A Test for Sandsted

[ClonedSIM]

Why do the odds chance of someone pass the pre-test...Art

I guess that was a question, although it didn't have a (?), and I'm not sure I understand exactly what you are asking.

If you're asking; Why have a Pre-test with Odds of 1:1000? I expect it is to weed out those who can actually demonstrate what it is they claim to be able to do, from those who merely submit a claim but can't come close to performing at the level of their claim. Thus, a lot of time and effort is saved, if there is no real reason to go through with the $1M challenge, and the tougher odds associated with it.

Jean

Another beautiful answer! I like the way you think, Jean!

 

[aarthrj3811]

Gee af1733....My mistake. I thought she said the math was interesting I just told her how intersting it was...Art

 

[Jean310]

I will not know the correct date after each trial. I will attempt to date each coin, right down my findings and send them back to Carl.

He'll tell me the dates.

In that case it looks like a problem involving 10 trials and each trial can have 40 possible outcomes. The outcomes are all part of the set {1960..1999}. (40 different dates)

I wonder what sort of accuracy you claim to be able to produce with your dowsing of the dates?

A test of this nature should probably have a definite level that was considered Success, as opposed to Failure. That way a conclusion could be reached based completely on objectivity.

Say for example that between you and Carl, it was decided that in order to demonstrate a real dowsing talent, you would need to get 6 of the 10 trials exactly right. This would equate to an accuracy level of 60%. If my math is correct, the Odds of getting 6 exactly correct simply by guessing (or chance) would be 1 in 839,663. Another way to say the same thing would be; if you performed these 10 trials a large number of times, you could expect to get 6 exactly correct, just by guessing, approximately 1 out of every 839,663 times you did the trials.

I think that's correct. Maybe Carl can jump in and correct me if I've missed something.

Also, if you don't have to hit the date exactly, but only get it within a certain +/- range, then that is a whole different problem and much different odds.

Jean

 

[ClonedSIM]

Gee af1733....My mistake. I thought she said the math was interesting I just told her how intersting it was...Art

She was just trying to answer your semi-question, Art.

Hey, Jean. You'll learn real quick that very little of what Art says has makes sense. Best just to ignore him unless you have a stiff drink or a handful of aspirin handy. Welcome to T-Net!

 

[Jean310]

Gee af1733....My mistake. I thought she said the math was interesting I just told her how intersting it was...Art

She was just trying to answer your semi-question, Art.

Hey, Jean. You'll learn real quick that very little of what Art says has makes sense. Best just to ignore him unless you have a stiff drink or a handful of aspirin handy. Welcome to T-Net!

Thanks. Guess I thought there was question. My mistake.

I've read other areas of T-Net, but just recently looked in here and got brave enough to inject a comment. I'll try to stay in the background.

Jean

 

[Real de Tayopa Tropical Tramp]

DUH,,I seem to be lost but how much more complicated is it to give a test of 1/1.000.000 to one of 1/ 1.000? It seems to me that a test of 1/1,000,000 would do far better job of weeding out those that may fail, as well as those that might succeed and just would be just as rapid..

Some how in the back of my mind lingers the question, what happens to the show if someone actually collected the money? At this point Randi's creditability and the show's value would vanish, There would be no need for him. since this would negate everything that he has been preaching. Some how I cannot see them allowing this to happen.

Tropical Tramp

 

[ClonedSIM]

Gee af1733....My mistake. I thought she said the math was interesting I just told her how intersting it was...Art

She was just trying to answer your semi-question, Art.

Hey, Jean. You'll learn real quick that very little of what Art says has makes sense. Best just to ignore him unless you have a stiff drink or a handful of aspirin handy. Welcome to T-Net!

Thanks. Guess I thought there was question. My mistake.

I've read other areas of T-Net, but just recently looked in here and got brave enough to inject a comment. I'll try to stay in the background.

Jean

Naaahhh, hang out. It can get fun at times.

