Dead Man's Chess

[Mark Felt]

If I had months to paint a picture, access to Rembrandat's work, multiple canvasses to work on, and a few books on painting, my paintings would still look like **** because I can't draw worth a damn, and no amount of time or practice will ever erase my complete lack of talent.

Your analogy is crap. Among other things, painting isn't chess. Let's boil it down.

If you had two months between finalising your brushstrokes, and the ability to visualise seven or eight brushstrokes past that, which is where the chessboards come in, then yes, you could imitate Rembrandat's work pretty damn closely. The difference between grandmasters and decent chessplayers isn't that grandmasters are fundamentally "better" at chess, but that they are faster to recognise the best moves, more confident, and more experienced. They know more.

You would get your ass handed to you.

Yes, in straight up play. Not in a playbymail game with a move every two months, if I had what I said before.

Have you found a case yet where a GM lost any kind of 1-1 game against a casual player?

At the table? No, and I doubt I ever will. Again, how is that relevant? We're talking about a game played by mail over 8 years, with moves once every two months, not one played face to face.

Oh, and assuming you're right, what exactly did Rollans expect to gain for his years of painstaking chess research? Besides an untimely death at the end of the game, that is.

Why do you assume that Rollans was behaving rationally?

Apparently not if you believe you can simply read some books, study your opponent, spend a long time thinking ahead and expect to play "an extremely close game" against the 3rd best player in the world.

Again, at the table. If that group had had two months between moves, I don't find it inconcievable that they could have beaten the GM.

Malerin, you are being intentionally obtuse, now. And Mark Felt is absolutely right. You are moving the goalposts.

The point I was making about losing in group demonstrations was not that they wouldn't win most matches, but that they could lose to someone BECAUSE THE TIME CONSTRAINTS FORCED THEM TO MAKE RUSHED ILL-ADVISED MOVES. Are you following the conversation at all?

That was the flip side of the coin to the discussion of a average to fair to good player being able, allowed sufficient time and any form of assistance whatsoever, to come up with optimal moves over an eight year game.

You even found results that show that your GM lost 3 games to club players. I cited the games I saw one lose at an exhibition here. WE WERE DISCUSSING EXHIBITIONS. You are the one who contended that a GM could never lose to an average player (FIDE Master is sort of average, by the way). I (we) contend that exhibitions, chess by mail, chess hopping on one foot, etc... are all casual/fun games and offer differences that can (not "must"... but can) give either party an advantage.

So now you have moved the goalposts and are holding that you only mean in "legitimate" tournament type competition? You realize of course that you're nailing your own coffin shut with that move? In short.... You are citing a goofy chess-by-mail game with a ghost. Just how are we supposed to judge that by Elo/FIDEstandards?

BTW, when someone says something like "I take/took the game seriously" in a general discussion, it does not mean a specific game/match. I was referring to the fact that I played chess often and seriously and was working towards getting a rating when life intervened (I had a kid to raise).

And I say you don't play the game because you take this ratings stuff far too seriously. Chess players don't make a lot on product endorsements or tournament wins. They make a few bucks lecturing and giving exhibitions and writing/solving chess problems for publications. It's not exactly the old NFL adage (on any given Sunday), but there are numerous instances of people losing (in "friendlies" or "gimmick games") to people so far below them that it would amount to an average player (me) against the ghost of Maroczy... And yes, I could win.

In fact, get the ghost of any known player, channel him/her through Sylvia or Edwards, and I bet you right now a thousand Euros that over a maximum of 48 moves and eight years of time (and no one to check on what I'm doing)and I will beat him. I mean this should be easy. Ghosts come back and play chess all the time in your world. I'm game. As soon as you get the ghost on the line PM me.

Okay, so my post was wasted. Nice.

 

[Foolmewunz]

Apparently not if you believe you can simply read some books, study your opponent, spend a long time thinking ahead and expect to play "an extremely close game" against the 3rd best player in the world.

