take your words for who's right and wrong?
>So I got better on this plateau. Could do just fine with strong players. But
>GMs and strong IMs would slaughter me; and I had no idea why. I'ld have good
>positions and lose. Couldn't make it go my way. I'ld listen to them say 'go
>here, this, go here, and what can black do?'; and I'ld think, but black is
>better (additive simplistic eval function). But black would lose. How ?
>Black has better position. But the key was 'what can black do?'. It takes
>some time to grasp that even in a 'better' position, you can still be lost;
>because 'what can you do?' Get that idea, and it is another plateau, the
>plateau that says, let the other guy get a 'better' position that he can do
>nothing with and then loses (Anand-Rebel).
Someone, for some odd reasons, believes that one could have a
better position when he's facing an inevitable loss. Here, you
are simply criticizing the accuracy of evaluation functions,
nothing more. Plus humans are FAR more likely to miss the
tactics that totally change their perceptions of position than
computers are. Many GMs considered their positions as superior
when they were dead lost.
>The idea that all that positional
>additive evaluation can be rendered meaningless.
Then substantiate your claims, and/or give us your vision of how
positional evaluation should be done. So far your argument, if
such thing exists at all, has been based on speculation *only*.
>My point ? Is that the guys on 'positional is long-range tactics' level of
>understanding don't understand or know or have any idea of the higher
>concepts (not claiming I do, btw) necessary to evaluate positions. They
>think it is a smooth slope, when it isn't. They think more additive
>knowledge, more depth is the solution. IMO it isn't. It is not how the thing
>works.
I have no idea what you're attempting to explain here, nor do I
think you know *how the thing works*
>a) benefit from low chess knowledge programmer - none.
>b) benefit from high knowledge programmer who can't describe or
>algorithmise - none.
>c) benefit from high knowledge programmer who can describe and
>algorithmise - large.
Give some examples of b) type programs and c) type programs.
>Quite. You need the explaining capability as well. Some of them can.
Some of them think they can but fail miserably. The thing is that no
one knows how GMs play chess, even GMs themselves. They just *play*.
>It count to get where 'we' are today. It counts for the current paradigm
>(dead-end street IMO). But it doesn't solve the holes [repeats Anand-Rebel
>RRvRBB example].
Again what alternatives do you suggest and what exactly are the
"holes"?
>Additive evaluation function, alfa-beta, nullmove, fast hardware, depth etc.
>can only get you so far. Fast, deep search is, IMO, a kind of unsatisfactory
>cheat, which, in any case, can be got round by application of human
>intelligence.
And any example of such "application of human intelligence"? Or is
this yet another BS talk?
>Wise men know that even a fool has something to say.
Every fool thinks he has something worthwhile to say, even if
he doesn't.
>On the other side he is an arrogant fool, spouting stupid opinions on
>everything from dismemberment of convicts to child-rearing. He makes out he
>knows everything about everything, PR's himself endlessly, any mention of
>Rebel (for example) will result in pages of stuff about how Crafty does this
>that and the other - by prolification of output he almosts creates a
>situation where no other program/programmer can get a look in. He certainly
>lies. He engages in untruthful character assassinations when the going gets
>tough. He invokes in others a desire to argue with him, to bring him down a
>few pegs, such that you can see the history of the newsgroup as a stream of
>arguments between Bob and a series of participants, all with the same urge.
This is a character assassination.
Phil Innes
leads neither to strong chess or to strong chess programs, IMO.
Agree. We are only trying to describe where we are now.
Intellectualising on anything is bloody sad, isn't it! Its as
if we don't know how to proceed. LOL.
define "pattern" algorithmically.
Lets not start flow-charting our approach.
I would like to do the Rolf-thing - read a few books, think it
over. Come back to this in a few days.
Excuse me for saying so, but this has been the most
interesting group of conversations in this ng for some time.
You ask if I am over 2200, yes.
Phil
Daniel Kang
> I would say three things. This extension of computer
> performance may not be linear.
But computer performance itself growing exponentially. So it
kind of evens out.
> There is no evidence that simple brute force approaches will
> solve pattern oriented problems.
There is no evidence that Chess is a pattern oriented problem
> Thirdly - and in my long initial post to this subject I
> perhaps burried it - is the following: We agree, do we not,
> that current playing strength is sub-2000. (Human players are
> not allowed instantaneous access to massive opening libraries,
> so lets make it even). I know from the human level of playing
> that the conditions that support human level play at 19xx have
> nothing to do with the kind of knowledge necessary to play at
> 2300.
Humans are allowed to memorize opening variations that they have
not made themselves. What is your point? Plus the existence of the
kind of knowledge you claim those higher rated possess is
debatable at best. Also from a theoritical point of view, chess is
a simple algorithmic problem. There's nothing 10^10-ply search cannot
solve. With absolutely no knowledge at all, 10000000000000000-ply
search can beat Kasparov and all of us.
Regards,
Daniel Kang
User77568
>Steve Blatchford wrote in message <6qs4se$qr5$1...@eros.clara.net>...
>But how did they get the 'right sort of wrapping' in the first place? Bob is
>*obviously* one of the best qualified posters in this group. He wrote
>Crafty, a clearly world-class program. He has a PhD (which does mean
>something, Rolf's and Seans protestations notwithstanding), and he is a
>professor. To me, these are 'credentials', not 'wrapping'. When you get on
>an airplane, I assume that you assume that your pilot has the right
>credentials. Would you be as trusting if the pilot didn't have the right
>'wrapping'? It doesn't mean that he can't fly, it's just that if he has the
>creds, then he probably can.
A pilot gets continual tests over the years. It's by far not the same with
academics. Too many possibilities to cheat.
>>It has been pointed out [danger, rgcc off-topic word coming] that this
>leads
>>to a kind of technical fascism, with masses and 'expert' leaders. Made all
>>the worse when the experts are wrong, or lie, or are playing games, or have
>>axes to grind. I think Rolf saw this, and in his own way, fought it. The
>>rest you know.
>>
>Rolf rarely talked about technical issues. He accused Ed of stealing, and
>Bob of being a fascist. How are these accusations related to a technical
>disagreement with one of the experts. You mentioned 'ad hominem' attacks
>before? Rolf was *the* major progenitor of them.
Last time I checked the record Rolf did not "accuse Ed of stealing". It was
character assassination. It was Ed insulting Rolf. Go figure.
What is a fascist? Is he gasolining human bodies? Does he censor people without
a cause on CCC? Or do you think of euthanasia or even abortion=murder after
birth? All ideas of "Bob". Go figure.
Rolf's arguments against such ideas and actions are 'ad hominem' attacks for
you?
Regards
Michael Chung
> Wow ! That's a very clever teaching method ! Your class seems in total
> thrall of your greatness ! Very impressive your majesty !
Well, I found it impressive.
> Still no answer to how to deal with the sort of problem demonstrated in
> Anand-Rebel RRvRBB, all programs say losing side is +2.5 pawns winning,
> because imbalanced material situation does not lend itself to assessment by
> simplistic additive materialisic evaluation function as expoused by
> gasp,gasp above. (Incedentally, Hyatt on ch 64, along with multitudes of
> other programmers all were quite convinced that Rebel was winning - as in my
> prog says +2.5, therefore I say +2.5, because I have no thought capability
> of my own)
Faster computers come with more storage space and faster access to this
storage. At the same time new ideas for algorithms are introduced. Taking
advantage of all this new stuff requires updated software to go with the
hardware. (This doesn't just apply to chess either.)
For example, increasing disk space allows endgame theory to be incorporated into
newer programs using endgame databases. Making some endgames a snap for
computers.
Making computers think like humans is an interesting idea. Of course, to
actually do this we would require... (guess). It's unlikely that we would want
computers to imitate humans. It would be much more difficult to get computers
to understand human 'aspiration' than to just solve chess.
Mike.
Michael Chung
> and mine: "there is a limit (as for human players) in linear
I'm afraid I also disagree with you. The node search speed improvement is
not linear, it's easily seen that it's exponential. And although there may
be a limit for computers, I think of the limit as where the result of
perfect play from every chess position is known, using brute-force or any
other algorithm. Only then is there no further computers can go--but until
then computers will continue to improve.
Mike.
Michael Chung
> performance may not be linear.
> solve pattern oriented problems.
Please forgive me but I feel a need to exercise what I learn in class:
In any position in chess there are a finite number of moves. So, a computer
can make the "best move" from any chess position in a finite amount of time.
There is evidence that computers keep getting faster at an exponential
rate. If this continues then the amount of time needed to find the best
move will very quickly approach zero. More specifically, the amount of time
required for this will eventually be less than 3 minutes.
This is enough to prove that given the assumptions, computers will
eventually play perfect chess.
mclane
>Last time I checked the record Rolf did not "accuse Ed of stealing". It was
>character assassination. It was Ed insulting Rolf. Go figure.
>What is a fascist? Is he gasolining human bodies? Does he censor people without
>a cause on CCC? Or do you think of euthanasia or even abortion=murder after
>birth? All ideas of "Bob". Go figure.
Oh no, not again. This is a computerchess newsgroup. Please discuss
your gasolining, euthanasia etc. stuff somewhere else. Maybe your wife
or your mum wants to listen about it. But not the readers of this
newsgroup.
>Rolf's arguments against such ideas and actions are 'ad hominem' attacks for
>you?
This is not point of the discussion here.
You have no clue what rolf wrote and what not.
spam, jam or post your off-topic stuff somewhere else.
>Regards
best wishes
mclane
Robert Hyatt
: "Christopher R. Dorr" <crd...@one.net> wrote:
:>Steve Blatchford wrote in message <6qs4se$qr5$1...@eros.clara.net>...
:>But how did they get the 'right sort of wrapping' in the first place? Bob is
:>*obviously* one of the best qualified posters in this group. He wrote
:>Crafty, a clearly world-class program. He has a PhD (which does mean
:>something, Rolf's and Seans protestations notwithstanding), and he is a
:>professor. To me, these are 'credentials', not 'wrapping'. When you get on
:>an airplane, I assume that you assume that your pilot has the right
:>credentials. Would you be as trusting if the pilot didn't have the right
:>'wrapping'? It doesn't mean that he can't fly, it's just that if he has the
:>creds, then he probably can.
: A pilot gets continual tests over the years. It's by far not the same with
: academics. Too many possibilities to cheat.
Not as many as you'd think. there are *plenty* of critics that look over
everything that is written and take instant offense at things that are
not verifiable... never been in that boat myself, never intend to ride
in it either...
:>>It has been pointed out [danger, rgcc off-topic word coming] that this
:>leads
:>>to a kind of technical fascism, with masses and 'expert' leaders. Made all
:>>the worse when the experts are wrong, or lie, or are playing games, or have
:>>axes to grind. I think Rolf saw this, and in his own way, fought it. The
:>>rest you know.
:>>
:>Rolf rarely talked about technical issues. He accused Ed of stealing, and
:>Bob of being a fascist. How are these accusations related to a technical
:>disagreement with one of the experts. You mentioned 'ad hominem' attacks
:>before? Rolf was *the* major progenitor of them.
: Last time I checked the record Rolf did not "accuse Ed of stealing". It was
: character assassination. It was Ed insulting Rolf. Go figure.
Backward. Rolf attacked Ed hundreds of times. Ed got tired of it and took
Rolf to court, which further incensed Rolf. That's all there was, continual
references to "the nazi criminal Schroeder"... over and over and over...
: What is a fascist? Is he gasolining human bodies? Does he censor people without
: a cause on CCC? Or do you think of euthanasia or even abortion=murder after
: birth? All ideas of "Bob". Go figure.
1. while I was a "moderator" (not a *censor*) we did not delete one single
post. We were accused of doing so by "Sean"... but we proved that his post
was still there, but an "operator error" on his part was not showing it. Not
our problem in giving lessons on operating netscape and a web site message
board. So on my watch, no one was "censored". The entire group of founders,
less a couple of us, voted to can Rolf. I was against it, so was Bruce and
Steve. The majority won. I'm now convinced it was the right action, albiet
for the wrong reasons, because he has repeatedly shown here that he can't stick
to "computer chess topics" for very long without going off into religion,
politics, and whatever else suits his agenda against Ed.
: Rolf's arguments against such ideas and actions are 'ad hominem' attacks for
: you?
"The Criminal Schroeder" is certainly an ad hominem. There were *many*
others...
: Regards
crd...@my-dejanews.com
> >*obviously* one of the best qualified posters in this group. He wrote
> >Crafty, a clearly world-class program. He has a PhD (which does mean
> >something, Rolf's and Seans protestations notwithstanding), and he is a
> >professor. To me, these are 'credentials', not 'wrapping'. When you get on
> >an airplane, I assume that you assume that your pilot has the right
> >credentials. Would you be as trusting if the pilot didn't have the right
> >'wrapping'? It doesn't mean that he can't fly, it's just that if he has the
> >creds, then he probably can.
>
> academics. Too many possibilities to cheat.
The point is credentials. Bob has them, Rolf and Sean don't. It's real hard
to cheat to get a PhD. I know enough pilots and PhD's to say that it's
probably a lot easier to cheat to get your pilot's license than it is to
cheat your way through your B.S., then your M.A., then your PhD. A PhD means
something. Period.
> >>It has been pointed out [danger, rgcc off-topic word coming] that this
> >leads
> >>to a kind of technical fascism, with masses and 'expert' leaders. Made all
> >>the worse when the experts are wrong, or lie, or are playing games, or have
> >>axes to grind. I think Rolf saw this, and in his own way, fought it. The
> >>rest you know.
> >>
>
> >Rolf rarely talked about technical issues. He accused Ed of stealing, and
> >Bob of being a fascist. How are these accusations related to a technical
> >disagreement with one of the experts. You mentioned 'ad hominem' attacks
> >before? Rolf was *the* major progenitor of them.
>
> Last time I checked the record Rolf did not "accuse Ed of stealing". It was
> character assassination. It was Ed insulting Rolf. Go figure.
You checked wrong then. Would you like some cited posts where Rolf accuses Ed
of stealing various things, including ChessBase databases?
> What is a fascist? Is he gasolining human bodies? Does he censor people
without
> a cause on CCC? Or do you think of euthanasia or even abortion=murder after
> birth? All ideas of "Bob". Go figure.
Rolfie? Is that you again? How's you get an AOL address. Anyway, all of these
things have been addressed before. If you are discounting Bob's opinion bcause
you call him a fascist, then that argument is *definitionally* ad hominem.
I've seen all this gibberish about Bob before on here, and the simple answer
is the same as it was before: It's just not true.
Exactly how does this work....you dislike Bob, and this somehow justifies Rolf
and Sean acting like morons? It's been a while since my logic courses in
college, but I'm pretty sure this doesn't fly <serious sarcasm>.
> Rolf's arguments against such ideas and actions are 'ad hominem' attacks for
> you?
Yes. Obviously. Of course. To any sentient being.
Is the above clear?
Chris
> Regards
>
-----== Posted via Deja News, The Leader in Internet Discussion ==-----
http://www.dejanews.com/rg_mkgrp.xp Create Your Own Free Member Forum
Robert Hyatt
:> I would say three things. This extension of computer
I don't believe hardware speeds are "exponential" geometric, perhaps,
but we are barely seeing a doubling every 2 years... which means we get
nearly an additional ply every 2 years, so long as this continues. I don't
believe we will see this doubling every two years much longer, because the
gigahertz barrier is a tough one, when electrical properties change quite
drastically as suddenly everything on the chip is an antenna for microwave
frequencies...
Chess is exponential, in the form W^D, where W is a known constant of
roughly 38, and D is the depth of the search... We don't have such
a relationship in hardware speeds... just the general 2x every 2 years
roughly...
Peter W. Gillgasch
> Oh no, not again. This is a computerchess newsgroup.
Exactly. I suggest that everybody should try not to get drawn
into obvious attempts of turning this into "who said what, why,
when to whom" again.
There are other places that could be used for that, much easier
to access those via AOL as well, especially if you are playing hyper
anon real chess openings and you are uncertain about the nature of chess
:)
The occasional joke and fair fist fight is ok, but please no primitive
mud throwing (I am always in favour of elegant mud throwing ;) and no
newsgroup history lessons... Let's move on.
-- Peter
Steve Blatchford
crd...@my-dejanews.com wrote in message <6r06kj$o0n$1...@nnrp1.dejanews.com>...
>In article <199808131631...@ladder01.news.aol.com>,
> user...@aol.com (User77568) wrote:
>> "Christopher R. Dorr" <crd...@one.net> wrote:
Sorry, but I (Steve Blatchford) protest. My name is attributed here, but I
wrote practically none of the quoted material.
>> >Steve Blatchford wrote in message <6qs4se$qr5$1...@eros.clara.net>...
>>
Although the indentation level suggests it, I did not write this.
