ICC for CCT11

[bob]

de Groot was the first who then discovered the importance of so called chunks that a GM is using. Chunk isnt just a memorized move or a exactly defined position but something in a configuration that is defined by certain chess content (where a single pawn on a2 might be uninteresting). The GM in his play is now searching for similarities that spring to his mind which makes his chess so outstanding compared to a patzer. The difference between a GM and a patzer is that a patzer tries to play memorized moves while a GM even in a opening vartiation doent play moves but he tries to recall the priciples, the gestalt, of chess positions. Here the eidetic talent is absolutely necessary because you search into the imagined future of chess situations and to handle the details it's absolutely necessary to recall the differences of the hundreds of position of a calculation. Here a patzer must by definition blunder because there are no *moves* to recall, but there are imagined pictures of positions plus the evaluations. Chunks are always in play. - Hope this helps a bit.

Never said they were not. But "chunks" are not eidetic memory.

Bob, if you think about a position and then search for images that you might already know and then calculate a lot, you need such a memory because it's too big to handle without.

Let me also explain something to the de Groot experiments. When he showed the positions during the AVRO tournament in 1938 he didnt tell the GM what he was searching. He didnt tell them now a scrumbled position. He just showed them. He was the very first to do that. Now think about the task. The GM didnt know that there was no sense in it whatsoever. So, with his usual view the GM couldnt store the "stupid" positions.

We should try to define what eidetic means. And also I am NOT talking about extremest idiots savants. Here you already agreed that eidetics ALONE doesnt make it in chess. Bob, and now is the moment you could also agree with me that exactly this is the reason why computerchess is so disturbing. You know that you violate the FIDE rules.

Feel free to cite the specific FIDE rules you think Crafty (and other programs) violate. Not some hazy half-written approximation, but a real rule. Then I'll respond.

[bob]

Add-on. During the AMBER in CB someone, John Nunn?, gave a difficult position for his collegues. They searched and couldnt find the mate in three. Some even claimed that there was something odd. And only they were dead on because what the one had missed is that another P was on e5 so that the K couldnt walk out there. The Mate in 3 is trivial for such players but take a P away they need a lot of time before they finally state that it's impossible to solve. However if the whole puzzle had been introduced as something odd where you had to look for strange things, then I am certain a GM would imediately have said but there is a P on e5 then the solution is easy. Know wgat I mean. I hope I could show you something where the "as such" shape of a brain isnt decisive but the social situation how something is presented. In other words, the GM were ONLY weak for the other positions because they didnt know about the different sort of tasks. If they had been informed they would have ran down the 32 pieces per board in minutes. Sense or non-sensical!

-all my personal opinions (just for the friendship with Dr. Bob)

So if a doctor is informed, before the fact, that you have a small rupture in your spleen, he is a genius for finding and repairing it. But if he doesn't know beforehand, it is ok that he doesn't find it and let you die?

[bob]

I don't disagree, except for memorization. I know literally hundreds of chess players, from 1000 USCF to GMs. I have played all of these players at some point in time. And once you pass the 1500-1600 level players, they will play their openings instantly, until they reach some point where either their memorized line runs out, or I make an unexpected move that requires analysis before proceeding. To claim that humans don't memorize opening moves (and their opening analysis preparation) is the mark of a patzer, IMHO. It is just a normal part of chess. Better players know more openings and know them deeper and broader.

Rolf simply doesn't know which end is up, here. As usual.

Bob, like Gerold or whoever, all those who claim that GM, I dont talk about patzer like most here like me, *memorize* opening lines and do that for years, meant, before they are finally mature GM, all are totally wrong.

Let me explain this triviality for me as a psychologist. A potential GM is early an eidetic. Which means he can learn/memorize the known theory in half a year maximum. But he wont do it, only a patzer would do this hoping he could become a champion.

It's very telling that Bob believes in such nonsense although he claims he is in continual contact with GM.

The trith is that GM in their development must understand the moves in all the lines. This means analysing the moves with their complete middle game and endgame implications. Otherwise you cant become GM. But simply memorizing the moves would only help beat a patzer but not a true GM! Because the difference between GM is how deep they understand the game as such.

Memorizing alone doesnt mean a thing in professional chess, I mean the moves of the know theory. A true GM has his own theory that he's analysing over his whole life and he wont tell others. For sure a GM wont tell Bob what he's analysing.