Yes, and I've just bet you a thousand Euro on it. You really don't understand the game enough to take this wager. Actually, you could get a live grandmaster to take the challenge. Tell him/her that the only condition is that I get a month between moves and that I will not be on the honour system and have access to a Cray supercomputer which is running Deep Fritz, and I will acquire all updates to it or alternately use the latest version of Rybka that is available over those eight years.

 

[Foolmewunz]

snip


Okay, so my post was wasted. Nice.

No, we're both just arguing the same sort of obvious points to someone who can't or won't get them. (You can take my place in the match against Capablanca's ghost or Anand - I believe he's the highest rated player now - since no one can check behind the curtains, we shift off any time we want. )

 

[Foolmewunz]

Aside:

We used to have a bunch of serious chess players on these forums, but I don't know if any are still lurking.

I had heard of the old flim flam routine back in the 60s. There are chess hustlers out there (see Washington Square Park - chess corner - they can point out the money players), some with fairly high ratings, if not at Tal's level.

The demo that Malerin mentions is a scam that could win you No. Players times $1.00 or times $5.00 quite easily. Supposedly it was common in Europe, although it could've been pre WWW urban legend.

It goes something like this. Visiting GM plays a few games at a good chess club in a town he's visiting. Maybe even a demo. (This is all just so that he can get a feel for the club and its members' abilities.) After probably drawing or losing a few games that he shouldn't, comes the hustle...

"Here, I'll tell you what.... your whole hundred and ten person club against me. You guys get to play white, and can have three minutes to vote on every move... All of you. Between a hundred and ten guys, surely you'll come up with the right moves. A draw goes to you, though, so all you have to do is draw. How about say I give you a buck each if I lose or draw, but you give me give bucks each if I win?"

It's a perfect con. Why? They are average players with no great high poobah of chess in their ranks. He makes sure of that in advance. And they are going to vote, amongst 110 people and come up with average moves. Invariably, to take down a GM you have to have one or two inspired moves that look like they're going to get you killed. 110 average players won't generally come up with that move. And, like all con artists and hucksters, he guides them away from the correct move or two if they should stumble upon them.

 

[Malerin]

Malerin, you are being intentionally obtuse, now. And Mark Felt is absolutely right. You are moving the goalposts.

The point I was making about losing in group demonstrations was not that they wouldn't win most matches, but that they could lose to someone BECAUSE THE TIME CONSTRAINTS FORCED THEM TO MAKE RUSHED ILL-ADVISED MOVES. Are you following the conversation at all?

But this is such a red herring. Korochoi was under NO time restraints. He had weeks to evaluate each move.

That was the flip side of the coin to the discussion of a average to fair to good player being able, allowed sufficient time and any form of assistance whatsoever, to come up with optimal moves over an eight year game.

You still have not provided any evidence at all that a casual player has ever beaten a GM in a 1-1 game. Playing 30 people at the same time is another red herring because Korochoi was PLAYING ONLY ONE PERSON. I thought you were the Google Master?

You even found results that show that your GM lost 3 games to club players.

Yes, playing 30 other people AT THE SAME TIME. Again, totally unrelated to what Korochoi's situation.

I cited the games I saw one lose at an exhibition here. WE WERE DISCUSSING EXHIBITIONS.

One of the games you claimed you saw involved a grandmaster, the other a master. No names. Your story is apocrphyal and unverifiable. That's supposed to be bad evidence, according to you guys.

You are the one who contended that a GM could never lose to an average player (FIDE Master is sort of average, by the way). I (we) contend that exhibitions, chess by mail, chess hopping on one foot, etc... are all casual/fun games and offer differences that can (not "must"... but can) give either party an advantage.

And I have yet to see any proof that a casual player has ever defeated a GM 1-1. Use your mad Googling skilz and find some!