>> >But how did they get the 'right sort of wrapping' in the first place?
Bob is
>> >*obviously* one of the best qualified posters in this group. He wrote
>> >Crafty, a clearly world-class program. He has a PhD (which does mean
>> >something, Rolf's and Seans protestations notwithstanding), and he is a
>> >professor. To me, these are 'credentials', not 'wrapping'. When you get
on
>> >an airplane, I assume that you assume that your pilot has the right
>> >credentials. Would you be as trusting if the pilot didn't have the right
>> >'wrapping'? It doesn't mean that he can't fly, it's just that if he has
the
>> >creds, then he probably can.
>>
Nor did I write this.
>> A pilot gets continual tests over the years. It's by far not the same
with
>> academics. Too many possibilities to cheat.
>
Nor this.
>
>The point is credentials. Bob has them, Rolf and Sean don't. It's real hard
>to cheat to get a PhD. I know enough pilots and PhD's to say that it's
>probably a lot easier to cheat to get your pilot's license than it is to
>cheat your way through your B.S., then your M.A., then your PhD. A PhD
means
>something. Period.
>
>
I did write this, even though the indentation marks don't add up to my name
at the top.
>> >>It has been pointed out [danger, rgcc off-topic word coming] that this
>> >leads
>> >>to a kind of technical fascism, with masses and 'expert' leaders. Made
all
>> >>the worse when the experts are wrong, or lie, or are playing games, or
have
>> >>axes to grind. I think Rolf saw this, and in his own way, fought it.
The
>> >>rest you know.
>> >>
>>
I did not write this.
>> >Rolf rarely talked about technical issues. He accused Ed of stealing,
and
>> >Bob of being a fascist. How are these accusations related to a technical
>> >disagreement with one of the experts. You mentioned 'ad hominem' attacks
>> >before? Rolf was *the* major progenitor of them.
>>
I did not write this.
>> Last time I checked the record Rolf did not "accuse Ed of stealing". It
was
>> character assassination. It was Ed insulting Rolf. Go figure.
>
>
Nor this, nor anything else that follows.
>You checked wrong then. Would you like some cited posts where Rolf accuses
Ed
>of stealing various things, including ChessBase databases?
>
Nor this.
>> What is a fascist? Is he gasolining human bodies? Does he censor people
>without
>> a cause on CCC? Or do you think of euthanasia or even abortion=murder
after
>> birth? All ideas of "Bob". Go figure.
>
Nor this.
>Rolfie? Is that you again? How's you get an AOL address. Anyway, all of
these
>things have been addressed before. If you are discounting Bob's opinion
bcause
>you call him a fascist, then that argument is *definitionally* ad hominem.
>
>I've seen all this gibberish about Bob before on here, and the simple
answer
>is the same as it was before: It's just not true.
>
>Exactly how does this work....you dislike Bob, and this somehow justifies
Rolf
>and Sean acting like morons? It's been a while since my logic courses in
>college, but I'm pretty sure this doesn't fly <serious sarcasm>.
>
Nor this.
>> Rolf's arguments against such ideas and actions are 'ad hominem' attacks
for
>> you?
>
Nor this.
Please be more careful.
Steve Blatchford
Steve Blatchford
Christopher R. Dorr wrote in message <35d20...@news.one.net>...
>
>Steve Blatchford wrote in message <6qs4se$qr5$1...@eros.clara.net>...
>
>>>>Yet again demonstrating the disastrous set of relationships within this
>>>>newgroup and wider.
>>>>
>>>>a) the general assumption that who is saying something is more important
>>>>than what is being said.
>>>
>>>When someone who has consistently demonstrated a complete lack of
>>>sensibility and substance speaks, their opinions necessarily carry less
>>>weight; i.e. when an idot has said ridiculous things 99 times in a row,
>the
>>>100th thing tends to be ignored.
>>>
>>
>>In the extreme case this would be reasonable. Except that here (rgcc and
>>ccc, computer chess in general, as well as chess itself) most posts are
>>pre-assessed by readers before they are read. You'll note endless
>references
>>to the writer of a post being this that or the other negative, rather than
>>to a direct addressing of the content. I'm taking Hyatt to task for this
>>elsewhere. IMO he's a particularly strong exponent of the ad hominem
>>argumentative style.
>
>
>I've been around this newgroup for four or five years...the only times I
>have seen Bob react in an ad hominem manner was when he had been previously
>personally attacked by the poster. In general, his style has appeared (to
>me) to be very reasonable. Are you stating that he should not 'give as good
>as he gets?'. That would appear to be a difficult to maintain position.
>
One could argue that in his position as group-god/guru he should be setting
an example.
What I am stating is that he often is responsible himself for the
instigation of 'personal' comments into threads. Not always, and not
exclusively, but often enough.
>
>>Partly this comes from chess itself, where there is a very strong pecking
>>order in value of output, based on elo. icc even has a filter based on
elo,
>>no ? This does have the advantage of keeping supposedly low-grade patzer
>>comments under control; but also creates unpleasant side effects. Here,
for
>>example.
>>
>>And partly it comes from a natural tendency to believe and accept the
words
>>of 'experts'. Most here are casual readers, who do they believe ? Lacking
>>background technical knowledge in a very confused field, they tend to
>>respect programmers output, titled persons output; and, in an argument
>>situation, if they are going to take sides, its much safer to side with
>>those with the right sort of wrapping paper for their output.
>
>
>But how did they get the 'right sort of wrapping' in the first place? Bob
is
>*obviously* one of the best qualified posters in this group. He wrote
>Crafty, a clearly world-class program. He has a PhD (which does mean
>something, Rolf's and Seans protestations notwithstanding), and he is a
>professor. To me, these are 'credentials', not 'wrapping'. When you get on
>an airplane, I assume that you assume that your pilot has the right
>credentials. Would you be as trusting if the pilot didn't have the right
>'wrapping'? It doesn't mean that he can't fly, it's just that if he has the
>creds, then he probably can.
Nice analogy. I have an analogy too. You're on a plane and the pilot has a
heart attack. So does the co-pilot. Now what ?
Elect the nearest person in blue uniform ? Possibly a good idea at the time.
But does this mean this person is now qualified to fly planes for evermore
wearing four gold stripes ?
BTW. For airplane read computer research AI lab. And for nearest person read
lab technician. For heart attack read got fed up, realised it was going
nowhere and had no application and went and did something more useful.
>
>>It has been pointed out [danger, rgcc off-topic word coming] that this
>leads
>>to a kind of technical fascism, with masses and 'expert' leaders. Made all
>>the worse when the experts are wrong, or lie, or are playing games, or
have
>>axes to grind. I think Rolf saw this, and in his own way, fought it. The
>>rest you know.
>>
>
>
>Rolf rarely talked about technical issues. He accused Ed of stealing, and
>Bob of being a fascist. How are these accusations related to a technical
>disagreement with one of the experts. You mentioned 'ad hominem' attacks
>before? Rolf was *the* major progenitor of them.
The behaviour of Rolf is one of life's great mysteries. You might even think
somebody else with another purpose was pulling his strings, given the almost
complete lack of logical reason for some (many?) of his attacks. However,
some of Rolf's attacks were with foundation. He certainly called Bob
something or other over the industrial processing into gasoline and
dismemberment of criminals issue. And, Bob has now effectively stopped this
off-topic opinion making of his. He learnt and is to be commended for that.
Shows that movement onwards is possible, no ?
>
>>>>
>>>>b) relatedly, the only acceptable viewpoints being those wrapped up in
>the
>>>>'correct' type of wrapping paper.
>>>>
>>>>c) relatedly again, the application by readers of their own choice of
>>>>wrapping paper (for condemnation or approval) when there is uncertaintly
>>as
>>>>to origin.
>>>>
>>>>d) relatedly again, the idea that every new poster must fall into one of
>>>the
>>>>
>>>>to some :-)
>>>>
>>>>but why look for the real gold when you have the fool's go(l)d already ?
>>>
>>>
>>>What on earth does this mean?
>>
>>Hyatt is the fool's god. My opinion. The gems of wisdom are to found
>>elsewhere. Often amongst his detractors.
>
>
>I disagree (as I believe the sizeable majority of newsgroup members would).
>Bob's ideas and explanations have greatly added to my understanding of
>computer chess.
I'll agree explanations. But not ideas. Those I don't see. Seriously.
>He has indeed passed many 'gems of wisdom'.
Yup. Teacher.
>Please find me
>something of a remotely similar value ever posted by Rolf or Sean.
>
Sean not. But Sean is a troll, a puppet being worked out of EST time zone by
somebody with some purposeful axe to grind.
Rolf generates a load of cantankerous, argumentative, rude crap. Sure. He
also generates intelligent prose with ideas and new ways of seeing the
world. And on-topic. He gets shat on whatever he does now, on-topic or off.
Not good IMO.
This paragraph is worthy of another thread. I will try and start it in a few
days (too busy elsewhere at the moment). Remind me, if you want, later next
week.
Improvement in elo of comps versus improvement in elo of humans. Yes. Talk
about that.
And the mentioning (sort of) above, similar to other comments in other
threads (Daniel? is it), of computer chess improvement by AI techniques as
against 'trying to make it think like a human'.
I think this dichotomy comes out of the bean-counter paradigm, suggesting
the ways are divergent and distinct. IMO it's a false dichotomy that
represents ideas in such a way as to hamper progress. I'll take this up next
week. And no, I' m not being evasive :)
Your evidence would be correct :) Often the biggest patzers have made the
'strongest' programs, sure. In the current bean-counter paradigm. True.
>
>>>>4. Thus the crop of current chess programmers (not just Hyatt) are way
>way
>>>>below GM strength. At 200 elo per step, that's 8 steps give or take. The
>>>>quantum at which they don't understand that they don't know that they
>>don't
>>>>know is just too large to even think about.
>>>>
>>>
>>>As you said above, chess may well be one of those activities that one
>>'does'
>>>rather than 'knows how to do'. If this is the case (and I'm not saying it
>>>is) what benefit would come from being a GM programmer who could play GM
>>>chess, and not describe (either in english or algorithmically) how it was
>>>done?
>>
>>None at all.
>>
>>a) benefit from low chess knowledge programmer - none.
>>
>>b) benefit from high knowledge programmer who can't describe or
>>algorithmise - none.
>>
>>c) benefit from high knowledge programmer who can describe and
>>algorithmise - large.
>
>
>Again, our GM/programmer can algorithmise (Is this really a word?! :) )
No. I made it up. But it's useful and you know what it means. Like
bean-counter. We know what that means too. New words have descriptive and
meaning-building capabilities that you should not knock. You bad person, you
:)
>how
>*he* plays chess. WIll this benefit his program? Doubtfuls, as we have
>already decided that the fundamental way each of them approaches and plays
>chess is very different.
I like use of phrase "we have already decided". I know you have. Built this
dichotomy out of the bean-counter paradigm. But it's a false dichotomy. More
next week.
>
>>
>>>
>>>>5. To know what a GM knows and these program/programmer combinations
>don't
>>>>(eg a whole massive lot), and to be able to explain it to someone else,
>>and
>>>>to be able to quantify it somehow, and *algorithmise it* requires the
>>>>programmer to be there himself. And they are not. They are way way off.
>>>
>>>
>>>Again, if a GM can't explain it, how could he/she program it?
>>
>>Quite. You need the explaining capability as well. Some of them can.
>>Although many chess books are make-work and fill-space, several actually
>>contain useful info, which implies these people do exist.
>>
>
>But even if the can explain it, how does it help their program?
By giving the capability to algorithmise the knowledge. If this results in
algorithmising of dynamic knowledge, or non-additive knowledge, then its a
big deal forwards over static, materialistic bean-counting.
Yeah, yeah, yeah. Bean counter paradigm. I know it well. You keep repeating
the dichotomy. False.
Bean-counter paradigm again. We will talk later.
Read the gasolining threads.
Read the dismemberment threads.
Read the baseball bat threads.
Read many of the Rolf threads.
Read the Jakarta threads.
And many more.
>He
>doesn't seem particularly self-promoting to me.
Well there we differ :)
>The reason the talk usually
>focus on Crafty is because the source and ideas are readily available. We
>can say 'Look in thisfile.c at line 120, and see how this workd. Can't do
>that with the commercial programs.
Very i n t e r e s t i n g, Dr Watson :)
And depends entirely what you want to talk about, what your approach is,
what paradigm you're in.
Programming knowledge ? Crafty useful. Look at source code and algos.
Or chess knowledge ? Commercials useful. Look at PVs and evaluations. Varied
eval functions. Varied knowledge elements.
The fact that the former talk is much more prevalent reflects something or
other, no ?
>
>Personally, I have a great deal of respect for Bob, and a great deal of
>appreciation for the knowledge that hes has given freely in this forum.
Passes on. Teacher. Let's make that clear.
>
>If Rolfie or Sean would like to contribute anything of meaning, perhaps our
>respect for them would increase. But for now, it seems that they are
>content to attack people personally, and try to garner attention for
>themselves by acting out.
Couldn't agree more. Although Rolf's latest was not wholly without value, I
thought. Even if Thorsten didn't like it.
Steve
Dan Homan
>
> I don't believe hardware speeds are "exponential" geometric, perhaps,
> but we are barely seeing a doubling every 2 years... which means we get
> nearly an additional ply every 2 years, so long as this continues. I don't
> believe we will see this doubling every two years much longer, because the
> gigahertz barrier is a tough one, when electrical properties change quite
> drastically as suddenly everything on the chip is an antenna for microwave
> frequencies...
>
> Chess is exponential, in the form W^D, where W is a known constant of
> roughly 38, and D is the depth of the search... We don't have such
> a relationship in hardware speeds... just the general 2x every 2 years
> roughly...
Of course this time relation is also exponential.
Speed ~ 2^(t/2)
Where t is the time in years, but I agree that this cannot keep up
forever. Even if clever people at IBM and elsewhere can continue to
push back the limits set by material properties, we will eventually run
up against the limits imposed by quantum mechanics and the speed of
light.
- Dan
>
> --
> Robert Hyatt Computer and Information Sciences
> hy...@cis.uab.edu University of Alabama at Birmingham
> (205) 934-2213 115A Campbell Hall, UAB Station
> (205) 934-5473 FAX Birmingham, AL 35294-1170
>
>
---------------------------------------------------------------------
Dan Homan d...@vlbi.astro.brandeis.edu
Physics Department d...@quasar.astro.brandeis.edu
Brandeis University ho...@binah.cc.brandeis.edu
---------------------------------------------------------------------
http://www.astro.brandeis.edu/BRAG/people/dch.html
---------------------------------------------------------------------
Robert Hyatt
: Hi Steve,
: Mike.
I would not want to even try to make a computer emulate a human. Would
you *really* want Crafty to stop playing chess, and post some off-topic
stuff here instead? :)
Think how many articles it could post per day if it can search 300K
nodes per second in chess. :)
Bob
mclane
>The behaviour of Rolf is one of life's great mysteries.
! :-))
> You might even think
>somebody else with another purpose was pulling his strings,
maybe he was paid to make noise here ? Maybe an enemy of ed paid rolf
to ruin ed's reputation by throwing lies here ?!?
>Rolf generates a load of cantankerous, argumentative, rude crap. Sure. He
>also generates intelligent prose with ideas and new ways of seeing the
>world. And on-topic. He gets shat on whatever he does now, on-topic or off.
>Not good IMO.
your comments do not reflect the truth IMO.
We TALKED with him in old times. about computerchess. Why ? because he
did not insulted anybody at this stage.
HIS own increase of insults has made it impossible to talk with him in
a civilized way. I remember e.g. that he called me by telephone and
have still his strange emails, to help him against Ed Schroeder.
If he would not have began to insult and start a campaign against ed
for no reason, we would not have changed out reaction towards him.
If you go in your pub and there is a newbie and first you talk with
him and next day you come again, he is again there, but suddenly
asking you strange things about your friends, and insulting them, than
it is up to you to further talk with him or not. You go somewhere else
and want you peace, but he follows. He cries in the whole pub: Ed is
xyz. One day the owner of the pub throws him out. but he comes back.
Now the owner of the pub calls the police. The judge decides that he
is not allowed to go into this pub again. Thats it.
If rolf posts on topic i am willing to discuss anything with him, as
far as it makes sense. I know bob would do the same, because it is his
spirit to talk with anybody too.
It is rolf causing these trouble. Because he insists of referring to
his old insults instead of coming with on-topic-post without insults.
We have shown great tolerance here. No other human beeing has ever got
so much attention. So he should be satisfied. That at least the whole
pub tried very hard to throw him out. If this is what he likes, ...
>>As I said before, I have never seen Bob act like an 'arrogant fool'.
>Read the gasolining threads.
>Read the dismemberment threads.
>Read the baseball bat threads.
