I do kno that if you are playing a Master quite a bit above you, and he is playing several dudes at a time- blind: then making some odd opening moves can throw him off.
I just tried this strategy against Chessmaster9000 and got my butt kicked.
Thanks Duck, Glee 
Top chess players don’t examine each piece to see what it can do. This would take too long (and be really boring). Remember they are trying to analyse several moves ahead, which, using ‘brute force’ would involve looking at say 1000 positions just to see one move ahead. Grandmasters lop off all but 3-4 possible moves each time, using their judgement.
What they do instead is to use patterns (pieces arranged in a certain way), knowledge of similar positions, strategy and a general feeling of how well they are doing.
In this thread (posts 32 + 40) I comment further on Kramnik’s blunder:
This is the computer that beat Grandmaster Michael Adams 5.5 - 0.5. Michael is the English no 1 and ranked in the World top 20. :eek:
Hydra runs on purpose built hardware, but perhaps a version for the PC will emerge. They like to keep the openings book + other stuff secret, so this may not happen for a while.
This thread is very disappointing to me as a serious chess player.
The original thread title and allegations were complete rubbish :eek: (and of course no evidence was ever produced).
Now the above claim is not quite so bad. 
But what is your evidence?
Do you really mean blindfold simultaneous displays?
(These are quite rare, but here’s a few.)
Grandmaster v Guernsey club; 6-0
Grandmaster v French celebrities; 11.5-0.5
I’ve done this myself (only two boards though!) and I assure you that weak play is easy to beat.
One of my mates (former British Champion) does this about once a year and again has no trouble with unorthodox openings.
I suppose if all the players were in a conspiracy and deliberately all played a similar opening there might be some confusion. But that’s just testing the experts memory, not his chess skill.
No quotes, but I believe it is common enough that someone can provide instant ratification from the top of his head. Chess masters have a much easier time remembering piece placements that are the product of a game than just random piece placements. This would add a bit of substance to the theory of foolish play confusing a blindfolded chess master. Not so for a computer, of course, which should remember any piece placement just the same.
Stupid question – how does one play blindfolded? I can only assume touching the opponent’s pieces (to guage position) is allowed?
I believe the moves are announced to the blindfolded player in chess notation.
No touching is allowed. It’s all done by memory and you call out your moves. (edit: Rather, the moves are called out by an intermediary usually.)
This study was done sometime before 1975, since I learned about it from people doing computer chess at Illinois around then.
Non-chess masters remembered real positions and random placements equally well, while masters remembered real positions much better than random placements.
Knowing a tiny bit about the internals of chess programs (though I suck at chess) I can’t conceive of how a strategy like given in the OP would work, unless the program was so badly designed that it got confused if an opening was not in its library. Perhaps it would work if the program were set at a level where I could beat it.
Karpov-Miles 1…a6, maybe. 
It’s true that there are specific things that chess programs are bad at, and types of positions where they can be made to look stupid. However, the method in the OP would only work against a very weak program, which should be easy to defeat using normal play as well.
There’s some nice information on Tim Krabbe’s “Defending Humanity’s Honor” page about this kind of thing. It includes a few examples of tricking a computer by giving up material for a long-term attack. The computer can’t see the strength of the attack because it takes many moves to develop, but a strong human player can see quite quickly that the attack is inevitable.
One instance of that quote is from Twain’s A Connecticut Yankee in King Arthur’s Court, when the protagonist’s clever scheme is foiled by the incompetence of his adversaries.
I remember i saw somewhere were an average joe beat 10 chess champs…It was real brilliant. He sat them all in a circle and had them face outward. He then went around the circle making his moves… The thing is… he was actually having them play EACHOTHER! he would just remember the moves each one did the play them on the opposite side of the circle
Yes. And I wasn’t talking about “weak play” I was talking about an unusual opening. Sure, in general those are weaker. I don’t know this myself, just what I was told by a regular chess Master who used to attend and play at several such exhibitions.
I assure you Derren Brown is not an ordinary bloke. 
Of course we knew what he was going to do because there are only 3 ways to beat 9 strong players (including 4 Grandmasters) simultaneously:
- bribe the players (sadly he didn’t)
- get Kasparov or Hydra to play for you (costs money!)
- use the mirror chess trick
He did two other pieces of magic.
The first was an envelope switch, which was excellently done.
The second was to beat the 9th player. Now Derren is not a strong player (to put it mildly), so he must have been fed moves. But we didn’t see anything during the 3+ hours of transmission…
From trying this myself:
- non-chess players only remember a few pieces in either case
- strong chess players can remember chess positions perfectly after say 10 seconds; they would struggle to remember 50% of placements in a random position
The explanation is that top players use patterns a lot. A particular pawn formation in the centre or a group of pieces defending a King can be spotted immediately as previously known.
Miles told me Karpov was really annoyed afterwards, both at Miles for doing it and at himself for letting it get to him.
If I remember correctly, the opening quickly transposed into a Sicilian-type formation, so Karpov really should have done better.
The opening has hardly ever been seen in Grandmaster chess since…
It’s true that if you can play a plan beyond the computer’s ‘horizon’, it will work. The problem these days is a) getting such a possible position b) coping with the incredible processing power of modern machines.
Fair enough.
Incidently, who was your guy?
Mine is Paul Littlewood (former British Champion).
Actually I believe it transposed into a Polish Defense.