So now you have moved the goalposts and are holding that you only mean in "legitimate" tournament type competition? You realize of course that you're nailing your own coffin shut with that move? In short.... You are citing a goofy chess-by-mail game with a ghost. Just how are we supposed to judge that by Elo/FIDEstandards?

I'm moving the goalposts to exclude games that have nothing in common with the game Korochoi played (e.g., playing 30+ people simultaneously). You're right, it's completely unreasonable to demand evidence remotely resemble the question at hand. If you could find a GM who had a heart attack in the middle of a game and accidently knocked over his king, I'm sure you would rejoice, "Aha! See, a GM CAN lose to a lowly ranked player!".

And I say you don't play the game because you take this ratings stuff far too seriously. Chess players don't make a lot on product endorsements or tournament wins. They make a few bucks lecturing and giving exhibitions and writing/solving chess problems for publications. It's not exactly the old NFL adage (on any given Sunday), but there are numerous instances of people losing (in "friendlies" or "gimmick games") to people so far below them that it would amount to an average player (me) against the ghost of Maroczy... And yes, I could win.

Then I'm sure you can find one. And you may not make as much playng chess as poker, but to discount tournaments as sources of both pride and money is absurd. Veselin Topalov, top player in the world, was a pretty busy guy in 2007 (and 2008). In 2004, the winner of the World Chess Championship walked away with about $100,000. I wouldn't turn that down. Would you? And to claim GM's don't take their ratings seriously because there's not millions in prize money ignores the fact that in nearly every sport, ratings and standings are jealously coveted because people like to be recognized as the BEST at things.

 

[Mark Felt]

I'm moving the goalposts to exclude games that have nothing in common with the game Korochoi played (e.g., playing 30+ people simultaneously).

Then why aren't you excluding every game that was played at the table?

 

[Malerin]

Your analogy is crap. Among other things, painting isn't chess. Let's boil it down.

If you had two months between finalising your brushstrokes, and the ability to visualise seven or eight brushstrokes past that, which is where the chessboards come in, then yes, you could imitate Rembrandat's work pretty damn closely. The difference between grandmasters and decent chessplayers isn't that grandmasters are fundamentally "better" at chess, but that they are faster to recognise the best moves, more confident, and more experienced. They know more.



Yes, in straight up play. Not in a playbymail game with a move every two months, if I had what I said before.



At the table? No, and I doubt I ever will. Again, how is that relevant? We're talking about a game played by mail over 8 years, with moves once every two months, not one played face to face.



Why do you assume that Rollans was behaving rationally?



Again, at the table. If that group had had two months between moves, I don't find it inconcievable that they could have beaten the GM.



Okay, so my post was wasted. Nice.

The two of you are so clueless it would be hysterical if this were not a supposed "skeptics" forum. Now it's just sad. Without a scintilla of proof given, on pure supposition, both of you claim, with the fervitude given to claims about "truths" in the Bible, that you can play a GM to a draw (or even a close game) by literally reading a few books, studying past games and concentrating "real hard" for a "long time". You can't produce one game where this has ACTUALLY HAPPENED, yet you assert it with the same force as a mathematical axiom.

Seriously, what do you REALLY think?

 

[Jimbo07]

I will... use the latest version of Rybka that is available over those eight years.

You wouldn't need a month between moves!

Right now, the wins on the very best computer are still giving and taking with very best humans. Eight years from now, I expect computer wins will be consistently decisive.

Wiki has a cool list:

# 1997, Deep Blue wins a six-game match against Garry Kasparov.
# 2002, Vladimir Kramnik draws an eight-game match against Deep Fritz.
# 2003, Kasparov draws a six-game match against Deep Junior.
# 2003, Kasparov draws a four-game match against X3D Fritz.
# 2005, Hydra defeats Michael Adams 5.5-0.5.
# 2005, a team of computers (Hydra, Deep Junior and Fritz), wins 8.5-3.5 against a rather strong human team formed by Veselin Topalov, Ruslan Ponomariov and Sergey Karjakin, who had an average ELO rating of 2681.
# 2006, the undisputed world champion, Vladimir Kramnik, is defeated 4-2 by Deep Fritz.