>Read many of the Rolf threads.
>Read the Jakarta threads.
>And many more.
You should not try to jump in and attack Bob Hyatt here.
If you do so, you are just another of those guys, trying to make the
impossible happen. That is: get more reputation yourself by throwing
mud on Bob. This will not work. You should know this.
I am not bob's best friend. We had many heated discussions in which I
called him nasty words. I have a different point of view concerning
HOW of computerchess. But i have to admit that your statements against
him are not substantial and are only trying to copy others.
Please stop this.
>Couldn't agree more. Although Rolf's latest was not wholly without value, I
>thought. Even if Thorsten didn't like it.
It is not the whole text i could not live with, it is that he tries to
insult by putting in a massive block of on-topic his old insults.
This is not the thing i like, right. If he wants to talk about
computerchess he should do it, but without insults. Than we have no
problems to answer him "steve".
>Steve
best wishes
mclane
Christopher R. Dorr
Steve Blatchford wrote in message <6r14sa$vq$1...@eros.clara.net>...
>
>
>
>crd...@my-dejanews.com wrote in message
<6r06kj$o0n$1...@nnrp1.dejanews.com>...
>>In article <199808131631...@ladder01.news.aol.com>,
>> user...@aol.com (User77568) wrote:
>>> "Christopher R. Dorr" <crd...@one.net> wrote:
>
>
>Sorry, but I (Steve Blatchford) protest. My name is attributed here, but I
>wrote practically none of the quoted material.
>
Indeed, I did misquote Steve. I apologize. I got lost in a sea of quotes,
and responses, and quotes of responses, and so on.
>
>Please be more careful.
>
>Steve Blatchford
>
I will try :)
Thanks,
Chris
Robert Hyatt
: Christopher R. Dorr wrote in message <35d20...@news.one.net>...
This is possible... but you have to take such "in context". IE I
do not *ever* make personal comments to people here that have not first
made personal comments to me first. You might not see theirs, particularly
if you haven't been around this newsgroup for several continuous years, but
that doesn't mean comments weren't made earlier. I do not take "newbies"
to task, I don't insult them, I try to answer their questions, as do many
others.
But we do have a "few" that can earn a quick chastisement from me, because
they have done it dozens of times previously...
:>
:>>Partly this comes from chess itself, where there is a very strong pecking
nope... but I might ask "anyone flown a 747 before?" if not, "anyone
flown a jet of any kind before?" if not, "anyone flown anything at all
before?" Because any of those three types of people would be better to
trust than *nothing*...
: BTW. For airplane read computer research AI lab. And for nearest person read
: lab technician. For heart attack read got fed up, realised it was going
: nowhere and had no application and went and did something more useful.
:>
:>>It has been pointed out [danger, rgcc off-topic word coming] that this
:>leads
:>>to a kind of technical fascism, with masses and 'expert' leaders. Made all
:>>the worse when the experts are wrong, or lie, or are playing games, or
: have
:>>axes to grind. I think Rolf saw this, and in his own way, fought it. The
:>>rest you know.
:>>
:>
:>
:>Rolf rarely talked about technical issues. He accused Ed of stealing, and
:>Bob of being a fascist. How are these accusations related to a technical
:>disagreement with one of the experts. You mentioned 'ad hominem' attacks
:>before? Rolf was *the* major progenitor of them.
: The behaviour of Rolf is one of life's great mysteries. You might even think
: somebody else with another purpose was pulling his strings, given the almost
: complete lack of logical reason for some (many?) of his attacks. However,
: some of Rolf's attacks were with foundation. He certainly called Bob
: something or other over the industrial processing into gasoline and
: dismemberment of criminals issue. And, Bob has now effectively stopped this
: off-topic opinion making of his. He learnt and is to be commended for that.
: Shows that movement onwards is possible, no ?
your statement is *totally* wrong. If you find the "dismemberment" post,
you see that was a statement of outrage about someone talking about how
"inhuman" it is to exercise capital punishment. They picked the state of
Alabama, and a particular person that was just executed (at the time of
the post). I explained the crime, why electrocution might have been
"too easy a punishment" and ended with the "two bus example, explaining how
the parents of the victim of this criminal probably felt." It was stretched
far beyond that, just as was the "gasoline" issue. Which is typical of
Rolf... to repeat something said one time, for hundreds of replays...
:>
:>>>>
:>>>>b) relatedly, the only acceptable viewpoints being those wrapped up in
: Yup. Teacher.
Please go back and find *two* on-topic posts by him...
:>>
:>>>
:>>>>>
:>>>>>C. Filled by meaningless babble (Paradigm shift? Please...I thought
There is no "bean counter" paradigm. That's a false legacy mentioned by
a few, but completely inaccurate. It came from the "faster is dumber,
slower is smarter" discussions. When someone mentioned "rebel=smart/slow,
crafty=dumb/fast" and I pointed out "hmmm... rebel is considerably *faster*
than crafty" this discussion sort of went away mysteriously...
:>
:>
:>>For myself, I am quite sure this step, plateau, step, plateau development
That's why we study history. What has happened will likely happen again.
:>
:>>>>4. Thus the crop of current chess programmers (not just Hyatt) are way
:>way
:>>>>below GM strength. At 200 elo per step, that's 8 steps give or take. The
:>>>>quantum at which they don't understand that they don't know that they
:>>don't
:>>>>know is just too large to even think about.
:>>>>
:>>>
:>>>As you said above, chess may well be one of those activities that one
:>>'does'
:>>>rather than 'knows how to do'. If this is the case (and I'm not saying it
:>>>is) what benefit would come from being a GM programmer who could play GM
:>>>chess, and not describe (either in english or algorithmically) how it was
:>>>done?
:>>
:>>None at all.
:>>
:>>a) benefit from low chess knowledge programmer - none.
:>>
:>>b) benefit from high knowledge programmer who can't describe or
:>>algorithmise - none.
:>>
:>>c) benefit from high knowledge programmer who can describe and
:>>algorithmise - large.
:>
:>
:>Again, our GM/programmer can algorithmise (Is this really a word?! :) )
: No. I made it up. But it's useful and you know what it means. Like
: bean-counter. We know what that means too. New words have descriptive and
: meaning-building capabilities that you should not knock. You bad person, you
: :)
I haven't seen a definition of "bean counter" that is anything other than
a vague attempt at being insulting. Do you have a different definition, or
is it "more of the same"??
:>how
:>*he* plays chess. WIll this benefit his program? Doubtfuls, as we have
:>already decided that the fundamental way each of them approaches and plays
:>chess is very different.
: I like use of phrase "we have already decided". I know you have. Built this
: dichotomy out of the bean-counter paradigm. But it's a false dichotomy. More
: next week.
:>
:>>
:>>>
:>>>>5. To know what a GM knows and these program/programmer combinations
:>don't
:>>>>(eg a whole massive lot), and to be able to explain it to someone else,
:>>and
:>>>>to be able to quantify it somehow, and *algorithmise it* requires the
:>>>>programmer to be there himself. And they are not. They are way way off.
:>>>
:>>>
:>>>Again, if a GM can't explain it, how could he/she program it?
:>>
:>>Quite. You need the explaining capability as well. Some of them can.
:>>Although many chess books are make-work and fill-space, several actually
:>>contain useful info, which implies these people do exist.
:>>
:>
:>But even if the can explain it, how does it help their program?
: By giving the capability to algorithmise the knowledge. If this results in
: algorithmising of dynamic knowledge, or non-additive knowledge, then its a
: big deal forwards over static, materialistic bean-counting.
I have a good idea who is writing this now... :)
:>
:>>>
:>>>>6. The programmers know that the programs are lacking knowledge. Big
: Read the gasolining threads.
There was *no* thread. One article by me with this as a one line
comment at the end. Rolf made it a thread by repeating it over and over.
: Read the dismemberment threads.
Ditto. *one* post by me, with the two busses to make a point.
: Read the baseball bat threads.
This was started by "Elvis" I agreed.
: Read many of the Rolf threads.
Exactly...
: Read the Jakarta threads.
ditto...
: And many more.
:>He
:>doesn't seem particularly self-promoting to me.
: Well there we differ :)
:>The reason the talk usually
:>focus on Crafty is because the source and ideas are readily available. We
:>can say 'Look in thisfile.c at line 120, and see how this workd. Can't do
:>that with the commercial programs.
: Very i n t e r e s t i n g, Dr Watson :)
: And depends entirely what you want to talk about, what your approach is,
: what paradigm you're in.
: Programming knowledge ? Crafty useful. Look at source code and algos.
: Or chess knowledge ? Commercials useful. Look at PVs and evaluations. Varied
: eval functions. Varied knowledge elements.
: The fact that the former talk is much more prevalent reflects something or
: other, no ?
Hi Chris... grammar gives all away after a while...
:>
:>Personally, I have a great deal of respect for Bob, and a great deal of
:>appreciation for the knowledge that hes has given freely in this forum.
: Passes on. Teacher. Let's make that clear.
:>
:>If Rolfie or Sean would like to contribute anything of meaning, perhaps our
:>respect for them would increase. But for now, it seems that they are
:>content to attack people personally, and try to garner attention for
:>themselves by acting out.
: Couldn't agree more. Although Rolf's latest was not wholly without value, I
: thought. Even if Thorsten didn't like it.
: Steve
--
Steve Blatchford
Cameron Hayne wrote in message ...
>In article <6qs4se$qr5$1...@eros.clara.net>, "Steve Blatchford"
><sblat...@clara.net> wrote:
>
>>Additive evaluation function, alfa-beta, nullmove, fast hardware, depth
etc.
>>can only get you so far. Fast, deep search is, IMO, a kind of
unsatisfactory
>>cheat, which, in any case, can be got round by application of human
>>intelligence.
>
>(with current machines) fast or deep enough to produce consistent wins
>against the best human players.
Even this statement is not value-free. You imply that the goal is to 'win
against the best'. IMO this is a bean-counting goal. Other goals might be
interesting chess, fun chess, good analysis, surprising candidate move
generator and so on.
>
>But in calling it a "cheat" I think you betray your true feelings
Always :)
>- you seem to have something against this arguably mechanistic method
>just because it is mecanistic.
No. I'm against it because it's the product of an outdated, dead-end
paradigm that I don't approve of and I would like the others to move on from
in order to make progress. The term mechanistic (amongst others) I use in
order to give you a word-picture of part of what I mean.
>A "satisfactory" chess playing computer
>for you would seem to require using similar methods to those used by
>human chess players.
B i g topic. Next week, sorry.
>That's fine if that's what interests you but
>I think that a chess playing program should be measured by its results
>(i.e. wins, losses, draws) not by the way it obtains those results.
We know this. It's a way all those in the bean-counter paradigm think.
Measure the results, preferably zillions, put them in spreadsheet, draw
pointless statistical conclusions. Imagine this is science.
Just a clue: if you (not you, one is meant) are weak at chess then the
machine aided bean-counting methodology and programming faster, big
machines, fast code is what interests you (one). That's what you do. That's
what you talk about. Most programmers are weak chessplayers. Most testers
are weak chessplayers too, believe it or not. If the aforementioned wish to
create a fool's forum where they can talk low-level drivel to each other and
others in the same boat, with the ability to censor out anyone who tells
them the truth, so be it. But don't expect any progress like that.
And the way it obtains results (ie the way it plays chess) is IMO rather
important. But there you go.
Steve
>
>--
>Cameron Hayne (ha...@crim.ca)
>Centre de recherche informatique de Montreal
Steve Blatchford
Robert Hyatt wrote in message <6qsol9$onn$1...@juniper.cis.uab.edu>...
>Cameron Hayne <ha...@crim.ca> wrote:
>: In article <6qs4se$qr5$1...@eros.clara.net>, "Steve Blatchford"
>: <sblat...@clara.net> wrote:
>
>:>Additive evaluation function, alfa-beta, nullmove, fast hardware, depth
etc.
>:>can only get you so far. Fast, deep search is, IMO, a kind of
unsatisfactory
>:>cheat, which, in any case, can be got round by application of human
>:>intelligence.
>
>: I can see that it might be called "unsatisfactory" since it is not not
yet
>: (with current machines) fast or deep enough to produce consistent wins
>: against the best human players.
>
>: - you seem to have something against this arguably mechanistic method
>: just because it is mecanistic. A "satisfactory" chess playing computer
>: for you would seem to require using similar methods to those used by
>: I think that a chess playing program should be measured by its results
>: (i.e. wins, losses, draws) not by the way it obtains those results.
>
>of us use "cheat" as a verb, with one notable exception from times past.
>When you think about it, you might conclude that this post was written
>by someone you know, then the grammer was polished by someone *else* that
>you probably know, before being deposited here. I will tell you my
>suspicious off-line if you are interested. But it does "fit" well... :)
>
R
O
T
F
L
We must have a game of poker one day.
But make sure you bring loads of money with you ;-)
Steve
>
>: --
>: Cameron Hayne (ha...@crim.ca)
>: Centre de recherche informatique de Montreal
>
User77568
>In article <199808131631...@ladder01.news.aol.com>,
> user...@aol.com (User77568) wrote:
>> "Christopher R. Dorr" <crd...@one.net> wrote:
>> >Steve Blatchford wrote in message <6qs4se$qr5$1...@eros.clara.net>...
>>
>> >But how did they get the 'right sort of wrapping' in the first place? Bob
is
>> >*obviously* one of the best qualified posters in this group. He wrote
>> >Crafty, a clearly world-class program. He has a PhD (which does mean
>> >something, Rolf's and Seans protestations notwithstanding), and he is a
>> >professor. To me, these are 'credentials', not 'wrapping'. When you get on
>> >an airplane, I assume that you assume that your pilot has the right
>> >credentials. Would you be as trusting if the pilot didn't have the right
>> >'wrapping'? It doesn't mean that he can't fly, it's just that if he has the
>> >creds, then he probably can.
>>
>> A pilot gets continual tests over the years. It's by far not the same with
>> academics. Too many possibilities to cheat.
>The point is credentials. Bob has them, Rolf and Sean don't. It's real hard
>to cheat to get a PhD. I know enough pilots and PhD's to say that it's
>probably a lot easier to cheat to get your pilot's license than it is to
>cheat your way through your B.S., then your M.A., then your PhD. A PhD means
>something. Period.
1. How do you know about Rolf or Sean?
2. Bob. Somewhere I read that he finally "made" his PhD some 15 years after
his M.A.
3. Look at Bob's political output.
>> >>It has been pointed out [danger, rgcc off-topic word coming] that this
>> >leads
>> >>to a kind of technical fascism, with masses and 'expert' leaders. Made all
>> >>the worse when the experts are wrong, or lie, or are playing games, or
have
>> >>axes to grind. I think Rolf saw this, and in his own way, fought it. The
>> >>rest you know.
>> >>
>>
>> >Rolf rarely talked about technical issues. He accused Ed of stealing, and
>> >Bob of being a fascist. How are these accusations related to a technical
>> >disagreement with one of the experts. You mentioned 'ad hominem' attacks
>> >before? Rolf was *the* major progenitor of them.
>>
>> Last time I checked the record Rolf did not "accuse Ed of stealing". It was
>> character assassination. It was Ed insulting Rolf. Go figure.
>You checked wrong then. Would you like some cited posts where Rolf accuses Ed
>of stealing various things, including ChessBase databases?
Yes, of course.
>If you are discounting Bob's opinion bcause
>you call him a fascist, then that argument is *definitionally* ad hominem.
I checked that. Gasolining and dismembering of human beings wasn't brought up
by Rolf. Rolf criticised it. Fascist then was just a summary. Go figure.
Of course this is not about technical computerchess issues.
>I've seen all this gibberish about Bob before on here, and the simple answer
>is the same as it was before: It's just not true.
Gibberish isn't true. I agree.
>Exactly how does this work....you dislike Bob, and this somehow justifies Rolf
>and Sean acting like morons? It's been a while since my logic courses in
>college, but I'm pretty sure this doesn't fly <serious sarcasm>.
I fully agree here. The method is wrong. You dislike someone and then act like
a moron. I checked that. Rolf didn't follow that line. Go figure.
>> Rolf's arguments against such ideas and actions are 'ad hominem' attacks for
>> you?
>Yes. Obviously. Of course. To any sentient being.
>Is the above clear?
No.
--
Christopher R. Dorr
User77568 wrote in message
<199808141510...@ladder01.news.aol.com>...
Their behaviour and positngs have demonstrated a lack of knowledge that
would be inconsistent with posessing a PhD. If either of them has one, I
would like to know. Especially Sean, as he is supposedly 12 years old.
>
>2. Bob. Somewhere I read that he finally "made" his PhD some 15 years
after
>his M.A.
So what? He's got it.
>3. Look at Bob's political output.
Again, so what?
