There is an interesting trick that was common for a time and still gets referenced by magicians where even the worst chess player can guarantee that they will win at least half of n chess games where n is any even number. They will also be able to appear to need very little time when they get to each table.
The trick of course is just to play each player against another. The magician acts as a proxy and makes the moves of one player against another. Of course it takes a little practice at memorisation but magicians are often practised at mnemonics anyway.
It's hilariously improbable to chess players, but some of my non-playing friends thought it was clever.
I actually don't think this is true. As I understand it, a given routine will have a number of high-risk tricks that may work, say, only 1-in-3 times. If they do work, they're highly effective, precisely because they're so implausible, and if they don't work you can just pretend it's a bluff and move onto plan B. It's all about effective showmanship and managing expectations.
(Disclaimer: not a magician!)
One of my friends had a card trick that involved me selecting a card. Before starting the trick, he jokingly picked a card at random from the deck, and asked me if that was my card. He had 1/52 chances of being right, but it would have been an insane trick if it had worked.
He guessed wrong (naturally), shrugged and went on with the actual trick as though the first attempt was just a joke, to be forgotten because an actual trick was about to take place.
And we don't want symmetry on one board, but between two boards.
Try it yourself vs a few instances of your favourite chess software or just here http://www.chess.com/play/computer (requires Java). Open a few instances and see how can you chuck away a move.
Playing a game blindfolded seems like incredible feat if you are not serious/competitive chess player. Playing several sounds like magic. The thing is though that at certain level almost anybody can play decent blindfolded game, especially vs much weaker opponent. Playing several takes some practice but it's mainly connected with not being confused (anchoring positions to something so it's easy to switch). It's really not that much more taxing on the brain as chess players have tons of positions memorized anyway. I am talking about 2-3-4 games. 10 is already very impressive to me and things like 20+ is insane even for competitive chess player standards.
I don't think about individual pieces placed in a board but rather interactions and moves. Pieces are not located at random, you can remember where pieces are because you remember the plans, the threats and the moves that lead to a setup. If I had to remember seemingly random arrangements of pieces it would be a completely different story and much harder.
Might help that I also played Shogi rather intensely and after being used to Shogi, the chess board feels very small.
When I'm alert, jacked up on caffeine, I can keep track of 100s of classes, 1000s of methods and member variables. When, I'm in the flow, I basically write the program in my head and let the fingers be my mouth.
There's a fair amount of new code... I'd say 10000 - 15000 lines, some of which I generated.
Maybe she wasn't ranked at the time, but a quick search revealed she played chess at an international level in her youth.
If a few players use the same openings and get themselves into similar positions it's very hard for the opposing player to keep track of the small differences that accrue.
Anyway, this reminds me of an anecdote I read long ago in a Tim Krabbe piece:
"A grandmaster once played a 10-board blind simul somewhere. Knowing the ropes of blind simuls, he varied his games right from the start, maybe opening two with 1.e4, two with 1.d4, one with 1.b3, and so on. To his suprise, all of his opponents played 1...b6. On the second move, five of them played 2...Bb7, and the other five 2...Ba6. On the third, three of the five players who had played 2...Bb7 now played 3...Bc8 and the other two played 3...Ba6, while three of the five who had played 2...Ba6 now played 3...Bb7, and the other two 3...Bc8. On move 4, the grandmaster saw bishops everywhere. After move 5, he excused himself and went to the toilet where he was happy to find a window big enough to let him through."
I agree it probably works better in blindfold play, but it works well in normal simul too (I used it successfully once, though that wasn't against a GM, and I had fairly recently stopped playing competitive chess).
Now going way off topic, but Krabbe's cycling book "The Rider"[1] is probably the best book about cycling I've ever read. Along with "Once a Runner"[2] I think it probably the best piece of sporting fiction ever:
My whole life had only one goal: making that last wheel, here, now. I was wasted. But that elusive finish line, eight, seven, six and a half meters in front of me, kept my hope and desire awake. I coughed and slobbered. I remembered the words of advice ‘Shift, when you’re really, truly at the end of your rope, to a higher gear.’ I shifted. A few hysterical kicks on the thirteen, then clenched power of a mortal struggle. I was there. I was sitting on that last wheel. I was in the lead group.
[1] http://cyclingtips.com.au/2010/11/the-rider/
[2] http://www.slate.com/articles/arts/culturebox/2008/12/speed_...
Similar moves with small differences to encourage mistakes? Very different to make the mental switching harder? Very esoteric openings to avoid "played this a 1000 times" situations for the grandmaster?
The GM's positional ability is almost certainly better than yours.