A Test for Sandsted

[Real de Tayopa Tropical Tramp]

=SWR Um...what is "shehs
"********
Obviously a very important part of a psychological trap swr. You will now have to figure it out yourself. As our briliant statisticians have said, "I am not here to teach you".

Tropical Tramp

 

Amazon Forum Fav 👍

Treasure of Victoria Peak - Grab it through Amazon!

[ClonedSIM]

Any results on Sandy's test yet?

 

[Real de Tayopa Tropical Tramp]

HI ROOM
isn't it about time to discuss Dowsing rationally and try to suggest ways to improve it instead of just mocking, ridiculing it and some of it's posters.etc ?

Even Carl, one of the most devoted attackers, has admitted to dowsing, so since there is no longer a question of it's existence, why don't we start on ways to improve it.

As to how it works , that is outside of any of our indvidual expertise, so put that aside, and discuss dowsing rationally

Tropical Tramp

 

[Oroblanco]

Greetings,

Dang what an interesting test! I have never heard of someone able to tell the dates of coins by dowsing before, not to cast aspersions but boy that seems like it would be really difficult with the 'tools' to be used.

I also don't wish to cast aspersions on the test parameters either but the odds may not be quite so 'long' as we are thinking here - for if I were to just guess away at ten possible dates, then compare them with ten randomly-picked coins, I would be having in effect, ten guesses for each of ten coins - unless the coins are sent one at a time, with the guess being ONLY for EACH individual coin, NOT just given a pack of ten, then dowse or guess ten possible dates. See what I mean? This is not too different from many state lotteries - you pick a bunch of numbers, then your picks are compared with the set of numbers drawn, and each number you 'guess' is compared to the whole group of numbers - the odds of picking at least SOME of them correctly is not anywhere near the 1:1,000,000 or even the 1:1000 odds. So if this test is going to be run, I suggest sending ONE coin, getting ONE dowse, then sending the next coin for the next dowse, until all ten are done without comparing the dates guessed to the coins until all TEN are finished.

May I ask a question here, how on earth can one dowse the dates of coins, in practice? Do you use a plumb-bob over a chart of dates, (just guessing at how this might be done) or some other way? If this works, it pretty much would have to be some kind of ESP or mental ability, it seems to me - and if it works, I would sure like to have a set of lottery numbers to play! An interesting test at any rate, I hope you fellows will decide to post the results here for us?

Oroblanco

 

[Jean310]

Greetings,

Dang what an interesting test! I have never heard of someone able to tell the dates of coins by dowsing before, not to cast aspersions but boy that seems like it would be really difficult with the 'tools' to be used.

I also don't wish to cast aspersions on the test parameters either but the odds may not be quite so 'long' as we are thinking here - for if I were to just guess away at ten possible dates, then compare them with ten randomly-picked coins, I would be having in effect, ten guesses for each of ten coins - unless the coins are sent one at a time, with the guess being ONLY for EACH individual coin, NOT just given a pack of ten, then dowse or guess ten possible dates. See what I mean? This is not too different from many state lotteries - you pick a bunch of numbers, then your picks are compared with the set of numbers drawn, and each number you 'guess' is compared to the whole group of numbers - the odds of picking at least SOME of them correctly is not anywhere near the 1:1,000,000 or even the 1:1000 odds. So if this test is going to be run, I suggest sending ONE coin, getting ONE dowse, then sending the next coin for the next dowse, until all ten are done without comparing the dates guessed to the coins until all TEN are finished.

May I ask a question here, how on earth can one dowse the dates of coins, in practice? Do you use a plumb-bob over a chart of dates, (just guessing at how this might be done) or some other way? If this works, it pretty much would have to be some kind of ESP or mental ability, it seems to me - and if it works, I would sure like to have a set of lottery numbers to play! An interesting test at any rate, I hope you fellows will decide to post the results here for us?

Oroblanco

It is my understanding the group of ten coins would be selected at random from a larger group of coins containing, I believe a total of 40 unique dates. The group of ten would be dowsed, but the dowser would NOT know the results of each individual dowse (guess) because the coin packets would not be opened. The ten coin packets would then be sent back to the sender (unopened). The sender would verify the packets were still in their sealed condition, and THEN be mailed back for the dowser to open and score his test. That's my understanding.... Carl could comment further.