>>> >>It has been pointed out [danger, rgcc off-topic word coming] that this
>>> >leads
>>> >>to a kind of technical fascism, with masses and 'expert' leaders. Made
all
>>> >>the worse when the experts are wrong, or lie, or are playing games, or
>have
>>> >>axes to grind. I think Rolf saw this, and in his own way, fought it.
The
>>> >>rest you know.
>>> >>
>>>
>>> >Rolf rarely talked about technical issues. He accused Ed of stealing,
and
>>> >Bob of being a fascist. How are these accusations related to a
technical
>>> >disagreement with one of the experts. You mentioned 'ad hominem'
attacks
>>> >before? Rolf was *the* major progenitor of them.
>>>
>>> Last time I checked the record Rolf did not "accuse Ed of stealing". It
was
>>> character assassination. It was Ed insulting Rolf. Go figure.
>
>
>>You checked wrong then. Would you like some cited posts where Rolf accuses
Ed
>>of stealing various things, including ChessBase databases?
>
>Yes, of course.
Check out Deja News - look up 'Rolf Teuschen and stolen'. Please get back to
us on the results.
>>If you are discounting Bob's opinion bcause
>>you call him a fascist, then that argument is *definitionally* ad hominem.
>
>I checked that. Gasolining and dismembering of human beings wasn't brought
up
>by Rolf. Rolf criticised it. Fascist then was just a summary. Go figure.
Again, these attacks (no matter who brings them about) are definitionally
'ad hominem'. Do you understand this?
>Of course this is not about technical computerchess issues.
>
>>I've seen all this gibberish about Bob before on here, and the simple
answer
>>is the same as it was before: It's just not true.
>
>Gibberish isn't true. I agree.
>
>>Exactly how does this work....you dislike Bob, and this somehow justifies
Rolf
>>and Sean acting like morons? It's been a while since my logic courses in
>>college, but I'm pretty sure this doesn't fly <serious sarcasm>.
>
>I fully agree here. The method is wrong. You dislike someone and then act
like
>a moron. I checked that. Rolf didn't follow that line. Go figure.
Rolf didn't follow this line? You are joking, right? Rolf id guilty of the
most atrocious behaviour that I have witnessed in *any* newsgroup. Period.
>>> Rolf's arguments against such ideas and actions are 'ad hominem' attacks
for
>>> you?
>
>>Yes. Obviously. Of course. To any sentient being.
>
>>Is the above clear?
>
>No.
Which part didn't you understand?
Chris
Steve Blatchford
Robert Hyatt wrote in message <6r1fr0$95m$1...@juniper.cis.uab.edu>...
I think this is true. You do have an annoying tendency, however, to assume
abc is actually xyz, then rst, then pqr and react personally on abc because
you had a run in with rst/xyz/pqr some time previously.
So, you're not a bad man, Bob, ok ? Just a LOUD one :-)
>
>But we do have a "few" that can earn a quick chastisement from me, because
>they have done it dozens of times previously...
>
Just as long as it's not either with sticks on Billy or baseball bats on
anyone else, no problem.
And, remember, there's a difference between doling out a smack in the face
and banning someone, right ?
Lower bound = 0 :)))
Not so. And shut up about it forever more. You were busted to hell on this
one. Your comments were unbelievably stupid. No doubt they'll be used
against you for evermore by anyone who wants to get at you.
Just stop even trying to defend yourself. Admit you were a giant pillock and
never mention it again.
[ this space intentionally left blank for Bob to admit he was a giant
pillock ]
Am not your slave. But he did post yesterday a good one that Thorsten
promptly shat on.
Oh, yes there is. Next week I'll set it out exactly what is meant by the
bean-counter paradigm and why it is just not good enough. We can have a good
blast from you too, expectedly.
[ space intentionally left blank for Rolf to get historical around the
Mulberry bush ]
[ space intentionally left blank for the imposter to make a joke about Rolf,
historical and hysterical ]
Yes. See above.
The first reference to it's 'insulting' nature was made by Don Dailey on
Fool's Forum. But nobody took it up (possibly because they'ld been banned :)
Yes it is insulting. It does imply a deficit in chess knowledge on the part
of the practitioner. Arguably [space left here for Rolf ] it implies an
emotional/personality deficit too. But that's not my problem.
Bean-counter is an exceptionally accurate and descriptive term. We need
these terms.
Bob, Bob, Bob. You've been through so many names so far. Either you're d e a
d slow or just plain w r o n g.
Do you really want to open this one up again ?
Do you want to be reminded of your slow yet gradually escalating, repeated
posts, repeated over a long period, where you advocated death, industrial
processing into gasoline, dismemberment by greyhound bus, on murderers, then
rapists, then drug dealers.
It was an on-going murderous fantasy, played out here in front of our eyes,
including in front of someone who witnessed one of his comrades executed in
a truly barbarous manner by a fascist regime.
It was off-topic.
It was insane.
It was disgusting.
Now stop it. Stop trying to defend yourself. There is no defence. Shut up.
Apologise if you can.
>
>: Read the dismemberment threads.
Likewise.
>
>Ditto. *one* post by me, with the two busses to make a point.
Shut up. It (they, numerous posts) were utterly disgusting.
>
>: Read the baseball bat threads.
>
>This was started by "Elvis" I agreed.
As was this one.
>
>: Read many of the Rolf threads.
>
>Exactly...
On both sides. You were not an angel.
>
>: Read the Jakarta threads.
>
>ditto...
Likewise.
I think you've learnt plenty over the last years. As have many others. Let's
see ....
Steve
mclane
>Just a clue: if you (not you, one is meant) are weak at chess then the
>machine aided bean-counting methodology and programming faster, big
>machines, fast code is what interests you (one). That's what you do. That's
>what you talk about.
<Most programmers are weak chessplayers.
?!?
no.
>Most testers
>are weak chessplayers too, believe it or not.
no.
>If the aforementioned wish to
>create a fool's forum where they can talk low-level drivel to each other and
>others in the same boat, with the ability to censor out anyone who tells
>them the truth, so be it.
nonsense.
A tester has to find out the strength about chess programs. Which
methods he uses is not important. it counts how good his estimation
and judgement comes to the truth.
A critics of books does not have to be a good writer. He only has to
read much and know all the important books.
A mechanician of cars does not need to drive very good like
schumacher.
All workers work in a team. You have chess advisors, hardware
advisors, programmers, to create a strong program and to stand the
competition.
> But don't expect any progress like that.
Nonsense. Absolute nonsense. I have not seen any master-chess player
programming a chess program...
also i don'T know any strong chess player who is able to find out the
differences between engines very fast. This is something complete
different. Somebody who produces wine must not be able to drink much
wine, or he is not able to taste different wine qualitites. But he
produces the wine.
You get people specialized in any level, any part.
You don't get these old classic GOOD IN ANYTHING people.
>And the way it obtains results (ie the way it plays chess) is IMO rather
>important. But there you go.
So what ?! Tell me why rebel makes progress ? Why mchess makes
progress. Why genius has made NO progress ?!
Tell me why ?
Ossi and his advisors are strong chess players. Richard maybe not.
Why has Zarkov made big progress ?!
Why has tiger made enormousæ„Žrogress since november 97 ?!
Tell me: if your theory is right, christophe is a strong chess player
and tiger knows much, or ?!
IMO it is the teamwork that counts. only with the heart-work of many,
you get a better quality. It needs that people work with their heart.
>Steve
>>
>>--
>>Cameron Hayne (ha...@crim.ca)
>>Centre de recherche informatique de Montreal
best wishes
mclane
Tord Kallqvist Romstad
Yes.
>>Most testers
>>are weak chessplayers too, believe it or not.
>
>no.
Yes. :-)
Sorry, but this discussion is completely pointless if we do not define
the term "weak chess player".
Tord (a very weak chess player by any reasonable definition of the word)
Steve Blatchford
Tord Kallqvist Romstad wrote in message ...
>In article <6r1r23$4ul$1...@steve.prima.ruhr.de>, mclane wrote:
>>"Steve Blatchford" <sblat...@clara.net> wrote:
>>>Most programmers are weak chessplayers.
>>
>>?!?
>>no.
>
>Yes.
>
>>>Most testers
>>>are weak chessplayers too, believe it or not.
>>
>>no.
>
>Yes. :-)
>
>Sorry, but this discussion is completely pointless if we do not define
>the term "weak chess player".
One condition is output of statement "chess is tactics" or "positional is
just long-range tactics".
This is on too low a step to go beyond bean-counting. And also its
justification.
Steve
Robert Hyatt
: Dan Homan wrote:
:>
:>
:> > I don't believe hardware speeds are "exponential" geometric, perhaps,
:> > but we are barely seeing a doubling every 2 years...
: Exponential : k^x
: Geometric: x^k
Sorry, wasn't thinking... of course you are right. :)
The "Growth" is on a linear curve is what I meant.. IE two times
faster every two years, pretty steadily for many years now. Naturally
if you figure twice as fast in 2 years, 4 times faster in 4 years, 8 times faster in 6 years, 16 times faster in 8 years... it is growing... but the
"rate" is pretty well constant, and, I'm afraid, going to start to flatten
out once we get past the current crop of potential gigahertz machines. I
don't see how they work myself, but I've run on at least one, so unless
someone is full of voodoo, it does. somehow... but it defies all the laws
of electronics I learned during my HAM radio days... :)
: Therefore exponential by your own statement: maybe -ish if 2 is
: increasing. Geometric comes into it not at all, it is not a decayed
: case of exponential... (Don makes same point I believe) ;-)
:> > which means we get
:> > nearly an additional ply every 2 years, so long as this continues. I don't
:> > believe we will see this doubling every two years much longer,
: Agreed. It is already longer.
Not really longer... IE My P6/200 is just over 2 years old. We now have
P6/450 (if you count the Xeon as "here"...)
:> > because the
:> > gigahertz barrier is a tough one, when electrical properties change quite
:> > drastically as suddenly everything on the chip is an antenna for microwave
:> > frequencies...
: Optics provides the answers... and puts Kryotech out of business.
I'm not sure it provides *all* the answers. Some, yes, but it brings in
its own set of problems, and is still constrained by "C"...
:> > Chess is exponential, in the form W^D, where W is a known constant of
:> > roughly 38, and D is the depth of the search... We don't have such
:> > a relationship in hardware speeds... just the general 2x every 2 years
:> > roughly...
:
:> Of course this time relation is also exponential.
: "has been" for "is", then I agree. ;-)
:
:> Speed ~ 2^(t/2)
:> Where t is the time in years, but I agree that this cannot keep up
:> forever.
: It already isn't :-)
: there's enough empirical data already to force a change for t > c1995 to
: say (comparing say Jan 1st positions):
: Speed = k * 2 ^ ( t / (2 + z * (t - 1996) ) )
: Where t is year number and z > 0.
: If z << 1, the speed relationship is approximately exponential.
: To what extent z = z(t) instead of constant needs to be monitored.
: Where z is constant, Dan, the effect that t effluxing will have on the
: *denominator* is linear increasing, thus producing a decreasing rate of
: change on the ratio (the power).
: This is wholly intentional - if you are interested, I can expand on my
: reasons for this.
:> Even if clever people at IBM and elsewhere can continue to
:> push back the limits set by material properties, we will eventually run
:> up against the limits imposed by quantum mechanics
: itself advancing, though, in a most confusing and unpredictable way.
: Any limit therefore imposed is liable to be foundd at some point not to
: be a limit. If God is a Statistician, she has not yet revealed most of
: the rules to us.
:> and the speed of light.
: Tachyons notwithstanding, Don? ;-)
: IAE, this "eventually" is very far off IMO, and parallelization will be
: working really well by then...
:> - Dan
:> > Robert Hyatt Computer and Information Sciences
:> > hy...@cis.uab.edu University of Alabama at Birmingham
:> > (205) 934-2213 115A Campbell Hall, UAB Station
:> > (205) 934-5473 FAX Birmingham, AL 35294-1170
:> ---------------------------------------------------------------------
:> Dan Homan d...@vlbi.astro.brandeis.edu
:> Physics Department d...@quasar.astro.brandeis.edu
:> Brandeis University ho...@binah.cc.brandeis.edu
:> ---------------------------------------------------------------------
: Kind regards
: fca
: c/o Unnamed Departments of Physics, Mathematics and Computer Science
Robert Hyatt
: crd...@my-dejanews.com wrote:
:>In article <199808131631...@ladder01.news.aol.com>,
:> user...@aol.com (User77568) wrote:
:>> "Christopher R. Dorr" <crd...@one.net> wrote:
:>> >Steve Blatchford wrote in message <6qs4se$qr5$1...@eros.clara.net>...
:>>
:>> >But how did they get the 'right sort of wrapping' in the first place? Bob
: is
:>> >*obviously* one of the best qualified posters in this group. He wrote
:>> >Crafty, a clearly world-class program. He has a PhD (which does mean
:>> >something, Rolf's and Seans protestations notwithstanding), and he is a
:>> >professor. To me, these are 'credentials', not 'wrapping'. When you get on
:>> >an airplane, I assume that you assume that your pilot has the right
:>> >credentials. Would you be as trusting if the pilot didn't have the right
:>> >'wrapping'? It doesn't mean that he can't fly, it's just that if he has the
:>> >creds, then he probably can.
:>>
:>> A pilot gets continual tests over the years. It's by far not the same with
:>> academics. Too many possibilities to cheat.
:>The point is credentials. Bob has them, Rolf and Sean don't. It's real hard
:>to cheat to get a PhD. I know enough pilots and PhD's to say that it's
:>probably a lot easier to cheat to get your pilot's license than it is to
:>cheat your way through your B.S., then your M.A., then your PhD. A PhD means
:>something. Period.
: 1. How do you know about Rolf or Sean?
: 2. Bob. Somewhere I read that he finally "made" his PhD some 15 years after
: his M.A.
Incorrect. I finished my BS in 1970. I worked steadily until 1981 when
I started to work on my MS part-time. I finished that in 1983. In 1985
I left where I was working/teaching and went to school full-time to work
on my phd, which I finished in 1988, taking almost exactly 3 years. Not
sure where your data came from, but it's broken... And I didn't earn a
"MA" anyway... "MS" was the degree...
: 3. Look at Bob's political output.
???
mclane
>mclane wrote:
>> Ossi and his advisors are strong chess players. Richard maybe not.
Strange, i have seen him having problems to input genius moves ...
whatever...
>Czub, Thorsten?
right-brain + 2 * -2000x currywurst-hormons / pyramids * left brain !
I prefer playing Mau Mau or civilisation2...:-))
>Kind regards
>fca
best wishes
mclane
Robert Hyatt
: Robert Hyatt wrote in message <6r1fr0$95m$1...@juniper.cis.uab.edu>...
I don't know. I've never "banned" anyone in my life.
: Not so. And shut up about it forever more. You were busted to hell on this
: one. Your comments were unbelievably stupid. No doubt they'll be used
: against you for evermore by anyone who wants to get at you.
: Just stop even trying to defend yourself. Admit you were a giant pillock and
: never mention it again.
Never going to happen. I have my view on capital punishment. I have my
view on how a parent might react to horrible crimes against their children.
That won't change. I can live with my opinion. No one else has to.
"busted to hell" is a strong phrase. I don't consider it applicable in
my case. hard to bust someone for their firmly held convictions, because
it would mean I would have to change my mind. I haven't yet. On this
case, anyway...
: [ this space intentionally left blank for Bob to admit he was a giant
: pillock ]
<it remains blank>
:>
:>There is no "bean counter" paradigm. That's a false legacy mentioned by
:>a few, but completely inaccurate. It came from the "faster is dumber,
:>slower is smarter" discussions. When someone mentioned "rebel=smart/slow,
:>crafty=dumb/fast" and I pointed out "hmmm... rebel is considerably *faster*
:>than crafty" this discussion sort of went away mysteriously...
: Oh, yes there is. Next week I'll set it out exactly what is meant by the
: bean-counter paradigm and why it is just not good enough. We can have a good
: blast from you too, expectedly.
Feel free. If it's interesting I'll participate. If not, I'll ignore...
: Yes. See above.
: The first reference to it's 'insulting' nature was made by Don Dailey on
: Fool's Forum. But nobody took it up (possibly because they'ld been banned :)
: Yes it is insulting. It does imply a deficit in chess knowledge on the part
: of the practitioner. Arguably [space left here for Rolf ] it implies an
: emotional/personality deficit too. But that's not my problem.
: Bean-counter is an exceptionally accurate and descriptive term. We need
: these terms.
Hmmm... if "certain programs" are written with such "accurate and
discriptive ideas" they are "in trouble" from the get-go... :)
: Bob, Bob, Bob. You've been through so many names so far. Either you're d e a
: d slow or just plain w r o n g.
No longer wrong, I'm afraid.. :)
: Do you really want to open this one up again ?