Looks like it's slowly starting to tilt toward the computers...

 

[Mark Felt]

The two of you are so clueless it would be hysterical if this were not a supposed "skeptics" forum. Now it's just sad. Without a scintilla of proof given, on pure supposition, both of you claim, with the fervitude given to claims about "truths" in the Bible, that you can play a GM to a draw (or even a close game) by literally reading a few books, studying past games and concentrating "real hard" for a "long time". You can't produce one game where this has ACTUALLY HAPPENED, yet you assert it with the same force as a mathematical axiom.

Seriously, what do you REALLY think?

Woah, woah, woah, back the fun bus back the funk up.

both of you claim, with the fervitude given to claims about "truths" in the Bible, that you can play a GM to a draw (or even a close game) by literally reading a few books, studying past games and concentrating "real hard" for a "long time".

Yes, we claim exactly that. What decides matches is time. With no time limit as such, with access to every match available by the GM in question, with access to games played by a great many other GMs(I'd need that, I don't know if Foolmewunz would), yes, we could play any given GM to a draw. We would lose flat out on our face in any game with a time limit, but with two months to prepare for each move, we're at a great advantage compared to a normal game. How long does a GM need to come up with the perfect move for any given situation? Not very long. How long to come up with the perfect seven moves in advance? Longer, but still a split second compared to any other level of play. If we nullify that tremendous advantage, all he has over us is his knowledge and experience, and those are both replicable and able to be studied.

You can't produce one game where this has ACTUALLY HAPPENED, yet you assert it with the same force as a mathematical axiom.

Have you been reading our posts at all, ever? Neither of us has ever denied that in a straight face to face game, we would lose to a GM, and so would any other player not ranked as a GM. This wouldn't be such a game.

 

[Malerin]

You wouldn't need a month between moves!

Right now, the wins on the very best computer are still giving and taking with very best humans. Eight years from now, I expect computer wins will be consistently decisive.

Wiki has a cool list:

# 1997, Deep Blue wins a six-game match against Garry Kasparov.
# 2002, Vladimir Kramnik draws an eight-game match against Deep Fritz.
# 2003, Kasparov draws a six-game match against Deep Junior.
# 2003, Kasparov draws a four-game match against X3D Fritz.
# 2005, Hydra defeats Michael Adams 5.5-0.5.
# 2005, a team of computers (Hydra, Deep Junior and Fritz), wins 8.5-3.5 against a rather strong human team formed by Veselin Topalov, Ruslan Ponomariov and Sergey Karjakin, who had an average ELO rating of 2681.
# 2006, the undisputed world champion, Vladimir Kramnik, is defeated 4-2 by Deep Fritz.

Looks like it's slowly starting to tilt toward the computers...

How long before they pass a Turing Test?

 

[Foolmewunz]

The two of you are so clueless it would be hysterical if this were not a supposed "skeptics" forum. Now it's just sad. Without a scintilla of proof given, on pure supposition, both of you claim, with the fervitude given to claims about "truths" in the Bible, that you can play a GM to a draw (or even a close game) by literally reading a few books, studying past games and concentrating "real hard" for a "long time". You can't produce one game where this has ACTUALLY HAPPENED, yet you assert it with the same force as a mathematical axiom.

Seriously, what do you REALLY think?

You're taking this kind of personal, aren't you? What kind of reaction do you expect from a skeptics' forum where the constant cry is for "evidence"?

Other than that, so's we don't seem to be ganging up on you.... What Mark Felt Just Said.

 

[Foolmewunz]

You wouldn't need a month between moves!

Right now, the wins on the very best computer are still giving and taking with very best humans. Eight years from now, I expect computer wins will be consistently decisive.