Like a single game of chess is much more than just the played moves, but all the analyses around each single move.

But this is something a normal patzer cant understand. I'm a patzer but psyhological studies allowed me to understand how the brain of a GM might function.

Your point is irrelevant. A patzer is a legal chessplayer -- memorized moves and all -- no matter how weak, no matter how much it doesn't understand in its "patzerness". If a chess program plays its openings like a patzer, and the rest of the game like a GM, that's no violation of rules or ethics or anything, neither does it mean it's not a complete chessplayer.

Simple.

Please don't ruin these discussions by introducing actual facts into them.

Thanks...

Matt is totally wrong on this topic. I must insist: a chess program cheats (compared to the FIDE rules) because it looks into material, book etc, what no human player is allowed to use, in special he cant use computer help during the game. This is so basic and trivial that I cant imagine why we all missed that for over 4 decades. Computer vs computer of course is ok.

Sorry, but a computer does not look at printed material on the fly. The computer, just like the GM, pores over a book on opening moves, decides what to keep and what to discard, makes a note of winning/losing/drawing percentages, and then builds upon this information as games are played. And notice I did _not_ tell you whether I was talking specifically about the computer or the GM, because they both do the _same_ thing. Once a game starts, all they can use is their "pre-prepared analysis that they have built over a few days or a few months, along with their _memory_." You don't want a computer to use a hard drive during the game? Fine by me, I just ran on a box with 512gb of RAM. The machines I use in CCT are 12GB. My book is a few megabytes. It all fits into DRAM anyway. You are so far out in left field here, you are not even in the parking lot any longer, you are across town.

[bob]

For those who ask what this DB, Kasparov, IBM, CC, science, ethos, classical chess stuff should have to do with the topic that the actual CCT is played online on ICC, when this is now the almost authentic "play" of machines without any personal human interference/intervention - let me define what a machine in chess should be looking like before we could call it a chess player, that played chess after the rules:

for the friends of computerchess it sounds odd if I define that it's cheating if a machine simply got rammed down its throat the whole data of the history of chess when we all know well that the same machine could never understand the whole details. Technical engineers in computer sciences might say what they want but a true human chessplayer plays the lines he has analysed for years and not lines he has not understood. Because if he would do that he would lose in the longer run against his competitors in human chess.

So, the definition goes like this. Everything what a machine finds out of itself, what is within its radius, its realm, that is allowed to implement. But not something a machine would never even try or plan.

If you will get humans to play by that same rule, I'll go for it. Humans can no longer buy or study any book that contains any games of any kind. They have to learn by playing OTB, and nothing else.

And we won't have a GM in 50 years.

A machine should be a certain physical entity that once connected to energy sources, then is prepared to play chess games. The opponent can push the button of the time schedules. The rest is alone the job of the machine.

Been that way since my first computer chess event.

A machine should learn by itself, by experience of course, that there exist different opponents who play different styles. All that must be programmed, but it's not allowed to implement for the machine, what it must discover on the job so to speak. Then we have a true chess playing robot.

Also already done...

If he then begins to understand chess he can make progress in higher leagues. But shortly after his birth he cant by no means directly start a match against the World Champion. Should be trivial!

Unless the world champion agrees. And DB didn't "start" with Kasparov. It started in 1985. And played hundreds of strong players in human events, as it climbed into the official GM ranks based on its rating... Do you ever study things before spouting off? Apparently not.

Everything else, the whole practice of implementing this or that detail is fine if it's thought to remain in the artificial world of chess, but a chess playing robot must be differently trained on chess. If this doesnt succeed then it's simply impossible to create a chessplaying robot. Then the tools can still be very valuable for us human players.

In this line of arguments I strongly reject the assumption of a chess machine that has now surpassed the abilities of human chess players. Wrong! Until now it was only shown that with disregarding the rules of chess a complex of tools with the helping hand of a computational engineer can win games against a human player who is bound to respect the FIDE rules!

Friends! If you ever have played a game of chess in your school perhaps you will remember that you would have won many games if a human advisor, at best a GM, would sit next to you and whisper all the best moves. But without you blundered and lost the game. Know what I mean?

In fair competition mode you cant play against a machine. But you can well use the machine like all PC databases and tools to become better in your own play.