: Do you want to be reminded of your slow yet gradually escalating, repeated
: posts, repeated over a long period, where you advocated death, industrial
: processing into gasoline, dismemberment by greyhound bus, on murderers, then
: rapists, then drug dealers.
Sure, because your memory is as faulty as some of your definitions. I posted
*once* about the "gasoline" issue, then had to re-explain it several times
since one poster liked to twist and cut to make it into something it wasn't.
Ditto for the bus post, which is much easier to find on Deja... And which
was started by "theDoDo" about Alabama executing some idiotic murderer...
: It was an on-going murderous fantasy, played out here in front of our eyes,
: including in front of someone who witnessed one of his comrades executed in
: a truly barbarous manner by a fascist regime.
: It was off-topic.
: It was insane.
: It was disgusting.
: Now stop it. Stop trying to defend yourself. There is no defence. Shut up.
: Apologise if you can.
I appologize when I figure I am wrong. In this case, the US still has a
death penalty. We are a democracy. Ergo, the majority want this still. I
am among that majority. End of story...
User77568
>>
>>>In article <199808131631...@ladder01.news.aol.com>,
>>> user...@aol.com (User77568) wrote:
>>>> "Christopher R. Dorr" <crd...@one.net> wrote:
>Their behaviour and positngs have demonstrated a lack of knowledge that
>would be inconsistent with posessing a PhD. If either of them has one, I
>would like to know. Especially Sean, as he is supposedly 12 years old.
I would say you should take usenet postings with the necessary grain of salt.
You should do your own reasearch. And let nobody think for you. I can't
understand why people believe that Sean is supposedly 12 years old.
He did never talk about. Last time I looked he did quote people who had
insulted him as such a schoolboy.
I read about you that you are a USCF master. Do you play computerchess? With
what program? Did you play in tournaments this year?
>>
>>2. Bob. Somewhere I read that he finally "made" his PhD some 15 years
>after
>>his M.A.
>So what? He's got it.
So what? That his PhD doesn't prove what you wanted to pretend?
>>3. Look at Bob's political output.
>Again, so what?
Yes, I begin to understand.
>>>> >>It has been pointed out [danger, rgcc off-topic word coming] that this
>>>> >leads
>>>> >>to a kind of technical fascism, with masses and 'expert' leaders. Made
>all
>>>> >>the worse when the experts are wrong, or lie, or are playing games, or
>>have
>>>> >>axes to grind. I think Rolf saw this, and in his own way, fought it.
>The
>>>> >>rest you know.
>>>> >>
>>>>
>>>> >Rolf rarely talked about technical issues. He accused Ed of stealing,
>and
>>>> >Bob of being a fascist. How are these accusations related to a
>technical
>>>> >disagreement with one of the experts. You mentioned 'ad hominem'
>attacks
>>>> >before? Rolf was *the* major progenitor of them.
>>>>
>>>> Last time I checked the record Rolf did not "accuse Ed of stealing". It
>was
>>>> character assassination. It was Ed insulting Rolf. Go figure.
>>
>>
>>>You checked wrong then. Would you like some cited posts where Rolf accuses
>Ed
>>>of stealing various things, including ChessBase databases?
>>
>>Yes, of course.
>Check out Deja News - look up 'Rolf Teuschen and stolen'. Please get back to
>us on the results.
Where are the cited posts? I told you before that I checked. And I didn't find
what you tried to pretend.
>>>If you are discounting Bob's opinion bcause
>>>you call him a fascist, then that argument is *definitionally* ad hominem.
>>
>>I checked that. Gasolining and dismembering of human beings wasn't brought
>up
>>by Rolf. Rolf criticised it. Fascist then was just a summary. Go figure.
>Again, these attacks (no matter who brings them about) are definitionally
>'ad hominem'. Do you understand this?
You don't understand. If a medical doctor tells you "that is cancer" he's not
attacking you ad hominem. If you don't understand that, then how do you think
you could judge if Rolf was correctly diagnosing Robert Hyatt's output as
fascist. I read Rolf's a psychologist and a doctor?
>>Of course this is not about technical computerchess issues.
>>
>>>I've seen all this gibberish about Bob before on here, and the simple
>answer
>>>is the same as it was before: It's just not true.
>>
>>Gibberish isn't true. I agree.
>>
>>>Exactly how does this work....you dislike Bob, and this somehow justifies
>Rolf
>>>and Sean acting like morons? It's been a while since my logic courses in
>>>college, but I'm pretty sure this doesn't fly <serious sarcasm>.
>>
>>I fully agree here. The method is wrong. You dislike someone and then act
>like
>>a moron. I checked that. Rolf didn't follow that line. Go figure.
>Rolf didn't follow this line? You are joking, right? Rolf id guilty of the
>most atrocious behaviour that I have witnessed in *any* newsgroup. Period.
Ok, let's not argue. Could you give cited posts at least for this theory here?
>>>> Rolf's arguments against such ideas and actions are 'ad hominem' attacks
>for
>>>> you?
>>
>>>Yes. Obviously. Of course. To any sentient being.
>>
>>>Is the above clear?
>>
>>No.
>Which part didn't you understand?
All those where you had no evidence.
User77568
>"Steve Blatchford" <sblat...@clara.net> wrote:
>>Couldn't agree more. Although Rolf's latest was not wholly without value, I
>>thought. Even if Thorsten didn't like it.
>It is not the whole text i could not live with, it is that he tries to
>insult by putting in a massive block of on-topic his old insults.
>This is not the thing i like, right. If he wants to talk about
>computerchess he should do it, but without insults. Than we have no
>problems to answer him "steve".
Thorsten, I read his long posting "Fast food...".
Where did Rolf hide new insults?
Please give us some examples. People often talk about Rolf but they can't give
proof for their judgements. Like Chris Dorr in this thread.
Michael Chung
> nearly an additional ply every 2 years, so long as this continues. I don't
> gigahertz barrier is a tough one, when electrical properties change quite
> drastically as suddenly everything on the chip is an antenna for microwave
> frequencies...
>
Okay... I agree that in the past we saw doubling every 2 years or so. Then don't
you agree that hardware speeds are exponential in the form: C*2^(T/2) where C is
the current hardware speed and T is the time relative to the present in years.
This is the same as C*38^(A) where A is approx. T/10.5 where the 38 is the same
base constant you used for chess. For brute force algorithms we would see another
ply every 10.5 years. I never claimed it was a short time but this shows eventual
perfect chess given games with limited plies.
This is true if the current trend continues. Yes there could be a barrier, but we
could spend forever talking about what _could_ happen. What is important is what
_is_ happening and what will probably happen.
And that's my point here... we can discuss the 1Ghz barrier and other such things
in another place.
Best regards,
Mike.
Michael Chung
> push back the limits set by material properties, we will eventually run
> light.
We do not need it to continue forever, only about 2000 years.
Mike.
Michael Chung
> word'perfect'.
I will not define, but I will try to clarify what I mean. When I say perfect play I
mean never playing inferior moves.
Mike.
Rolf Tueschen
>Making computers think like humans is an interesting idea.
Who exactly did propose to "make" computers "think" like humans? I must
have missed the posting. I also saw Bob reacting to the "idea", but he also
didn't mention the original source.
Phil did talk about the very special case of human GM chess. But he did NOT
reason as if better for yesterday than today all programs should be changed
into "human GM chess".
What I've understood and already did write my article "Fast food..." about
that is the necessity to keep the difference in mind. You see the point? At
first sight after 30 years of always better performing mixture of
brute-force, alpha/beta and so on, the protagonists of this technique (and
especially a double Wchampion here still residing) might confuse these
performances with *real* chess. And the most important reason for such a
confusion is for Phil and for me too the tricky implementation of high
classed human chess history as a guideline for the machines up to move 25
if necessary. This alone does guarantee you a jump from the amateur status
of <2000 up to fabulous 2400. Understood the depth of 12 to 14 ply overall.
The latter gives you the safety against short-ranged blunders. Effective
learning features do their best to make the programs look like real
masters.
But this is all based on a trick with definitions.
If we see two humans play over a chessboard in Central Park N.Y., then we
"say" that they play chess. In case GM Roman Dzin. played there times ago,
this is a true statement. But what if "coffeehouse" players hammered down
their moves while talking surrealistic nonsense to confuse the opponent?
They might play on the base of three or four ply and a few tricks they've
learnt in their youth. Sacs on f7 or h2-h4. Or the Budapest Gambit. Diemer?
This is also chess. But not "GM chess". Now how should programmers who are
rather weak amateur players have a clue of real "GM chess"? Isn't it a
fallacy of thirty year long practice that you begin to confuse you machine
based performancies with human master or even GM play? Especially if you
happen to realize performances on ICC of over 3000? After you have admitted
the relativity of bullet or blitz do you still have the guts to stay
cautiously on scientific grounds or is it a real danger to be taken away by
wishful thinking? --- That is the point of experts of the theory of
science, Phil and Rolf. ---
In chess this is a rather moot debate. It's a game. It's not real life. But
look at all the machines e.g. in our modern aeroplanes and military areas
anyway. There too much naivety about the *real* perforamnces of todays
robots did cost us casualties in the thousands at least. And always we come
later to correct the tiny "errors" as if we've had a little game of "chess"
in Central Park. "Nuuuh, pawn was still on h2?? Not possible. I was sure I
had moved it to h3. Ok. That's completey uninteresting then. Let's start a
new game. Ok? And stop blowing that smoke into my eyes. I can't see no
board..."
Basically this is all about the definition of the real strength of the
thought process of humans and robots. Now chess is in its creative version
(amateurs happen to perform this once in a lifetime and super GMs day by
day) a game about exceptions. It's a double staged problem. To detect
possible exceptions and then to solve them in a given time. It should be
noticed that both parts are non-trivially to master.
Here is my personal solution for the actual almost-GM performances of the
chessmachines. 95% of their success is based on the so-called "surprise"
effect. *All* so-called "matches" human v. machine are played with unknown,
brand new versions of the machines. Also the best machines on ICC are
tweisted and corrected through constant "coaching" by their creators almost
dailey. This is good for the suspense but it doesn't prove in no way the
"strength" of the robots. It more so speaks for the cleverness of the
operators/programmers.
It's a difficult field for laymen. Scientists however detect the most
important factor at the instant. It's so trivial and basic. The *human*
influence cannot be denied.
Still it's a funny event for each chessplayer. Because in a way players are
like gamblers. And often there's money at stake. But this has nothing to do
with fairness or sports. It's a gamble. With one difference. In the long
run the human GMs will win because the possibilities to tweak and twist are
rather limited. And low-rated operators/programmers can't be informed about
the newest discoveries in human GM chess. I said "in the long run". Means
that a GM had to invest some months of research. Obviously 20000 or 50000
or even 100 000 are NOT enough of a recompensation. Kasparov didn't do that
even for possible 1 000 000 dollars. In his case a real preparation wasn't
possible at all. Because the opposing machine was in its actual
chesspersonality like a white sheet of paper. What should a gambler do in
that case? Of course he took the 700 000 or some for no real fight at all.
Honestly! Would anybody here around go into an aeroplane with Deep Blue (of
the second game of the 1997 event) as auto-pilot??
I would still prefer Kasparov! What are your opinions? :)
Rolf Tueschen
>: Rolf generates a load of cantankerous, argumentative, rude crap. Sure. He
>: also generates intelligent prose with ideas and new ways of seeing the
>: world. And on-topic. He gets shat on whatever he does now, on-topic or off.
>: Not good IMO.
>Please go back and find *two* on-topic posts by him...
I'll bite. And want to top your claim.
How much for each computerchess ON-TOPIC from Rolf above n=2?
=======================================================
I once gave 100 § US for Robert Hyatt contributions in three challenges.
How about the same right now?
I'm game. I doubt that Bob will join me.
Peter W. Gillgasch
> fwiw, I could not agree more.
>
> Especially when the aforesaid mud-thrower has said so! ;-)
The key word here was "elegant". Snipping everything with the
exception of a single word and then refering to things which have been
said or not been said hartly qualifies as "elegant" but as "primitive".
-- Peter [confessing elegant mud thrower non-anonymous]
michael adams
>
> EvaNova1 <evan...@aol.com> wrote:
> :>e) the neat and easy solution of having a devil's pit into which to throw
> :>any unacceptable viewpoint. "he's just like Rolf, innit?"
>
> : Steve, I'm new here in this group. I still haven't decided which program to
> : buy. Fritz or Crafty. Now you talk about Rolf. Could you or someone explain the
> : main differences of these programs?
>
> : Kind regards
>
> : Eva
>
> First, you can't buy Crafty as it is free. Second, Fritz is a stronger
> program with a more polished front-end. Third, you wouldn't want a "Rolf"
> if it were given to you as part of a 100 million dollar lottery.
>
> :)
>
> --
> Robert Hyatt Computer and Information Sciences
> hy...@cis.uab.edu University of Alabama at Birmingham
> (205) 934-2213 115A Campbell Hall, UAB Station
> (205) 934-5473 FAX Birmingham, AL 35294-1170
monopolises.
michael adams
>
> Michael Chung <m...@my-dejanews.com> wrote:
> :> I would say three things. This extension of computer
> :> performance may not be linear.
> :> There is no evidence that simple brute force approaches will
> :> solve pattern oriented problems.
>
> : Please forgive me but I feel a need to exercise what I learn in class:
>
> : In any position in chess there are a finite number of moves. So, a computer
> : can make the "best move" from any chess position in a finite amount of time.
>
> : There is evidence that computers keep getting faster at an exponential
> : rate. If this continues then the amount of time needed to find the best
> : move will very quickly approach zero. More specifically, the amount of time
> : required for this will eventually be less than 3 minutes.
>
> : This is enough to prove that given the assumptions, computers will
> : eventually play perfect chess.
>
> I don't believe hardware speeds are "exponential" geometric, perhaps,
> but we are barely seeing a doubling every 2 years... which means we get
> nearly an additional ply every 2 years, so long as this continues. I don't
> believe we will see this doubling every two years much longer, because the
> gigahertz barrier is a tough one, when electrical properties change quite
> drastically as suddenly everything on the chip is an antenna for microwave
> frequencies...
>
> roughly 38, and D is the depth of the search... We don't have such
> a relationship in hardware speeds... just the general 2x every 2 years
> roughly...
>
> Robert Hyatt Computer and Information Sciences
> hy...@cis.uab.edu University of Alabama at Birmingham
> (205) 934-2213 115A Campbell Hall, UAB Station
> (205) 934-5473 FAX Birmingham, AL 35294-1170
Robert Hyatt
: Robert Hyatt wrote:
:> f...@accountant.com wrote:
:> :> On 14 Aug 1998, Robert Hyatt wrote:
:> :> > I don't believe hardware speeds are "exponential" geometric, perhaps,
:> :> > but we are barely seeing a doubling every 2 years...
:>
:> : Exponential : k^x
:>
:> : Geometric: x^k
:>
:> Sorry, wasn't thinking... of course you are right. :)
: In return, that special "tell" string which when fed to crafty when
: playing makes it move randomly for two moves... please? You know, that
: string you type in when you want to entice that GM in for a whole game
: series (drop the first, then hammer him). Not breaking any rules
: thereby.
:> The "Growth" is on a linear curve is what I meant..
: Now *if* Growth (rate, growth by itself gives even more problems) =
: Rate of increase = Slope, and Slope (e^x) = e^x
: i.e. still exponential not linear
: It is frying pan to fire time ;-)
: better therefore we define "Growth" as "Log".
: Log ( k a ^ x ) = Log k + x Log a == ax + b i.e. linear.
: :-)
: Only trying to help, as always.
If you plot this on log paper, you get a straight line... which was
the point of this in the beginning... the hardware speed is improving
by a factor of two every two years, which means, basically, that every
two years the clock speed is doubling... and this growth rate, doubling
every two years, will remain constant for a few more years, but must
start tapering off pretty soon... how much I don't know, because this
tapering off was predicted to start several years ago and it hasn't yet
that I've noticed..
:> IE two times
:> faster every two years, pretty steadily for many years now. Naturally
:> if you figure twice as fast in 2 years, 4 times faster in 4 years, 8 times faster in 6 >years, 16 times faster in 8 years... it is growing... but the
:> "rate" is pretty well constant, and, I'm afraid, going to start to flatten
:> out once we get past the current crop of potential gigahertz machines. I
:> don't see how they work myself, but I've run on at least one, so unless
:> someone is full of voodoo, it does. somehow... but it defies all the laws
:> of electronics I learned during my HAM radio days... :)
:>
:> : Therefore exponential by your own statement: maybe -ish if 2 is
:> : increasing. Geometric comes into it not at all, it is not a decayed
:> : case of exponential... (Don makes same point I believe) ;-)
:>
:> :> > which means we get
:> :> > nearly an additional ply every 2 years, so long as this continues. I don't
:> :> > believe we will see this doubling every two years much longer,
:>
:> : Agreed. It is already longer.
