eight bits on Sep 27 2008, 05:34 PM, said:

Hello, Maybis

Good for you to try to get some evidence about all of this.

I would be happy to help you with the statistical part, except that I am unsure what you did. Depending upon what you did, there might not be much point in a statistical analysis, or even much real possibility of performing one.

Methods: I didn't understand

So, is this with feedback or not (does the caller know the history of the cards or not at each guess)?

If the caller knew the history, the probability of red or black (and presumably you did the same thing with A-6 and 8-K?) would not be a constant 50% on each trial.

Sample size: I also didn't understand

So, is that 125 red-black and 125 high-low trials with you inside, plus another 125 red-black and 125 high-low trials with you outside? For a grand total of 500 trials?

If only 250 trials overall, then how were they divided between inside and outside?

And which of you contributed the idea that using only half the deck would prevent "card counting"? Who supplied the cards? Who shuffled?

As you can see, there's a lot more to this than algebra.

See, that's the kind of feedback I'm looking for. I have pretty much a shoestring budget on this (currently looking for a new job, so not a lot of spare money at the moment,) so I have to use what's on hand, such as a deck of ordinary playing cards. Myself and two friends, and the psychic, (who was brought in by a friend,) met on a saturday evening, and spent a few minutes brainstorming. One of my friends was a skeptic, who believes all psychic phenomenae are just clever tricks, and the other friend is (and I mean no offense to anyone with this stereotype,) A crystal-collecting, new-age hippie-type who is a hardcore believer in all things paranormal. I do not recall exactly who came up with what idea, after maybe fifteen minutes, we had a plan everyone agreed on.

The entire deck would be shuffled, but only twenty-five cards would be revealed. This would, while not eliminating card counting, still reduce the accuracy. Turns alternated, First with me as the person who shuffled the deck and revealed cards, with my two friends present as witnesses. It took maybe five minutes to conduct the test. Then, I took a walk around the block, coming back after my friends conducted the test. There would be a few minutes between each round, where results were tallied (unfortunatley, just using basic algebra...no statisticians among us). So, figure ten or fifteen minutes for one round of testing ( two trials per round.) It was a little over an hour to complete the red/black test, then we took a break for dinner, removed the sevens, and did the high/low test.

As for the shuffling, I shuffled and dealt when I was present, and I believe it was the skeptic who shuffled and dealt during my absences. Since we alternated back and forth, I think that canceled out any fatigue over time, or at least balanced it between the two different 'control groups'.

So, yes, a total of 500 cards were flipped. (I did look around for my notes from that night, but I am in the middle of moving out of my apartment, so I can't find them to review.) We just had a simple 2 column tally system to record outcomes, so It isnt exactly a precise log of each individual card's results. We figured the percentage out as we went along, then at the end of the night we added up everything for the final answer (that is, after I flipped 25 cards, we had a number like 10 out of 25 right, then at the end of the nightwe had something like 51 out of 125 *not the exact numbers, mind you, but hopefully it gives you an idea of our system.)

It will be a few weeks before I might find time to a similar trial with cards (talking my friends into killing a saturday night again with the same thing will be tough.) But if you had suggestions on how to refine the test, or even a different type of test altogether, I will gladly try to get more refined results. I just started with playing cards since they were handy.