What is actually playing on CCT isnt a chess robot but the said "complex" consisting out of encyclopedical tomes and all kind of unallowed support.

What planet are you from? Crafty's book, for example, as used on the last CCT event is _exactly_ 1.6mb in size, and a significant part of that is not chess moves. I claim that is _far_ smaller than what most GM players know. That was produced by looking over a couple of hundred thousand games, extracting what was considered important, ignoring the rest, and then building a small database containing statistics on the results of those games. Crafty then used this book on the server to play thousands of games and "learn" from those games which of the openings it had kept are good, which are bad, and it adjusts the database based on that. Then it used that in CCT-11. Not very big. A single position is 16 bytes. the book contains about 10,000 positions. Really huge, right? It doesn't understand any of that, even though it played games and made notes about how the scores looked right out of book, and whether it won or lost the game?

You really do have no clue about this subject. Absolutely no clue.

I just got the idea of a good comparison. What actually happens in CC is this: the creators of dance choreography sit together in a room and they play their championship but without the dancers. Without dancing at all. But the different steps are analysed b computers and then they are measured on certain parameters. On display we could also see puppets that dance the given choreography.

Who on this world would now reason that this could replace true shows with true human dancers? But on display the choreographic beauty is on a level that authentic human dancers could never reach. Would we then speak of a triumph of the computer displayed dance over the actually leading couples in the sport?

Please think about better ideas to understand why the theory is flawed that sees a machine wise superiority over human records.

Reality is that computers beat humans in OTB games. There is no other measure that is important. The rules of chess do not specify how you have to choose a move, they only specify which moves are legal and which are not. The rules of chess don't specify how you learn your openings, they only define which moves are legal and which are not. The rules don't require that you "understand" why you make a move, they only require that you do make it or lose on time.

[bob]

You can insist all you want. But you are still wrong. My program has no access to printed materials. It does have access to its own enormous memory. The reason it was not missed is that most everyone "gets this". Humans can memorize openings. Computers can do it far better. Computers can't "generalize" knowledge, humans can do it extremely well.

You have tried to hijack this thread _twice_ now. It was originally about where the best location to hold CCT events is. You turned it into a deep blue rant, and now a computers-cant-play-legal-chess rant.

why don't you just fade away???

Stop spreading untrue allegations, Bob!

The CCC actually has some thirty (30) threads on page 1 the most actual one. In a single (1) thread I comment because I have my personal opinions and you then accuse me of disturbing this here in CCC? Why do you answer me at all? Only to insult me or because I have something to say that you want to discuss? I thought we had the same rights with our opinions on chess.

The you are also wrong that my positions have nothing to do with the topic of the CCT on ICC which allegedly is a progress because now the play is not depending on operating hands. Coming from the DB2 debate I proposed the creation of a playing machine entity that is completely independent of cheating possibilities. Therefore the topic DB2 vs Kasparov with the mistreatment against Kasparov by the IBM operators is important. Becuse it speaks against these computerchess people who have lost contact with human chess and FIDE.

Secondly the question of legal FIDE chess. Yes, I claim that the actual computer players are violating the rules of FIDE. But I am sure that entities could be created where a development happens from scratch to perhaps GM status over some months and years. These entities can then play in human tournaments. For now this is impossible because the machine now have extra rights which is called cheating in human chess because they have illegal help during a game. It's all so easy.

Just quote the FIDE rule(s) we are violating. I have asked for this a half-dozen times. You _never_ respond. Most of us can conclude why that is...

[bob]

You have tried to hijack this thread _twice_ now. You turned it into a deep blue rant, and now a computers-cant-play-legal-chess rant. why don't you just fade away? Would you please stop posting such nonsense. You have burned out my bullshit detector so many times, I can't convince anyone to buy me another one. It is _impossible_ to eliminate all cheating possibilities. It is just yet another attempt to start a long and fruitless discussion about nothing. Which is about the only thing you are qualified to discuss herein. At least _their_ arbiters agree, and that carries a lot more weight than your irrational opinion on the subject.
There are medications to treat this kind of schizo-behaviour, you should see what you can find and come back down to planet earth, from wherever it is you are...