:>
:> Not really longer... IE My P6/200 is just over 2 years old. We now have
:> P6/450 (if you count the Xeon as "here"...)
: "Just over" is critical when such short times are being considered ;-)
: And you could maybe have got it a bit earlier. Contrary to "Samson"'s
: and "DoDo"'s kind assertions, I have experienced how tight Uni budgets
: can be.
No... I bought mine within a month of their being released... machine
cost me about 6,000 bucks for a 64mb P6/200. And at about the same time
this year, 400mhz machines were already available, so 2 years is not a
bad estimate, so far...
: And Xeon 450 does not arrive till August 24 per sysdoc.
yes... but the PII/400 was here before two years elapsed, the PII/450
was around in May or so...
: And P2/450, as "there are improvements to the Pentium Pro core"...
: So, fair to compare PP/200 (512K L2) with P2/400 non-Xeon. The only
: sensible bm I have for both is cpumark 32, and the ratio is 1.67 < < 2.
: And I am probably being generous with the 2 years here (intro-date to
: intro-date) Of course one comp pair proves nothing.
I got better results... the PII/400 clocked in at 1.82 with crafty... the
PII/450 clocked in at just over 2.00, and the xeon at 2.25 which is almost
perfect...
:
:> :> > because the
:> :> > gigahertz barrier is a tough one, when electrical properties change quite
:> :> > drastically as suddenly everything on the chip is an antenna for microwave
:> :> > frequencies...
:>
:> : Optics provides the answers... and puts Kryotech out of business.
:>
:> I'm not sure it provides *all* the answers.
: Why? No heat, no sweat, no problem. And frequency... :-))
Frequency isn't any better, yet, unfortunately... and we don't have
any purely optical technology running... there is digital circuits
in the loop, which has the same problem... and there is still "C" which
means to keep going faster, we have to keep going "smaller" and that is
a huge problem...
:> Some, yes, but it brings in
:> its own set of problems, and is still constrained by "C"...
: Bob, BTW an Expert is now visiting rgcc.... one who had not posted here
: for some months.
: Maybe if I tell him "Visual" C is optimised for Optical computers, he
: believes me?
: No, surely not, despite the fact that I thought he made one post in
: alt.stupid.morons (of course, by mistake).
: Kind regards
: fca
Robert Hyatt
: Still smarting over the P2/450 availability.
: Kind regards
: fca
In an attempt to make you "smarter" I was getting PII/500mhz benchmark
results *months* ago... :) Don't have access to one myself, but the
"grapevine" is pretty long here at times...
Also Digital's new processor is "eye-opening" when they start showing
up...
Sean Evans
>up...
>
>Robert Hyatt                   Computer and Information Sciences
>hy...@cis.uab.edu              University of Alabama at Birmingham
>(205) 934-2213Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â Â 115A Campbell Hall, UAB Station
>(205) 934-5473 FAXÂ Â Â Â Â Â Â Â Â Â Â Â Â Birmingham, AL 35294-1170
Robert Hyatt
: Hello Bob,
: Do you happen to know what the MHZ is on the digital chip
: Best regards,
: Sean Evans
Yes... unfortunately I can't say anything at present, as my "source"
asked me to keep it quiet until the official Digital announcement, which
won't be long. However, I will say this: don't pay a lot of attention to
the mhz speed. This processor is far faster than the old 21164, just on
pure internal design... which seems to have some things in common with the
pentium-pro approach to things, rather than the older "pentium/21164" way
of pipelining/superscalar execution.
can't wait myself...
User77568
>User77568 <user...@aol.com> wrote:
>: "Christopher R. Dorr" <crd...@one.net> wrote:
>:>Steve Blatchford wrote in message <6qs4se$qr5$1...@eros.clara.net>...
>:>But how did they get the 'right sort of wrapping' in the first place? Bob is
>:>*obviously* one of the best qualified posters in this group. He wrote
>:>Crafty, a clearly world-class program. He has a PhD (which does mean
>:>something, Rolf's and Seans protestations notwithstanding), and he is a
>:>professor. To me, these are 'credentials', not 'wrapping'. When you get on
>:>an airplane, I assume that you assume that your pilot has the right
>:>credentials. Would you be as trusting if the pilot didn't have the right
>:>'wrapping'? It doesn't mean that he can't fly, it's just that if he has the
>:>creds, then he probably can.
>: A pilot gets continual tests over the years. It's by far not the same with
>: academics. Too many possibilities to cheat.
>Not as many as you'd think.
Do you want to decide what I should think?
>there are *plenty* of critics that look over
>everything that is written and take instant offense at things that are
>not verifiable... never been in that boat myself, never intend to ride
>in it either...
>:>>It has been pointed out [danger, rgcc off-topic word coming] that this
>:>leads
>:>>to a kind of technical fascism, with masses and 'expert' leaders. Made all
>:>>the worse when the experts are wrong, or lie, or are playing games, or have
>:>>axes to grind. I think Rolf saw this, and in his own way, fought it. The
>:>>rest you know.
>:>>
>:>Rolf rarely talked about technical issues. He accused Ed of stealing, and
>:>Bob of being a fascist. How are these accusations related to a technical
>:>disagreement with one of the experts. You mentioned 'ad hominem' attacks
>:>before? Rolf was *the* major progenitor of them.
>: Last time I checked the record Rolf did not "accuse Ed of stealing". It was
>: character assassination. It was Ed insulting Rolf. Go figure.
>Backward. Rolf attacked Ed hundreds of times. Ed got tired of it and took
>Rolf to court, which further incensed Rolf.
I checked that. I'm not defending Rolf, I also read. He attacked Ed. But then
he wrote that Ed committed character assassination. For me this is a different
thing. Attacks and character assassination. Also Ed did not win his court
trial. Rolf wrote that Ed couldn't provide evidence for his claim that Rolf had
insulted him as nazi.
So why all this should further incense Rolf? Rolf might be no angel either but
what I checked spoke for Rolf.
>That's all there was, continual
>references to "the nazi criminal Schroeder"... over and over and over...
I checked that too. I found the "criminal" with "triple character assassination
of Rolf". Rolf as a victim should have the right to discribe his offender. As I
said I'm not defending Rolf. Criminal is probably too strong. But character
assassination is a serious offence.
Rolf then did explain several times that he as a German used the "nazi" as
"adjective". I couldn't find the wording "nazi criminal". Please show the
quote. Of course I read that Rolf defined his unjustified exclusion on CCC as
a fascist (or nazi) act. I checked that Rolf used this notion as the negation
of "democratic".
I checked all the events of autumn 1997. I found out that nobody had accused
Rolf of anything wrong on CCC.
I said I'm not defending Rolf. But false decisions remain false decisions. Also
nobody should be punished for things he didn't do.
Much of this "over and over and over" seems to be a consequence of the at least
unfair treatment Rolf got here.
>: What is a fascist? Is he gasolining human bodies? Does he censor people
without
>: a cause on CCC? Or do you think of euthanasia or even abortion=murder after
>: birth? All ideas of "Bob". Go figure.
>1. while I was a "moderator" (not a *censor*) we did not delete one single
>post. We were accused of doing so by "Sean"... but we proved that his post
>was still there, but an "operator error" on his part was not showing it. Not
>our problem in giving lessons on operating netscape and a web site message
>board. So on my watch, no one was "censored". The entire group of founders,
>less a couple of us, voted to can Rolf.
Who were the founders? What was their status before?
>I was against it, so was Bruce and
>Steve. The majority won. I'm now convinced it was the right action, albiet
>for the wrong reasons, because he has repeatedly shown here that he can't
stick
>to "computer chess topics" for very long without going off into religion,
>politics, and whatever else suits his agenda against Ed.
Allow me one question. You talk of his agenda against Ed. What about you
yourself? I was talking with Chris about your record and why Rolf called your
output "fascist". Didn't you read it?
>: Rolf's arguments against such ideas and actions are 'ad hominem' attacks for
>: you?
>"The Criminal Schroeder" is certainly an ad hominem. There were *many*
>others...
I checked that. For me Rolf did publish a statement. Character assassination
surely is a criminal behaviour. So all depends on the offence. I checked that.
Rolf repeated this over and over that Ed had accused him of misusing his
"academic titles" and of cheating his clients. Ed had insulted Rolf as a "pig"
because he had attacked, as Rolf said "nazi-like" methods "propagated" by one
of Ed's friends. And then Ed had accused and sued Rolf for insulting Ed as a
nazi.
For me "pig" isn't character assassination. Also a false accusation is not
necessarily character assassination. But if you look deeper then the notion
seems justified. Ed did surely violate Rolf's reputation.
I did some further research. I found out that you did the same against Rolf. I
checked that Rolf has attacked many events as "fascist" or "nazi-like". How
could you insult Rolf as the "worlds greatest nazi admirer"? I saw a lot of
Rolf's postings, but nowhere did he admire nazis.
Gasolining and dismembering? What about that? Euthanasia for Rolf? What about
that?
Did you ever reflect on Rolf as a human being? What could be possible reasons
for Rolf's sensitivity for all sorts of threats against the intact physical
status? I checked that. Rolf did never talk about. But others wrote about awful
physical handicaps. You really should be ashamed of your campaign against Rolf.
I think it's not your support of the existing death penalty in the USA that is
frightening for Rolf, but your atrociously violent fantasies about the tratment
of human bodies. Death penalty doesn't include torture and all the inhumanities
you did propose here in this chessgroup.
You can't rewrite history that is in dejanews.com.
Why did nobody, also Robert Hyatt, react positively on Rolf's peace proposals?
I checked that. All I could find were reactions like "Teuschen, lie down and
die".
Regards
Sean Evans
>he wrote that Ed committed character assassination. For me this is a different
>thing. Attacks and character assassination. Also Ed did not win his court
>trial. Rolf wrote that Ed couldn't provide evidence for his claim that Rolf had
>insulted him as nazi.
>
>So why all this should further incense Rolf? Rolf might be no angel either but
>what I checked spoke for Rolf.
Schroder beschliesse bis go to gericht und untergelegen. Schroder bauen CCC bis gehegt himself und jetzt ihnen tritt volk raus ihnen getan nicht like. Ihnen genommen mein losungswort away ohne angebracht gründe. Ihnen not bis be gestoppt. Volk darf erfahre bis gewehrt themselves auf öffentlichkeit und nicht gebrauch Ray Guns und barbwire bis gehegt themselves!
>I checked that too. I found the "criminal" with "triple character assassination
>of Rolf". Rolf as a victim should have the right to discribe his offender. As I
>said I'm not defending Rolf. Criminal is probably too strong. But character
>assassination is a serious offence.
<User77568>
>Gasolining and dismembering? What about that? Euthanasia for Rolf? What about
>that?
Wieder ein doppelt standard wo einige volk gekonnt gesagt was ihnen mögend und gekriegt away mit es. Mai ich gemacht ein freundlich suggestion, erfreue gebrauch dein wirklich name auf künftig, normally nur "Trolls" gebrauch falsch name.Â
Dein freund auf computer schach,
Phil Innes
Making computers think like humans is an interesting idea. Of course,
to
actually do this we would require... (guess). It's unlikely that we
would want
computers to imitate humans.
I believe that this is a major part of what we are about to
explore Michael. Of course the question is... to what degree
do we understand human thinking? Perhaps compare your
impressions with the on-going discussion here. Phil
Phil Innes
The node search speed improvement is
not linear, it's easily seen that it's exponential. And although there
may
be a limit for computers, I think of the limit as where the result of
perfect play from every chess position is known, using brute-force or
any
other algorithm. Only then is there no further computers can go--but
until
then computers will continue to improve.
May I defer any response to this comment. I should like first
to define a few terms, a) with Bob on the nature of the
*evaluation* of the algorithm, and b) with Rolf, on some
respectable terms that we can use to describe
"intelligence(s)!" (I haven't done either yet) Phil
Phil Innes
I don't believe hardware speeds are "exponential" geometric, perhaps,
but we are barely seeing a doubling every 2 years... which means we get
nearly an additional ply every 2 years, so long as this continues. I
don't
believe we will see this doubling every two years much longer, because
the
gigahertz barrier is a tough one, when electrical properties change
quite
drastically as suddenly everything on the chip is an antenna for
microwave
frequencies...
It is not a current techology in a commercial sense, but nano
second optical transmissions are already used in some
applications, and simple silicon photodiode semiconductors are
capable of femto-second differentials in receipt of data.
I anticipate myself a little, and would like to formulate the
question in a more structured way, but is Michael addressing
simply the clock speed of the processor as it effects search
tree capability:time?
Does this increase in speed (even using a hypothetical optical
computer with a speed increment of ten exp. six over a 1 gig
processor/bus) have any QUALITATIVE function?
ie: Is there any inherent increment in AI functionality? Or
does it just arrive at a similar conclusion faster than
earlier processors (Turing's thesis.) Phil
PS - Please skip this entirely if I have grossly
undersimplified the effect of processor speed. I should like
to take up this issue using a better "frame"
Daniel Kang
>Making computers think like humans is an interesting idea. Of course, to
>actually do this we would require... (guess). It's unlikely that we would want
>computers to imitate humans. It would be much more difficult to get computers
>to understand human 'aspiration' than to just solve chess.
Just to add another point, we should also be aware that computers
are already far more efficient than humans in playing chess. Even
Deep Blue does not come close to the processing power of a human
brain, yet it will defeat nearly every single human being on earth
except possibly very few. There is no need for the computers to
imitate human thinking, that, in my opinion, is inferior in terms
of efficiency.
Regards,
Daniel Kang
Robert Hyatt
:>Not as many as you'd think.
: Do you want to decide what I should think?
Just pointing out that your "thinking" is wrong here. Think what you
want, however...
:>: Last time I checked the record Rolf did not "accuse Ed of stealing". It was
:>: character assassination. It was Ed insulting Rolf. Go figure.
:>Backward. Rolf attacked Ed hundreds of times. Ed got tired of it and took
:>Rolf to court, which further incensed Rolf.
: I checked that. I'm not defending Rolf, I also read. He attacked Ed. But then
: he wrote that Ed committed character assassination. For me this is a different
: thing. Attacks and character assassination. Also Ed did not win his court
: trial. Rolf wrote that Ed couldn't provide evidence for his claim that Rolf had
: insulted him as nazi.
He attacked ed. He recently directly accused him of stealing copyrighted
material to use in his opening book. Several pointed out that what Ed and
the rest of us use in our books is absolutely not copyrightable... There
were many other cases..
: So why all this should further incense Rolf? Rolf might be no angel either but
: what I checked spoke for Rolf.
Check further. I can't think of anything that "speaks *for* Rolf"...
:>That's all there was, continual
:>references to "the nazi criminal Schroeder"... over and over and over...
: I checked that too. I found the "criminal" with "triple character assassination
: of Rolf". Rolf as a victim should have the right to discribe his offender. As I
: said I'm not defending Rolf. Criminal is probably too strong. But character
: assassination is a serious offence.
Rolf was no victim. Ed had nothing to say to Rolf, until Rolf erupted
here with his attacks on Ed. Rather than responding *here* Ed chose to
respond in a legal fashion. That was Ed's "only criminal act"... to
challenge Rolf's legal ability to make slanderous remarks in public, over
and over.
: Rolf then did explain several times that he as a German used the "nazi" as
: "adjective". I couldn't find the wording "nazi criminal". Please show the
: quote. Of course I read that Rolf defined his unjustified exclusion on CCC as
: a fascist (or nazi) act. I checked that Rolf used this notion as the negation
: of "democratic".
: I checked all the events of autumn 1997. I found out that nobody had accused
: Rolf of anything wrong on CCC.
: I said I'm not defending Rolf. But false decisions remain false decisions. Also
: nobody should be punished for things he didn't do.
This is incorrect. If I see *you* drunk and disorderly at a football game, I
will absolutely *not* invite you to my house. CCC was founded to eliminate
personal attacks. I'd have preferred to let Rolf stay until he shot himself
in the foot, which he most certainly would have done, but we certainly had
every right to exclude him. CCC is *not* public. It is a *private* facility
where we "invite" everyone to come, but where we reserve the right to excise
"bad actors." Rolf may have been excised sooner than I'd have preferred, but
there is *no* doubt that he would have been excised somewhere down the line.
And we clearly have the right to exclude anyone we want. And anyone there
has the right to exclude himself by leaving if they don't like the way CCC
is being managed. But *no* one has the "right" to go to CCC.
: Who were the founders? What was their status before?
I don't know what you mean "status." But there were several, Myself, Bruce,
Ed, Chris, Thorsten, Moritz, Dirk, Enrique, etc... IE the usual posters here..