Wiki has a cool list:

# 1997, Deep Blue wins a six-game match against Garry Kasparov.
# 2002, Vladimir Kramnik draws an eight-game match against Deep Fritz.
# 2003, Kasparov draws a six-game match against Deep Junior.
# 2003, Kasparov draws a four-game match against X3D Fritz.
# 2005, Hydra defeats Michael Adams 5.5-0.5.
# 2005, a team of computers (Hydra, Deep Junior and Fritz), wins 8.5-3.5 against a rather strong human team formed by Veselin Topalov, Ruslan Ponomariov and Sergey Karjakin, who had an average ELO rating of 2681.
# 2006, the undisputed world champion, Vladimir Kramnik, is defeated 4-2 by Deep Fritz.

Looks like it's slowly starting to tilt toward the computers...

They're already starting the computer v computer competition. I really don't follow the developments enough (it's a whole different sub-set of chessfreaks), but I'd heard for a couple of years that Rybka had a far higher Elo rating than Deep Fritz, and even consistently higher than humans, but I read somewhere in back-checking something for this thread that Fritz 10 v Rybka 3.2(? not sure and now can't find it) has Fritz out on top like 4 to 2. The computers will probably be asking for show up fees like tennis and golf pros, soon.

I think the difference in computers now is that they still go for the massive numbers crunching of moves, but also have better weighting for victory than before. The original programs all went for control of the center of the board, as there is such an obvious mathematical/numerical advantage in doing so. Such control, though tends towards draws. Live players could beat them by working around that known silicone preference. Now the programs look for winning conditions two or twenty moves ahead.

 

[ddt]

So it applies on both sides of the game. Winning or losing. I can play a pretty fair game of chess, but Korchnoi's third cousin four times removed could beat me in regulation time. But given a week between moves, access to books and friends for advice (and even a primitive 80s chess computer), and I'm pretty sure I could play him to a draw. And I was never ranked. (Never played enough tournaments when I was serious about the game.)

Has anyone here actually looked at the game? I don't think anyone yet posted a link, so here it is.

I only ever played chess recreationally, so don't take my word for it, but a number of things about the game struck me:

1) Korchnoi wins the pawn advantage in move 18 and doesn't let it go. By that time, all bishops and knights are out of the game;
2) in moves 19/20, the queens are captured, so both players are left with 2 rooks and pawns.
3) in moves 25/26, both players capture one of the other's rooks.

Point (1) is still in the opening phase; even by 1980s standards, opening theory extended to move 15-20, IIRC. Point (2) essentially means we're now in the endgame and we skipped the middle game altogether. At least from point (3) on, we're again in charted endgame theory about how to take advantage of an extra pawn. Correct me if I'm wrong on these counts.

I'm only an amateur, but I couldn't see where Korchnoi was in trouble during the game. He seemed quite eager to get the pieces as quick as possible from the board and go to the endgame.

So as I see it, the non-GM player could get most of the game from the existing chess literature, and he had 1 month per move to think about it. Seems quite enough to me.

 

[Cuddles]

But this is such a red herring. Korochoi was under NO time restraints. He had weeks to evaluate each move.

And you really think that a grandmaster had nothing better to do in those weeks other than study a friendly mail game with a nobody?

Playing 30 people at the same time is another red herring because Korochoi was PLAYING ONLY ONE PERSON.

Really? You honestly believe that a grandmaster chess player didn't play a single other game for 8 years? Really? As others have already said, it's painfully obvious that you're being deliberately stupid. Please stop it and actually try to participate in a sensible conversation.

 

[Malerin]

And you really think that a grandmaster had nothing better to do in those weeks other than study a friendly mail game with a nobody?



Really? You honestly believe that a grandmaster chess player didn't play a single other game for 8 years? Really? As others have already said, it's painfully obvious that you're being deliberately stupid. Please stop it and actually try to participate in a sensible conversation.