What I quote here is symbolic for Bob Hyatt writing stuff if he wants to attack someone ad hominem because he had opinions different to Bob Hyatt. This is just ugly as output of an academic. Because if it were only allowed to write Hyatt stuff then the rest here could stop posting and Bob could dictate what is the "truth". This is just laughable because we have a debate here and we take part in our free decision. I dont get it why Hyatt is entitled to defamate certain opinions! IMO this is against the charter of CCC which forbids personal attacks of other members.

In addition to your many other issues, you also don't understand "ad hominem attacks" either. Start at the beginning of what I wrote. Nothing "ad hominem" there at all. I do believe you have some sort of deep-rooted personality disorder, as evidenced by your continual off-topic rants here...

[mhull]

Matt is totally wrong on this topic. I must insist: a chess program cheats (compared to the FIDE rules) because it looks into material, book etc, what no human player is allowed to use, in special he cant use computer help during the game. This is so basic and trivial that I cant imagine why we all missed that for over 4 decades. Computer vs computer of course is ok.

I copy moves from Horowitz, NewInChess or interesting games to my internal storage (neurons) and "play openings like a patzer". Computers copy moves from the same sources into their internal storage and play them in the same way. This is so trivial that I can't imagine how you missed it.

A GM doesnt play this way, Matt, excuse me. A GM I have mentioned it many times plays variations he has analysed, understood and evaluated. Patzers and programs just copy what they have stored which is simply non-creative chess.

Matt, perhaps you are a bit cooler than Bob. Tell me what you think about this scenario:

your computer uses his stored human knowledge what he cant understand. I as opponent have tomes of opening books in my different storages. I have three secretaries who support me during a game. En plus I have three computers who do analyses of the position. In the ending I use table bases.

Are you ready to play me? Do you agree with me that I violate the FIDE rules of human chess?

But if I do it then all the computers do it too, all the Fritz and Crafty of this World. If they play against each other then it's ok, but not if they imposter that they could have any rights to challenge our best human chess players, our super GM. These computers are made to help us human players in our preparation. But everything else is nonsense and unfair and evil.

You haven't addressed my points at all, but instead repeated your own points, which my points have already addressed. I refuted your GM/patzer argument by showing that patzer play is legal chess. I refuted your cheating argument by the definition of respective internal storage media of humans and computers.

Any player, human or mechanical, which accesses its own internal storage is not cheating.

[Rolf]

You haven't addressed my points at all, but instead repeated your own points, which my points have already addressed. I refuted your GM/patzer argument by showing that patzer play is legal chess. I refuted your cheating argument by the definition of respective internal storage media of humans and computers.

Any player, human or mechanical, which accesses its own internal storage is not cheating.

Indeed I have addressed it. Hint: with my scenario. But Matt, do you really plan to enter this here and then talking with a deeply rooted disorder? What do you expect here?

Ok, just kidding. Since you have abstained from any insults I will answer you of course. You say internal storage. Do you really mean it? Then human players play with implantated chips for endgame tablebases. I prefer the modul mode. I get the necessary data for the next opponent the afternoon before the game. This is all ok for you? IMO all this would violate the eternal FIDE rules of chess. But computers use all that and you call it perfectly normal internal storage, right?

Therefore I would propose a true robot player for tournaments like CCT. Computers who only use selfdiscovered lines. Without any influence of programmers. He will program the feature as such that of the lines finding. The robot remains the same for 6 months. Of course it can itself improve his play after own analyses before and after tournament games. In these 6 months the different players (C) live in neutral zones outside the realm of the programmers. Then the FIDE rule cheating would stop. And that of the ELO number hystery.

All just IMO, 100x excuses.

[bhlangonijr]

You can insist all you want. But you are still wrong. My program has no access to printed materials. It does have access to its own enormous memory. The reason it was not missed is that most everyone "gets this". Humans can memorize openings. Computers can do it far better. Computers can't "generalize" knowledge, humans can do it extremely well.

You have tried to hijack this thread _twice_ now. It was originally about where the best location to hold CCT events is. You turned it into a deep blue rant, and now a computers-cant-play-legal-chess rant.

why don't you just fade away???

Stop spreading untrue allegations, Bob!

The CCC actually has some thirty (30) threads on page 1 the most actual one. In a single (1) thread I comment because I have my personal opinions and you then accuse me of disturbing this here in CCC? Why do you answer me at all? Only to insult me or because I have something to say that you want to discuss? I thought we had the same rights with our opinions on chess.