 

[Jean310]

DUH,,I seem to be lost but how much more complicated is it to give a test of 1/1.000.000 to one of 1/ 1.000? It seems to me that a test of 1/1,000,000 would do far better job of weeding out those that may fail, as well as those that might succeed and just would be just as rapid..Tropical Tramp

I really don't know the exact complexity of designing one test or another. It may well be nearly the same. However, if you investigate the JREF site and look at past history of those attempting to capture the $1M prize, the weeding out process is quite important because all indications are the preliminary (easier odds) test is not performed by the JREF group. The preliminary test is delegated to a different agency. I think in most instances it is administered by the Tampa Bay Skeptics Society, or at least some of them are done by that group. In that regard, it would make sense (at least to me) to have a weeding out process.

Jean

 

[ClonedSIM]

It's a nice critical way to view the testing process. Some folks here believe the pre-screening is a way for the JREF to get rid of people that could potentially complete the challenge and win the million. I personally believe that it's more of a laughable pretense they use for not attempting the challenge, or even applying for it.

Believe me, if you choose to stick around, you'll find some awfully fantastic claims by some of the dowsers here. But I won't try to lean you one way or the other, just skim through the dowsing threads and see for yourself.

Have a Happy New Year!

 

[aarthrj3811]

Ya Right af1733...no-one as ever passed the pre-test...No-one...No-one has ever taken the test for the big bucks....Art

 

[Real de Tayopa Tropical Tramp]

[=af1733 ]
HI Sandsted,. If you failed to get a single one correct that would be a negative way of provng that you are 100%
correct in your ablity. Carl about my post #88
Tropical Tramp
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Ummmm, so you're saying that Sandy would prove he has no ability to date coins?
***********

Perhaps I am wong, but the same odds are involved in calling them all wrong as all right no?

TRamp

 

[Mike(Mont)]

I'm no expert in odds, but if you have ten coins and someone asks you do date them, that could be you have coins from the last two hundred years somewhere around one in two hundred for each coin, so two hundred to the tenth power is two hundred with ten zeros after it. That's 2,000,000,000,000 to one. That's two trillion to one. Okay, lets assume it's only for the last one hundred years now the odds are 1000000000000 to one. And to be generous for the last fifty years it's only 500 billion to one. Okay, coins only from the last ten years is only ten billion to one (ten to the tenth power or ten with nine zeros after it). BTW the Powerball lotto is about a hundred eighty million to one.

 

[Carl-NC]

He'll tell me the dates.

No, once I see that the coin holders were not tampered with, I'll mail them back to you and let you open them. That way you will know I didn't switch coins on you, or just lie about the dates. When you get the coins back, you should also check to ensure I didn't tamper with them (more on this later).

In that case it looks like a problem involving 10 trials and each trial can have 40 possible outcomes. The outcomes are all part of the set {1960..1999}. (40 different dates)

Ahhh... good approach.

I wonder what sort of accuracy you claim to be able to produce with your dowsing of the dates?

A test of this nature should probably have a definite level that was considered Success, as opposed to Failure. That way a conclusion could be reached based completely on objectivity.

I think he's tentatively suggesting at least 6 correct dowses to within +/- 3 years each. So if one of the coins is 1970 and he dowses 1973, we call that correct (enough)... 1974 is not. So, based on this, we do have a definite level to objectively compare against.

If my math is correct, the Odds of getting 6 exactly correct simply by guessing (or chance) would be 1 in 839,663.

I think that's correct. Maybe Carl can jump in and correct me if I've missed something.

I believe it's 1-in-14,545,440. But it's late, and I'm tired.

I'm no expert in odds, but if you have ten coins and someone asks you do date them, that could be you have coins from the last two hundred years somewhere around one in two hundred for each coin, so two hundred to the tenth power is two hundred with ten zeros after it. That's 2,000,000,000,000 to one. That's two trillion to one. Okay, lets assume it's only for the last one hundred years now the odds are 1000000000000 to one. And to be generous for the last fifty years it's only 500 billion to one. Okay, coins only from the last ten years is only ten billion to one (ten to the tenth power or ten with nine zeros after it). BTW the Powerball lotto is about a hundred eighty million to one.