Because playing 40 people at the same time in 6 hours is the same as playing multiple games over 8 years.

http://www.marconews.com/news/2008/apr/19/forty-players-no-match-chess-grandmaster-marco-eve/

You're absolutely right. All the baseless suppositions are true. In fact, lets go further: a toddler could have beaten Korochnoi. No, wait! Moving the pieces randomly with no knowledge of the game whatsoever would have been enough. What GM would even look at the board in a match spanning 8 years, am I right?

Since you're completely making **** up at this point, let's go for the brass ring: Korochnoi COULDN'T play chess at all! Through an astonishing coincidence, all of his random moves just happened to catapult him to 3rd place in the world. Fortunately for Korochnoi, his luck held up, and he was able to fend off the medium'ss vicious onslaught and win the game.

True story.

 

[Mashuna]

Koronchoi couldn't play chess. He just channelled a grand master to make his moves for him. However, in this case, someone else happened to channel the same grand master. That's why the match ended in a draw.

 

[Civilized Worm]

And then there is of course this: Who wants to be known as the Grandmaster who lost to a ghost?

If he was worried about being known as the grandmaster who lost to a ghost he wouldn't have accepted the challenge in the first place.

So I think the naturalistic explanation is weak on motivation. This kind of hoax would take quite a lot of work to pull off (finding a high-ranking player to advise you, staying with the game for 8 years, and quite a bit of research). If the only reasonable explanation is "Hey, look how I fooled everyone", why did Rollans go to his grave with the secret? Don't people usually brag when they pull off a big hoax just for the hell of it? If Eisenbeiss was the mastermind, why has HE kept quiet all these years? HE certainly gained nothing from the game. And if its all a fake, why drag the game out for 8 years? The hoax would be just as valid if the game had finished in 1 year, and that would have left plenty of time to perpetuate OTHER hoaxes on unsuspecting GM's. But no attempt was ever made again.

Having successfully channelled a dead grandmaster would have looked good on any mediums CV would it not? I'm sure it must have upped his profile in medium circles. Was Rollans in the business of giving private readings? How much did he charge?

I never said the natural explanation was impossible. I pointed out that it was weak in certain areas, namely motivation and lack of explanation for the length of the game.

And you're right, we'll never know the full explanation for what happened. All we can do is see what explanations would be most likely without having to resort to unproven phenomena.

Is there a single case of Houdini exposing a medium who worked for FREE?

Why would he? Like James Randi, Houdini set his sights on the most high profile mediums, and they charge.

The two of you are so clueless it would be hysterical if this were not a supposed "skeptics" forum. Now it's just sad.

OMG so abusive wah wah wah!

 

[Jaggy Bunnet]

I'm not claiming it's impossible for a below 1800 rated player to defeat a Grandmaster. Just that it is very very unlikely.

Actually, your claim was:

When Grandmasters lose, they are invariably defeated by other grandmasters or prodigies who later become Grandmasters themselves (David Howell, Fabiano Caruana).

You didn't sat unlikely, you said that when they lost it was "invariably" to another grandmaster or a prodigy on their way to becoming a grandmaster.

Can you become a grandmaster only paying chess by mail?
If not, why not?

Answer those honestly and you start to understand how unimpressive your evidence is.

 

[Ersby]

This was discussed on a chess site in March:

http://www.chessgames.com/perl/chessgame?gid=1486372

My favourite quote:

I have no doubt in mind that was Maroczy playing white. I´m not so sure that Korchnoi was black though...

 

[YeahDude]

Malerin,

I have one simple question to ask you. Please answer it. What is the more likely scenario:

A ghost of a grandmaster was submitting moves via a mdeium to play a game of chess over an 8 year period.

or

A clever person realized that he could fool people by pretending to be channeling the 'ghost' of a grandmaster, but instead used some form of cheating to complete the match.

Simple and direct question, which scenario is more likely?