I should add that it is possible that faith healers etc know they are working entirely off the placebo effect, but are just very careful never, never to admit it because then the placebo effect would collapse.
If that is what you think, then presumably (as a New Ager) you despise them all for their lies.
I must say that that is not my experience of New Ager attitudes towards faith healers.
Exactly. On a slightly different topic; what the hell is “new” about New Age beliefs? Belief in ghosts, healing crystals, and general magic has been around forever. At least the people living in ancient times had an excuse. For modern, theoretically educated people to believe in this crap is inexcusable.
Regards
testy
I don’t have a dog 
Why do you refer to them as psi events if you don’t know that they weren’t a lucky guess? As you admit they could be?
If they regularly perform above chance, then they could pass a test. I don’t think you understand probabilities as well as you think you do. If I can guess head or tails 51 out of a hundred times (ie above chance but only just above chance) then all Randi has to do is require you to do that many many times. If you can do so, you will prove (at greater than million to one odds) that you can pick heads or tails at very slightly better than chance.
You seem to be confusing how much better than chance a psychic can do, with how many repetitions of that feat it would be necessary to do to prove you could do that.
Yer right, but what I said is also correct: very high odds or Randi loses money.
Absolutely. Contrariwise, any skeptic will glibly say, “Oh surely it was just the placebo effect,” even if there is evidence beyond that.
Well, he has displayed his abilities in the lab.
The skeptics say the experiments sucked; others say they were fine or partially fine. It’s a whole other argument.
Just to nitpick: I think you are referring to Ray Hyman, of the University of Oregon at Eugene. He told this story on a NOVA video and I have heard it from him in person.
To those asking for a list of challenges and detailed records of each, that may be forthcoming. If has been suggested on the JREF Forums, and in a personal communication with Mr. James Randi a few months ago, he said his staff is working on just such a venture to be published as a book. I get the impression it may be more in a readable, narrative vein than a heavily-footnoted scientific treatise, though. I agree such a document in either form would be welcome, although I doubt it will silence all the critics no matter how well-written it is.
As a slight hijack and niggling nitpick, while Randi and his followers are justifiably proud that no one has ever beaten their challenge I will point out that no one has ever beaten CSICON’s challenge either. The Committee for Surrealist Investigation of Claims of the Normal (CSICON) offers $100,000 to any “normalist” who can produce “a perfectly normal person, place or thing – or even an ordinary sunset. Or an average day.” The gauntlet has been thrown.
It was Ray Hyman and palm reading.
Anything could be a guess. There is no mathematical model for this stuff. If a person, just as a hypothetical, could read a note that you had written and shown to nobody else while it was still in your pocket, it would seem to be psi, wouldn’t it? Or fraud, if you had a theory about how he could have seen it.
I understand the issues. Whether it’s guessing heads or tails 100% in a row or some other method, Randi is going to require a certain high probability of failure in order to accept the terms. Would Randi accept my terms of guessing 10 heads/tails in a row? He’d be a fool to do so. More likely he would require me to guess 20 or more, as that would give him above 1/million odds, so his mathematically expected loss would be less than a dollar.
The same principle applies for guessing heads/tails with less accuracy. In theory he would let someone do so with more attempts, as you say, but the odds would similarly have to be in his favor. If he gives someone 1,000 trials, then he’s not going to be satisfied with 510 hits, as the person would have a 27.4% chance of succeeding.
It’s the total probability that matters.
Not necessarily: the applicants claims determine the timescale. If the applicant claims he can guess 10 coins correctly for 50% of any 10-flip sequence (which takes, what, 20 seconds?) that will likely be falsified quickly. If he claims he can do it for 0.2% of those sequences (ie. slightly better than chance) he is going to have to submit to many thousands, perhaps millions, of flips.
This is precisely the way statistical evidence works for any other discipline, from medical trials to particle physics. This is not “high odds”, this is simply statistical rigour.
Depends on what your claim is.
Claim#1: “Next Wednesday, I’ll be able to guess 10 coin tosses in a row correctly! But I’ll only be able to do it once, and never again. This is the sum claim I advance regarding my psychic ability.”
Randi will reject your claim as not being a claim of anything paranormal. Guessing 10 coin tosses in a row correctly, on one occasion, doesn’t prove anything paranormal whatsoever: it could easily be a lucky guess.
Claim #2: “I’m so good at coin tosses that I can guess 10 correctly in a row whenever I want!”
Randi will accept this claim for testing, and will set up an experiment where you have to prove this. The experiment might, for example, consist of your guessing 10 coin tosses correctly in a row on ten different occasions, these occasions to be chosen by you, under conditions chosen in part by you.
Would you consider his actions in either of these cases to be unfair?
Daniel
And I should add that this statistical rigour could be applied even to mediumship and the like: Surely a genuine medium could pick out the person with a given description (eg. an entire page of “mother called Betty died of cancer”-type details) better than chance?
Here you’re just going to have to admit you’re wrong. The total odds for any such event–I don’t care about the timescale or precise claim–can be calculated. And Randi isn’t going to accept anything but high odds.
Let’s look at a real-life example: the Ganzfeld experiments. People guess a symbol with 1/4 odds of getting it. The claim made is that the experimenter will average 33% whereas chance would be 25%.
You can calculate the odds quite concretely for any number of trials (although the calculations aren’t that easy without good software. I’m not going to calculate them now, although I did for the coin example above. The point I’m making should still be clear.).
Averaging 33% in 10 trials is still above chance. Probably not significantly significant and and bad deal for Randi.
Averaging 33% in 100 trials. Probably significant at this point, but I bet the odds are lower than 1/1000. Still a bad deal for Randi.
Averaging 33% in 1000. Now the odds of failure are going to be high. Definitely significant, and probably a good deal for Randi.
Point taken?
Point remains that such tests are difficult to devise. It’s not that people don’t devise experiments for remote viewing and mediumship, they do. But it is very easy to say “bullshit” to the way the probabilities are figured, etc.
Point incorrect.
“Bad deal for Randi”? What, do you think that if nobody wins the challenge in the next two years, he’s going to buy a yacht with that million bucks and live the high life?
It’s not his money.
He’s never going to get the benefit of a single dime of that money (assuming, of course, that he doesn’t develop psychic powers sufficient to win a challenge). So there’s no such thing as a “bad deal for Randi.”
You could say that someone’s winning the challenge would be an embarrassment for him, and that’s true. But the level of embarrassment has got nothing to do with the amount of money set aside for the challenge.
And you could say that he demands too high a success rate for folks entering the challenge. I’d just say that’s nonsense: he’s demanding the same deviation from the norm that any scientist demands when looking for significant results. The deviation for which he looks has got no relation whatsoever to the amount of money the challenger stands to win.
Once more: it’s not his money. That’s simply a canard.
Daniel
I don’t need to read the original thread or finish this one to agree with you on this. It sounds like, on that point, you’re doing exactly that.
Ah, I see what you mean: We’re both actually saying the same thing - A large sample space is necessary to make the probability of statistically significant results actually being down to blind luck incredibly low. Point taken.
Do we agree on the basis for a fair test, given that it will be subject to the statistical rigour we have discussed? Surely, given a page of explicit detail, somebody with genuine psychic powers should be able to consistently guess which person it described out of several?
Yes, it should be possible.
Do you trust that the Randi Foundation would execute such a test fairly and honestly?
If so, the question is why no famous applicants?