:>I was against it, so was Bruce and
:>Steve. The majority won. I'm now convinced it was the right action, albiet
:>for the wrong reasons, because he has repeatedly shown here that he can't
: stick
:>to "computer chess topics" for very long without going off into religion,
:>politics, and whatever else suits his agenda against Ed.
: Allow me one question. You talk of his agenda against Ed. What about you
: yourself? I was talking with Chris about your record and why Rolf called your
: output "fascist". Didn't you read it?
Nope. I try to read what is related to computer chess. Between that and
my daily email traffic of about 50-60, plus reading CCC, that's about all
the time I have. *I* don't have an "agenda". I don't sell chess software,
I don't work for a company that does. I've been posting here for I have no
idea how many years, and haven't changed appreciably since I started. So
"agenda" doesn't apply to me...
: I did some further research. I found out that you did the same against Rolf. I
: checked that Rolf has attacked many events as "fascist" or "nazi-like". How
: could you insult Rolf as the "worlds greatest nazi admirer"? I saw a lot of
: Rolf's postings, but nowhere did he admire nazis.
Quite easy. If someone uses the term "nazi" in 99% of his posts, there is
obviously some deep-rooted fascination with the term, and with the ideology.
What, I don't know, but it is definitely there. I rarely use that "word".
Rolf uses it like most of us use "the".
: Gasolining and dismembering? What about that? Euthanasia for Rolf? What about
: that?
Did you find the original posts? Gasolining as a response to the death penalty
and how "futile" it really is since nothing good at all comes from it. The
"dismembering" was a simple example of just how parents might feel about someone
that brutally beat, raped, and ultimately murdered their teen-age daughter,
after 12 brutal hours of torture. "thedodo" accused "alabama" of being a
barbaric state for electrocuting this guy. I pointed out that the girl's
parents probably had quite a different perspective. "simple example or analogy,
get it now"???
: Did you ever reflect on Rolf as a human being?
Absolutely, at first. I tried to talk to him many times, in fact. But his
true nature keeps returning, time after time. I have written him off. I won't
try again...
: What could be possible reasons
: for Rolf's sensitivity for all sorts of threats against the intact physical
: status? I checked that. Rolf did never talk about. But others wrote about awful
: physical handicaps. You really should be ashamed of your campaign against Rolf.
I have *no* campaign against Rolf. I pled his case on CCC. I talked to him
here while he was lambasting others. Please go back and read thru 1 year's worth
of deja here... your "picture" will change...
: I think it's not your support of the existing death penalty in the USA that is
: frightening for Rolf, but your atrociously violent fantasies about the tratment
: of human bodies. Death penalty doesn't include torture and all the inhumanities
: you did propose here in this chessgroup.
: You can't rewrite history that is in dejanews.com.
Never tried to rewrite history. I've left that to a couple of "others" here
that seem to delight in doing so.
: Why did nobody, also Robert Hyatt, react positively on Rolf's peace proposals?
: I checked that. All I could find were reactions like "Teuschen, lie down and
: die".
Perhaps, because we heard him cry "wolf" too many times? And we learned that
it was all a load of crap?
: Regards
Sean Evans
>Â Â Â Â I believe that this is a major part of what we are about to
>Â Â Â Â explore Michael. Of course the question is... to what degree
>Â Â Â Â do we understand human thinking? Perhaps compare your
>Â Â Â Â impressions with the on-going discussion here. Phil
Robert Hyatt
: Hi Bob, I see that you wrote: (responding to Michael Chung)
This is a common misconception... ie faster hardware doesn't "just produce
a move faster". On the contrary, my program will take the same amount of
time no matter what you run it on... but it *will* search deeper on faster
iron. And deeper search leads to qualitatively better play. A doubling of
speed is almost enough for another ply of search, and once a few pieces are
off, doubling gives a ply at least, maybe more...
So, bottom line, approximately every two years, we tack on another ply, and
thereby solve a whole new set of tactical issues that we couldn't solve two
years back... Sooner or later, we will solve enough that humans can't cope
any longer. How long it will take to do this is anybody's guess. But we
are *clearly* seeing progress, if you compare programs of today with programs
from 10 years ago...
Chris Whittington
Robert Hyatt wrote in message <6r5ikg$nku$1...@juniper.cis.uab.edu>...
Never before (well, maybe a few times) has such fraudulent use of language
been employed to mask over and deny a brilliantly presented and simple
truth.
As repeated here:
ie: Is there any inherent increment in AI functionality? Or does
it just arrive at a similar conclusion faster than earlier processors
(Turing's thesis.)
Off "we" go ...
>
>This is a common
not common, unstated before on this ng in the last three years
>misconception...
it's no misconception
> ie faster hardware doesn't "just produce
>a move faster".
oh yes it does. it is all it does.
>On the contrary, my program will take the same amount of
>time no matter what you run it on...
cheat. the program takes as long as you care to let it run. your decision,
not its, and not hardware. First mix of terminology, jumping at will from
hardware to program, with human muddled in somewhere, as if they were
mutually interchangeable and anyway the same thing.
>but it *will* search deeper on faster
>iron.
correct. but "Is there any inherent increment in AI functionality? Or does
it just arrive at a similar conclusion faster than earlier processors
(Turing's thesis.) "
> And deeper search leads to qualitatively better play.
quantitative. let's not pretend anybody did anything qualitative at all.
only the hardware got faster. bean-counting. Punkt.
>A doubling of
>speed is almost enough for another ply of search, and once a few pieces are
>off, doubling gives a ply at least, maybe more...
Every year thus putting off yet again the fateful day when you actually have
to have any chess knowledge in the thing.
>
>So, bottom line, approximately every two years, we
"we" ? :)))))). the hardware manufacturor did that. you just got out the
cheque book. cheat argument. not to mention the usual identity crisis.
>tack on another ply, and
>thereby solve a whole new set of tactical issues that we
"we" again. more identity crisis and inclusion of $4,000,000,000 Intel
processing plant.
>couldn't solve two
>years back...
As Phil points out, "you" (the program) could. "You" (the human, allegedly)
just left it running longer.
But: "Is there any inherent increment in AI functionality?"
>
> Sooner or later, "we"
identity crisis and fraudulent use of "we" yet again.
> will solve enough that humans can't cope
>any longer. How long it will take to do this is anybody's guess. But we
>are *clearly*
Danger: use of "clearly", "obviously" etc. hides erroneous argument. And
another "we".
>seeing progress, if you compare programs of today with programs
>from 10 years ago...
Brillant cheat ! Drones on about "hardware" x2 every year, then sudden
switch to "program" as comparitive element. I protest !
Typical confused thinking of the bean-counter paradigm. Gigantic muddle
between who does what and source of progress masked by use of word "we".
Keeps it this way in order to forever stall the issue of the non-existent
knowledge. Which "we" don't have.
Bob Hyatt, you win this weeks George Orwell 1984 Memorial Prize for Clarity
of Thought and Accurate Word Usage. Congratulations.
Chris Whittington
Chris Whittington
>
>Making computers think like humans is an interesting idea. Of course,
>to
>actually do this we would require... (guess). It's unlikely that we
>would want
>computers to imitate humans.
>
> explore Michael.
Certainly is. But already confusing language is being used. As in 'human
thinking'. The idea that there are two ways 'machine brute-force thinking'
and 'human-thinking' is, IMO, erroneous. And dichotomising the problem in
this way is part of the problem.
> Of course the question is... to what degree
> do we understand human thinking? Perhaps compare your
> impressions with the on-going discussion here. Phil
IMO it is necessary to first look at the language that is used. Language is
kind of thought, no ? The words you use are your thoughts. Without language
thoughts are not possible. Without realising which langauge terms are
hindering the move from the old to the new paradigm; and which new words are
required, progress will be very difficult; and discourse will break down
into non-understanding.
Chris Whittington
>
Chris Whittington
Robert Hyatt wrote in message <6r4ess$abr$1...@juniper.cis.uab.edu>...
>Sean Evans <spe...@home.com> wrote:
>: Hello Bob,
>
>: Do you happen to know what the MHZ is on the digital chip
>
>: Best regards,
>
>: Sean Evans
It doesn't matter.
Chris Whittington
>
>
>
>Yes... unfortunately I can't say anything at present, as my "source"
>asked me to keep it quiet until the official Digital announcement, which
>won't be long. However, I will say this: don't pay a lot of attention to
>the mhz speed. This processor is far faster than the old 21164, just on
>pure internal design... which seems to have some things in common with the
>pentium-pro approach to things, rather than the older "pentium/21164" way
>of pipelining/superscalar execution.
>
>can't wait myself...
>
Robert Hyatt
: Phil Innes wrote in message <35D63EBF...@sover.net>...
:>Michael Chung wrote:
:>
:>Making computers think like humans is an interesting idea. Of course,
:>to
:>actually do this we would require... (guess). It's unlikely that we
:>would want
:>computers to imitate humans.
:>
:> I believe that this is a major part of what we are about to
:> explore Michael.
: Certainly is. But already confusing language is being used. As in 'human
: thinking'. The idea that there are two ways 'machine brute-force thinking'
: and 'human-thinking' is, IMO, erroneous. And dichotomising the problem in
: this way is part of the problem.
The problem *I* always have with this can be summarized by the following
two points: (1) Lots of folks know how computers do what they do to play
chess, from search strategies to evaluation techniques. (2) *nobody* knows
how humans do what they do to play chess. In light of those two things, I
don't see any way to get into a deep discussion about how a computer can
"do it like a human" when we don't have a clue about how a human does "it".
:> Of course the question is... to what degree
:> do we understand human thinking? Perhaps compare your
:> impressions with the on-going discussion here. Phil
: IMO it is necessary to first look at the language that is used. Language is
: kind of thought, no ? The words you use are your thoughts. Without language
: thoughts are not possible. Without realising which langauge terms are
: hindering the move from the old to the new paradigm; and which new words are
: required, progress will be very difficult; and discourse will break down
: into non-understanding.
: Chris Whittington
:>
Chris Whittington
Robert Hyatt wrote in message <6r6qmj$583$2...@juniper.cis.uab.edu>...
>Chris Whittington <chr...@cpsoft.demon.co.uk> wrote:
>
>: Phil Innes wrote in message <35D63EBF...@sover.net>...
>:>Michael Chung wrote:
>:>
>:>Making computers think like humans is an interesting idea. Of course,
>:>to
>:>actually do this we would require... (guess). It's unlikely that we
>:>would want
>:>computers to imitate humans.
>:>
>:> I believe that this is a major part of what we are about to
>:> explore Michael.
>
>: Certainly is. But already confusing language is being used. As in 'human
>: thinking'. The idea that there are two ways 'machine brute-force
thinking'
>: and 'human-thinking' is, IMO, erroneous. And dichotomising the problem in
>: this way is part of the problem.
>
>The problem *I* always have with this can be summarized by the following
>two points:
>
>(1) Lots of folks know how computers do what they do to play
>chess, from search strategies to evaluation techniques.
>
Correct. This is called the bean-counter paradigm.
>(2) *nobody* knows
>how humans do what they do to play chess.
Not correct. It's only correct that the bean-counters don't know this.
That's why they are forced to bean-count.
>In light of those two things, I
>don't see any way to get into a deep discussion about how a computer can
>"do it like a human" when we don't have a clue about how a human does "it".
I know you don't :)
But that's ok, Bob.
I know it's my role to drag you screaming and kicking into the twenty-first
century :)
Chris Whittington
Robert Hyatt
: Robert Hyatt wrote in message <6r6qmj$583$2...@juniper.cis.uab.edu>...
:>Chris Whittington <chr...@cpsoft.demon.co.uk> wrote:
:>
:>: Phil Innes wrote in message <35D63EBF...@sover.net>...
:>:>Michael Chung wrote:
:>:>
:>:>Making computers think like humans is an interesting idea. Of course,
:>:>to
:>:>actually do this we would require... (guess). It's unlikely that we
:>:>would want
:>:>computers to imitate humans.
:>:>
:>:> I believe that this is a major part of what we are about to
:>:> explore Michael.
:>
:>: Certainly is. But already confusing language is being used. As in 'human
:>: thinking'. The idea that there are two ways 'machine brute-force
: thinking'
:>: and 'human-thinking' is, IMO, erroneous. And dichotomising the problem in
:>: this way is part of the problem.
:>
:>The problem *I* always have with this can be summarized by the following
:>two points:
:>
:>(1) Lots of folks know how computers do what they do to play
:>chess, from search strategies to evaluation techniques.
:>
: Correct. This is called the bean-counter paradigm.
:>(2) *nobody* knows
:>how humans do what they do to play chess.
: Not correct. It's only correct that the bean-counters don't know this.
: That's why they are forced to bean-count.
It's correct. Unless you can produce your Nobel prize that anyone who
discovers how the human plays chess would certainly produce. We have some
interesting projects going on here at UAB, between the computer science
department, the cognitive science group and some psychologists. And they
are simply trying to understand basic things about how young children "learn"
to match a few simple objects upon command. Playing chess goes way beyond
that...
:>In light of those two things, I
:>don't see any way to get into a deep discussion about how a computer can
:>"do it like a human" when we don't have a clue about how a human does "it".
: I know you don't :)
: But that's ok, Bob.
: I know it's my role to drag you screaming and kicking into the twenty-first
: century :)
: Chris Whittington
Nope... I'm going to plod right into it, learning as I go... :)
Sean Evans
<Dr. Hyatt snip>
>The problem *I* always have with this can be summarized by the following
>two points: (1) Lots of folks know how computers do what they do to play
>how humans do what they do to play chess. In light of those two things, I
>don't see any way to get into a deep discussion about how a computer can
>"do it like a human" when we don't have a clue about how a human does "it".
This topic has more light shining on it every year. The billions of
synapsis, the place where action potentials from one neuron release
chemicals that influence the firing of another neuron in the brain that
transmit thoughts (data) are becoming better and better understood. They
know the chemicals and how they work together they just cannot reproduce
them right now.
There may be a short cut though if we had a brain donation to connect to
your Cray System, any volunteers Bob???
Regards,
Sean
Peter W. Gillgasch
> Rolf
[ ad nauseam. sigh. ]
Bob, pleazzzzzzzeeeeeeeeeeeee.
-- Peter [sitting on a flak and recycling postings]
User77568
Here Robert Hyatt deleted an important paragraph. "A pilot gets continual tests
over the years. It's by far not the same with academics. Too many possibilities
to cheat." Either delete the complete topic or leave it completely there but
don't mess around with a snip here and there. Thank you.
>:>Not as many as you'd think.
>: Do you want to decide what I should think?
>Just pointing out that your "thinking" is wrong here. Think what you
>want, however...
>:>: Last time I checked the record Rolf did not "accuse Ed of stealing". It
was
>:>: character assassination. It was Ed insulting Rolf. Go figure.
>:>Backward. Rolf attacked Ed hundreds of times. Ed got tired of it and took
>:>Rolf to court, which further incensed Rolf.
>: I checked that. I'm not defending Rolf, I also read. He attacked Ed. But
then
>: he wrote that Ed committed character assassination. For me this is a
different
>: thing. Attacks and character assassination. Also Ed did not win his court
>: trial. Rolf wrote that Ed couldn't provide evidence for his claim that Rolf
had
>: insulted him as nazi.
I think you didn't understand my point. Because you wrote:
>He attacked ed. He recently directly accused him of stealing copyrighted
>material to use in his opening book. Several pointed out that what Ed and
>the rest of us use in our books is absolutely not copyrightable... There
>were many other cases..
I said I checked all that. And I checked that Ed did in fact commit character
assassination against Rolf. Your reaction "He attacked Ed" followed as if I had
said "No, Rolf did never attack Ed, but Ed did commit character assassination."
Obviously I had said "He attacked Ed." You did quote that!
You're not only messing around with manipulating the actual text but also the
public record everybody could check in dejanews.com.
Also you talk about events that can't be found in dejanews.com. With the very
same method Ed lost his court trial!
"He recently directly accused him of stealing copyrighted material to use in
his opening book." I couldn't find the evidence. Rolf did never talk about
opening books. Go figure. He talked about some "Millionbase" which has no
"opening book" at all.
Then you mess around again. I checked that. Rolf did not talk about "Ed and the
rest of us" and how they might use their "books".
Also you completely ignored Rolf's statement that he did not want to say that
Ed did steal. I checked that. The difference between your allegation and Rolf's
statement seems clear to me. Rolf talked about "data" he knew as coming
originally from INFORMATOR. Go figure.
The data of INFORMATOR is known as copyrighted.
I got an email from Rolf where he proved that Ed did use copyrighted data. I
saw the evidence. It's quite clear. Rolf asked me to keep the details
confidential. Let me state this. There was a debate about copyrightable data
and free data. What I saw was clearly copyrightable data. If stolen or whatever
is a different question.