The you are also wrong that my positions have nothing to do with the topic of the CCT on ICC which allegedly is a progress because now the play is not depending on operating hands. Coming from the DB2 debate I proposed the creation of a playing machine entity that is completely independent of cheating possibilities. Therefore the topic DB2 vs Kasparov with the mistreatment against Kasparov by the IBM operators is important. Becuse it speaks against these computerchess people who have lost contact with human chess and FIDE.

Secondly the question of legal FIDE chess. Yes, I claim that the actual computer players are violating the rules of FIDE. But I am sure that entities could be created where a development happens from scratch to perhaps GM status over some months and years. These entities can then play in human tournaments. For now this is impossible because the machine now have extra rights which is called cheating in human chess because they have illegal help during a game. It's all so easy.

Just quote the FIDE rule(s) we are violating. I have asked for this a half-dozen times. You _never_ respond. Most of us can conclude why that is...

Probably he is talking about FIDE chess laws article 12:

http://www.fide.com/component/handbook/ ... ew=article

Article 12: The conduct of the players
12.1

The players shall take no action that will bring the game of chess into disrepute.
12.2

1.

During play the players are forbidden to make use of any notes, sources of information, advice, or analyse on another chessboard.
2.

It is strictly forbidden to bring mobile phones or other electronic means of communication, not authorised by the arbiter, into the playing venue. If a player`s mobile phone rings in the playing venue during play, that player shall lose the game. The score of the opponent shall be determined by the arbiter.

12.3

The scoresheet shall be used only for recording the moves, the times of the clocks, the offers of a draw, matters relating to a claim and other relevant data.
12.4

Players who have finished their games shall be considered to be spectators.
12.5

Players are not allowed to leave the `playing venue` without permission from the arbiter. The playing venue is defined as the playing area, rest rooms, refreshment area, area set aside for smoking and other places as designated by the arbiter.
The player having the move is not allowed to leave the playing area without permission of the arbiter.
12.6

It is forbidden to distract or annoy the opponent in any manner whatsoever. This includes unreasonable claims or unreasonable offers of a draw.
12.7

Infraction of any part of the Articles 12.1 to 12.6 shall lead to penalties in accordance with Article 13.4.
12.8

Persistent refusal by a player to comply with the Laws of Chess shall be penalised by loss of the game. The arbiter shall decide the score of the opponent.
12.9

If both players are found guilty according to Article 12.8, the game shall be declared lost by both players.

But he's completely wrong. Those are not applicable to computers. There are specific regulations to computer chess players that not include any restriction about reading any information in its own storage device.

Therefore, chess engines are not violating any FIDE rules - at least of course outside the world this person have created in his own head.

[mhull]

You haven't addressed my points at all, but instead repeated your own points, which my points have already addressed. I refuted your GM/patzer argument by showing that patzer play is legal chess. I refuted your cheating argument by the definition of respective internal storage media of humans and computers.

Any player, human or mechanical, which accesses its own internal storage is not cheating.

Indeed I have addressed it. Hint: with my scenario.

My points address your scenario.

But Matt, do you really plan to enter this here and then talking with a deeply rooted disorder? What do you expect here?

Ok, just kidding. Since you have abstained from any insults I will answer you of course. You say internal storage. Do you really mean it? Then human players play with implantated chips for endgame tablebases.

Oh, so you concede the point on openings, since you're changing the subject to EGTBs. But EGTBs are also copied to internal storage.

I prefer the modul mode. I get the necessary data for the next opponent the afternoon before the game. This is all ok for you? IMO all this would violate the eternal FIDE rules of chess. But computers use all that and you call it perfectly normal internal storage, right?

Therefore I would propose a true robot player for tournaments like CCT. Computers who only use selfdiscovered lines. Without any influence of programmers. He will program the feature as such that of the lines finding. The robot remains the same for 6 months. Of course it can itself improve his play after own analyses before and after tournament games. In these 6 months the different players (C) live in neutral zones outside the realm of the programmers. Then the FIDE rule cheating would stop. And that of the ELO number hystery.

All just IMO, 100x excuses.

Computers have self-discovered all the endings in EGTBs. And just like humans, they share their knowledge with other computer players by publishing their findings. All programs using those findings (EGTB) have copied them to their internal storage, just like they do with opening books, just like humans learn openings and endings and store the shared knowledge internally to themselves.

All your arguments have now been addressed.