Why, Mike, I do believe you have a basic grasp of probabilities! (Art, pay attention!) However, you didn't read the conditions... A 40 year span... 10 coins... +/- 3 year allowable window.

Others... Please keep this topic limited to the Sandsted Test. There is another thread for discussing Randi.

- Carl

 

[Jean310]

I believe it's 1-in-14,545,440. But it's late, and I'm tired.

Yes, you are probably right. I got in a hurry and took some calculations I did when at first you were discussing a range of only 23 different dates S{1960..1982}. Later today I'll run through the Numb3rs again , and post what I come up with for 6 exactly correct.

One thing I do know. Mike is not correct, but I think he does not have the correct bounds for the problem .

Jean

 

[Jean310]

How could one consider this to be Objective Experimentation....in view of the fact that both tests are being performed by Skeptical societies? ( I believe we have already ruled out the possibility of them being remotely scientific in methodology. )

Judy

Being skeptical does not automatically infer that one is close-minded and cannot objectively arrive at an accurate and meaningful conclusion. Neither should we conclude that all self-proclaimed skeptics are in fact unbiased. If, in fact, it is desirous for two skeptical factions to mutually involve themselves in a test, then it becomes the goal for both parties to agree on an approach (hopefully scientific) that both are satisfied with, and will yield a meaningful objective conclusion. Else, why bother?

Jean

 

[Real de Tayopa Tropical Tramp]

[=. I personally believe that it's more of a laughable pretense they use for not attempting the challenge, or even applying for it.

***********

Moning Af, when are you applying for a place on the Olympic Gymnastic team ?

Tropical Tramp

 

[Jean310]

I believe it's 1-in-14,545,440. But it's late, and I'm tired.
- Carl

Okay, I went over the numbers again. First, I better verify that I have the problem defined correctly.

A single test consists of 10 trials.
Each trial will have an outcome from the set S{1960..1999} (one of 40 possible dates**)

There would be Comb(10,6) = 210 ways to get 6 exactly correct.

Neglecting round-off errors, the Probability of guessing 6 exactly correct dates would be 4.63317x10-8

If the Odds are the reciprocal, then the Odds of Guessing 6 Exactly Correct should be 1 in 21,583,515

Haven't done these for a while. Where did I go wrong?

Jean

(**) I suppose there is another possible way to consider this. Suppose the subject doing the Trials is completely confident in their dowsing and would not consider the possibility of making a mistake. In this case, the second Trial could only have 39 possible outcomes, not 40. And, so on, each subsequent trial having one less possible outcome than the previous one. If this were the case, then the Prob and Odds would be different.

 

[Carl-NC]

If the Odds are the reciprocal, then the Odds of Guessing 6 Exactly Correct should be 1 in 21,583,515

Yup, for that test method, your odds are exactly right. I used the approach that each successive test has one less possible target. I still need to check my math, though.

- Carl

 

[Mike(Mont)]

My numbers were for guessing the exact year for ten coins. So, whatever the rules are different than that and Carl comes up with 14,545,440 we'll use that number.

Let's compare this to a map dowsing exercise. There are 27,878,400 square feet in a square mile. Now if you use a quadrangle topo map and mark a pencil point that is approximately a twenty foot square or 20 x 20 = 400 square feet. So 27,878,400 divided by 400 is 69,696 different possible pencil points for a square mile (most dowsers say not to expect any better than 100'-200'). Now since Carl says the odds are 14,545,440 you divide that by 69,696 and you come up with the equivalent of over 208 square miles of search area and you would need to be within the pencil point accuracy in order to pass his test. And that assumes that he doesn't do like the other skeptic contest and probably make you take it over again because then the number would be 208 times 208 or 43,264 square miles (nearly the entire New York state) and be within one pencil point of accuracy. Good luck! If you are that good you wouldn't be wasting your time on a piddley twenty-five thousand.

 

[Henry Hartley]

Mke(Montana)
Thanks for exposing another Flim Flam attempt by Carl. Funny how he didn't mention odds until you called him on it. Another great "tell" and a glimpse into his motives.
Henry Hartley