>: So why all this should further incense Rolf? Rolf might be no angel either
but
>: what I checked spoke for Rolf.
>Check further. I can't think of anything that "speaks *for* Rolf"...
I already asked you if you wanted to decide what I should think. Now you try to
ignore what I reported about what I had seen in dejanews.com.
When I see that your claim can't be proven with the original postings then it
"speaks for" Rolf. It's not because of the quality of Rolf's output itself but
because of the failed confirmation of your (therefore false) allegation about
his output.
>:>That's all there was, continual
>:>references to "the nazi criminal Schroeder"... over and over and over...
>: I checked that too. I found the "criminal" with "triple character
assassination
>: of Rolf". Rolf as a victim should have the right to discribe his offender.
As I
>: said I'm not defending Rolf. Criminal is probably too strong. But character
>: assassination is a serious offence.
Where is the "cited post" with the "nazi criminal Schroeder"?
>Rolf was no victim. Ed had nothing to say to Rolf,
Character assassination is nothing? Hear hear!
>until Rolf erupted
>here with his attacks on Ed. Rather than responding *here* Ed chose to
>respond in a legal fashion.
Didn't Ed insult Rolf as a "pig" for his attack against Thorsten's "nazi-like"
ideas?
>That was Ed's "only criminal act"... to
>challenge Rolf's legal ability to make slanderous remarks in public, over
>and over.
I checked that. Why then did Ed challenge Rolf with completely invented
"insults"? Rolf allegedly had insulted Ed as a nazi?
>: Rolf then did explain several times that he as a German used the "nazi" as
>: "adjective". I couldn't find the wording "nazi criminal". Please show the
>: quote. Of course I read that Rolf defined his unjustified exclusion on CCC
as
>: a fascist (or nazi) act. I checked that Rolf used this notion as the
negation
>: of "democratic".
>: I checked all the events of autumn 1997. I found out that nobody had accused
>: Rolf of anything wrong on CCC.
>: I said I'm not defending Rolf. But false decisions remain false decisions.
Also
>: nobody should be punished for things he didn't do.
>This is incorrect. If I see *you* drunk and disorderly at a football game, I
>will absolutely *not* invite you to my house. CCC was founded to eliminate
>personal attacks. I'd have preferred to let Rolf stay until he shot himself
>in the foot, which he most certainly would have done, but we certainly had
>every right to exclude him. CCC is *not* public. It is a *private* facility
>where we "invite" everyone to come, but where we reserve the right to excise
>"bad actors." Rolf may have been excised sooner than I'd have preferred, but
>there is *no* doubt that he would have been excised somewhere down the line.
>And we clearly have the right to exclude anyone we want. And anyone there
>has the right to exclude himself by leaving if they don't like the way CCC
>is being managed. But *no* one has the "right" to go to CCC.
I said I'm not defending Rolf. Now again you did mess around with a confusing
manipulation. I had told you that I had found a lot of evidence that much of
Rolf's actions might be a "consequence of the unfair treatment in this group".
On CCC he was punished although he didn't misbehave on CCC.
Now you preferred to delete again the important context of my statement. Then
of course you could construct your case as if Rolf acted like a drunken man and
"disorderly". Like a "rebel without a cause".
The weakness of your construction is quite obvious with the answer on the
following question.
>: Who were the founders? What was their status before?
>I don't know what you mean "status." But there were several, Myself, Bruce,
>Ed, Chris, Thorsten, Moritz, Dirk, Enrique, etc... IE the usual posters
here..
I checked that. You are badly wrong again.
Bruce was no longer there when Rolf was excluded without a reason.
Ed was the one who just had lost his court trial. The agenda seems to be quite
obvious. Next. Chris, I think you mean Whittington and not Dorr, was the one
with Ed who had rejected Rolf's peace offers with the reason that Rolf were
Hitler. The agenda seems to be quite obvious. Next. Thorsten. Rolf had attacked
his "nazi-like" ideas. The agenda seems to be quite obvious. Next Moritz and
Dirk. Two close friends of Thorsten. The agenda seems to be quite obvious.
Finally Enrique. The one who was angry about Rolf because he had refused to
accept an invitation to Spain. This against the background of Rolf supposedly
having some "awfull physical handicaps".
A very special and mean agenda so to speak.
All these "posters" had one thing in common. They were no "usual" posters.
"Etc."? Who? Andreas Mader who had an axe to grind because he once was caught
at false allegations against Rolf? I checked that. Andreas Mader had to
apologize in public. Finally I found a Peter Schreiner. Also a close friend of
Thorsten who slept with him in the same small room during the WCCC 1997. The
agende seems to be quite obvious. Also against the background of no attacks or
insults at all from Rolf's side. "pitters" in special was not at all a "usual"
poster simply because he usually didn't post at all except for a technical
information all two months. I checked that.
Rolf as a German once had posed a question about "pitters" other job in
Germany. In some journal. But "pitters'" didn't answer the important question
for the Komputer Korner thread. I checked that. Rolf awarded the Medal of the
Week to Peter Schreiner for best communication. Ironically referring to
"Komputer Korner's" Gold Medals. Not really an attack, not even less an insult.
So the agenda seems to be quite clear.
I would prefer to talk of a little clique than of "usual posters".
>:>I was against it, so was Bruce and
>:>Steve. The majority won. I'm now convinced it was the right action, albiet
>:>for the wrong reasons, because he has repeatedly shown here that he can't
>: stick
>:>to "computer chess topics" for very long without going off into religion,
>:>politics, and whatever else suits his agenda against Ed.
>: Allow me one question. You talk of his agenda against Ed. What about you
>: yourself? I was talking with Chris about your record and why Rolf called
your
>: output "fascist". Didn't you read it?
>Nope. I try to read what is related to computer chess. Between that and
>my daily email traffic of about 50-60, plus reading CCC, that's about all
>the time I have. *I* don't have an "agenda". I don't sell chess software,
>I don't work for a company that does. I've been posting here for I have no
>idea how many years, and haven't changed appreciably since I started. So
>"agenda" doesn't apply to me...
I think we both agree that "many years posting here" and "haven't changed
appreciably since" doesn't necessarily prove the non-existence of any possible
reason for you to have a hidden agenda against Rolf.
But this was not my point. It should be allowed to ask if you have hidden
agendas at all. There must be a reason for your output, Rolf defined as
"fascist"? That was my question.
BTW here you did snip the paragraphs about Ed's character assassinating Rolf.
>: I did some further research. I found out that you did the same against Rolf.
I
>: checked that Rolf has attacked many events as "fascist" or "nazi-like". How
>: could you insult Rolf as the "worlds greatest nazi admirer"? I saw a lot of
>: Rolf's postings, but nowhere did he admire nazis.
>Quite easy. If someone uses the term "nazi" in 99% of his posts, there is
>obviously some deep-rooted fascination with the term, and with the ideology.
>What, I don't know, but it is definitely there. I rarely use that "word".
>Rolf uses it like most of us use "the".
Perhaps this is the moment where you could understand my doubts about someone
with a PhD 15 years after his BA.
I checked that. There's no evidence for the occurence of the "term 'nazi'" in
"99% of his posts". If then someone frequently wrote negatively about a certain
topic then how this could be the proof for "obviously some deep-rooted
fascination"? Must be sort of secret wisdom you used to achieve such heights
of higher nonsense. Also you forgot about Rolf being a German. Also you forgot
about his alleged personal concern as a physically handicapped. I checked that.
I did never find expressions of fascination. What I found was more something
like expressions out of despair, horror and terror.
>: Gasolining and dismembering? What about that? Euthanasia for Rolf? What
about
>: that?
>Did you find the original posts? Gasolining as a response to the death
penalty
>and how "futile" it really is since nothing good at all comes from it. The
>"dismembering" was a simple example of just how parents might feel about
someone
>that brutally beat, raped, and ultimately murdered their teen-age daughter,
>after 12 brutal hours of torture. "thedodo" accused "alabama" of being a
>barbaric state for electrocuting this guy. I pointed out that the girl's
>parents probably had quite a different perspective. "simple example or
analogy,
>get it now"???
I said that I checked that. I think that your reasoning is biased. For your
postings you try to find more or less reasonable explanations. But the same
sort of explanation with personal motivations or as sort of example or analogy
for something different, basically a sound conviction, a deep concern, you are
denying in Rolf's case. That's wrong.
I said I checked that. I read your original post and Rolf's reaction. I also
read many other posts in that context. Nobody did defend your positions about
gasolining and dismembering. You should accept that. Enrique and Chris
Whittington told you why your output was unacceptable.
It's not because anybody denied you the right to post your mindgames. In detail
your thoughtexperiments are allowed. But Enrique and Chris already made it
quite clear, you with your PhD shouldn't distribute such inhuman ideas and
methods. Rolf called them "fascist".
Nevertheless thinking about such topics is allowed and also scientists already
had thought about. But here in a computerchess group in usenet the like
postings are most unwanted. Here in this field these postings are a provocation
by definition.
I read that Rolf is a psychologist. I'm quite confident that Rolf had all the
necessary understand in the world for the emotions of parents who had lost
their child. The whole R.G.C.C surely would tolerate if the parents would post
something similar to yours into this group. But this is totally different to
your own postings. You had no right at all to post that into a computerchess
newsgroup.
If however you want to claim your rights to post the like ideas then please do
grant Rolf the same rights for his criticism.
I checked that. I understand that Rolf's "over and over and over" repetitions
came from the early negation of all sorts of criticism as such. I read an early
post from Bruce Moreland where he didn't argue with Rolf's theories but where
he denied Rolf the right to applicate his special theories at all. Later Bruce
corrected his post. But then the same thing happened again. Ed called Rolf a
pig and denied him to applicate his arguments on Thorsten. This is ridiculous.
Rolf criticised Thorsten's methods (how serious the might have been) as
"nazi-like". But Ed told Rolf that he should know that Thorsten was anything
but a nazi. Isn't it stupid? Because someone isn't a nazi he isn't able to post
"nazi-like" ideas?
This is the same with you too. You are also not at all a nazi or a fascist, I
guess. But still your output could be fascist. Wouldn't you agree with me?
Further. If your output would be fascist then it would be still completely
unclear if you were a fascist or fascinated by fascist thoughts.
Further. If however one could prove that your output in that special posting
had a supporting aspect, an approval, then a criticism might be justified.
Whatever you might think deep in your heart, whatever might be the practise of
the death penalty in your country or state, the details you had written in both
postings were inhuman, therefore unallowed approvals for fascist methods. The
practise of death penalty does exactly not include such methods as later
gasolining or dismembering of human bodies. These are murder fantasies itself.
And no democratic state would allow that. The death penalty as such is
problematic enough. But that's not the question here.
I would like to ask you if you couldn't accept this. I see a great chance for
peace in this group if you could accept that Rolf's criticism was basically a
justified and very noble reaction on your posting. On Thorsten too. But then
came the court trial without necessary evidence and then Rolf's exclusion on
CCC.
You can't rewrite this record. I checked that in dejanews.com.
>: Did you ever reflect on Rolf as a human being?
>Absolutely, at first. I tried to talk to him many times, in fact. But his
>true nature keeps returning, time after time.
I think you didn't understand me. I reflected on the past character
assassinations against Rolf, on Rolf's alleged "awful physical handicap".
"Absolutely" you said. But how could you do that without even knowing about him
personally? Of course if you defend "over and over again" the fascist methods
you once posted Rolf's "nature" will always keep returning. Only I wouldn't
call this "nature" but justified criticism, wouldn't you agree with me?
>I have written him off. I won't
>try again...
Fortunatelly I also found many "no, never again" in dejanews.com that were
decently corrected after a while.
>: What could be possible reasons
>: for Rolf's sensitivity for all sorts of threats against the intact physical
>: status? I checked that. Rolf did never talk about. But others wrote about
awful
>: physical handicaps. You really should be ashamed of your campaign against
Rolf.
>I have *no* campaign against Rolf. I pled his case on CCC. I talked to him
>here while he was lambasting others. Please go back and read thru 1 year's
worth
>of deja here... your "picture" will change...
You really have the ability to destroy the whole meaning of a paragraph. You
had no campaign, no, not at all. You did only accuse Rolf of using the nazi
term in 99% of his posts which is a false accusation, you did only accuse Rolf
of lying in all of his posts. When challenged and possibly recompensated with
100 $ you had a single example that couldn't survive a minute of thorough
linguistic analysis. But your claim was that Rolf had lied in hundreds of
cases, probably over thousand. So it was somewhat suspicious when you couldn't
find more than a single very weak example.
Then you did only post here to me half a dozen false statements about Rolf and
two confusing text demolitions.
But I'm ready to concede that you don't have a special agenda against Rolf.
That might be true. What about "a rather weak intellect" that always forced you
to overlook the deeper meaning of Rolf's criticism? Agreed Rolf's English
really "sucks" as someone once wrote. I checked that in dejanews.com.
>: I think it's not your support of the existing death penalty in the USA that
is
>: frightening for Rolf, but your atrociously violent fantasies about the
tratment
>: of human bodies. Death penalty doesn't include torture and all the
inhumanities
>: you did propose here in this chessgroup.
>: You can't rewrite history that is in dejanews.com.
>Never tried to rewrite history. I've left that to a couple of "others" here
>that seem to delight in doing so.
What if you had to rewrite history to avoid that it could happen again?
>: Why did nobody, also Robert Hyatt, react positively on Rolf's peace
proposals?
>: I checked that. All I could find were reactions like "Teuschen, lie down and
>: die".
>Perhaps, because we heard him cry "wolf" too many times? And we learned that
>it was all a load of crap?
Perhaps this was a little error in the transmission? Perhaps he cried "I'm
R-o-l-f, stop defaming me and correct the character assassinations. I need
you."
But then I'm not a psychologist.
Regards
Chris Whittington
User77568 wrote in message
<199808161827...@ladder03.news.aol.com>...
>
>Ed was the one who just had lost his court trial. The agenda seems to be
quite
>obvious. Next. Chris, I think you mean Whittington and not Dorr, was the
one
>with Ed who had rejected Rolf's peace offers with the reason that Rolf were
>Hitler. The agenda seems to be quite obvious. Next. Thorsten. Rolf had
attacked
>his "nazi-like" ideas. The agenda seems to be quite obvious. Next Moritz
and
>Dirk. Two close friends of Thorsten. The agenda seems to be quite obvious.
>Finally Enrique. The one who was angry about Rolf because he had refused to
>accept an invitation to Spain. This against the background of Rolf
supposedly
>having some "awfull physical handicaps".
>A very special and mean agenda so to speak.
>
>
[ snipped ]
>
>>Quite easy. If someone uses the term "nazi" in 99% of his posts, there is
>>obviously some deep-rooted fascination with the term, and with the
ideology.
>>What, I don't know, but it is definitely there. I rarely use that "word".
>>Rolf uses it like most of us use "the".
>
>Perhaps this is the moment where you could understand my doubts about
someone
>with a PhD 15 years after his BA.
>
>I checked that. There's no evidence for the occurence of the "term 'nazi'"
in
>"99% of his posts". If then someone frequently wrote negatively about a
certain
>topic then how this could be the proof for "obviously some deep-rooted
>fascination"? Must be sort of secret wisdom you used to achieve such
heights
>of higher nonsense. Also you forgot about Rolf being a German. Also you
forgot
>about his alleged personal concern as a physically handicapped. I checked
that.
>I did never find expressions of fascination. What I found was more
something
>like expressions out of despair, horror and terror.
>
>
[ snipped ]
>
>I think you didn't understand me. I reflected on the past character
>assassinations against Rolf, on Rolf's alleged "awful physical handicap".
>
>"Absolutely" you said. But how could you do that without even knowing about
him
>personally? Of course if you defend "over and over again" the fascist
methods
>you once posted Rolf's "nature" will always keep returning. Only I wouldn't
>call this "nature" but justified criticism, wouldn't you agree with me?
>
>
[ snipped ]
>>: Why did nobody, also Robert Hyatt, react positively on Rolf's peace
>proposals?
>>: I checked that. All I could find were reactions like "Teuschen, lie down
and
>>: die".
>
>
>>Perhaps, because we heard him cry "wolf" too many times? And we learned
that
>>it was all a load of crap?
>
>Perhaps this was a little error in the transmission? Perhaps he cried "I'm
>R-o-l-f, stop defaming me and correct the character assassinations. I need
>you."
>
>But then I'm not a psychologist.
>
>
>
I would like to add that "fca" and Hyatt were both extensively using the
expression
ROTFL
usually translated as Rolling On The Floor Laughing
but transmogrified into
Rotting On The Floor Lying
and using it to mock Rolf some weeks back.
I should have spoken out at the time and registered my disgust that this was
a mocking reference to Rolf's alleged seriously physically handicapped
status.
Chris Whittington
>Regards