In order for the dowser (Sandsted) to prove his dowsing was actually better than what someone might get from just guessing, he would have to correctly identify some number of the coins, which would reflect a real dowsing talent. I'm not sure this level of significance has been determined, but I believe the number 6 or more was discussed. However, it was also discussed that the (guess) would not have to be exactly "on the money", but that it could be within a range. Having a range to be in changes the odds quite a bit.

If the details of the test have been worked out, I don't think it has been published on the forum. Carl or Sandsted could comment further. Perhaps the idea has been abandoned.

Jean

 

[ClonedSIM]

A difference between Dowsing & Divination (predicting the future) is that Dowsing is only relative to past events, and the information obtained can, and has been used to assist the gambler to determine probabilities, but does not work to actually predict future events of who will wine a race, game, or lottery.

Is this why you always seem to be able to tell where the treasure was hidden, but can't tell if you'll actually recover it?

 

[xupz]

Is this why you always seem to be able to tell where the treasure was hidden, but can't tell if you'll actually recover it?

That's an asine question since I just explained adifference between, Dowsing & Divination, and you quoted me in your reply.

ANSWER: Because I do not do Divination. I cannot predict the future.

But, from mental Dowsing information, and my years of experience as a Professional Hunter/Salvor it helps me determine if there is the possibility of a recovery with the resources I have, without ever going to the Dowsed location.

I have Dowsed many hundreds of sites where a recovery would be impossible with my limited financial resources. I have followed up with a number of exploratory excavations in order to learn the feasibility of a recovery under most conditions. This field experience, and knowledge of Treasure Hunting allows me a distinct time and money saving advantage to cover a lot more area in a shorter period of time than with conventional searches for deep buried, and ancient treasures. Dell

Funny thing is if you were as good as you actually claimed to be, there's plenty of people in this world who would front you the money to claim these treasures. I'd imagine that route is completely pointless though because before they would front you anything it would require you PROVE to them you can actually do what you claim. The only problem with that is that dowsing doesn't work and you couldn't prove anything so you're not getting anyone to take the risk to finance any of your "dowsed" locations. Lovely cycle wouldn't you agree?

 

[Carl-NC]

I also don't wish to cast aspersions on the test parameters either but the odds may not be quite so 'long' as we are thinking here - for if I were to just guess away at ten possible dates, then compare them with ten randomly-picked coins, I would be having in effect, ten guesses for each of ten coins - unless the coins are sent one at a time, with the guess being ONLY for EACH individual coin, NOT just given a pack of ten, then dowse or guess ten possible dates.

Jean is correct. 10 coins have been drawn from a pool of 40. Each coin is sealed and marked with a number from 1 to 10. Sandsted will be dowsing the date of each individual coin, not 10 dates that any of the 10 coins might be.

- Carl

 

[Carl-NC]

Any results on Sandy's test yet?

Sorry this has taken so long. It took longer to collect all 40 dates than I thought. I have them, have randomly drawn the 10 coins, and am in the process of sealing them. They should mail out tomorrow or Wed.

- Carl

 

[Jean310]

Any results on Sandy's test yet?

Sorry this has taken so long. It took longer to collect all 40 dates than I thought. I have them, have randomly drawn the 10 coins, and am in the process of sealing them. They should mail out tomorrow or Wed.

- Carl

Good. Then the test is going forward. Is there a stipulation the results cannot be shared on the forum, or will we find out here? What was the date range that was agreed on?

Jean

 

[xupz]

What kind of coins?

 

[Carl-NC]

Good. Then the test is going forward. Is there a stipulation the results cannot be shared on the forum, or will we find out here? What was the date range that was agreed on?

I will post some info on how the coins are selected and sealed, and how the test proceeds. I will let Sandsted decide whether or not to share the results.

What kind of coins?

US Lincoln cents, 1960-1999, no mint marks. Yes, I know this includes zinc pennies, but the date range creates a fairly even split, so I'm not too concerned.

I still owe some statistical predictions.

- Carl

 

[Real de Tayopa Tropical Tramp]

Good morning room
Sheehs I leave for a few days to conclude a mining contract that may well give me a few million over the next few years and when I return I see that the mice have come out of their holes and are running amok, squealing and fighting over "look at me, see how brilliant I am".

However, I still see that not one of them has even attempted to address the data in the now infamous posts #6 & #88. So the obvious conclusion must be that they simply cannot, so they just run away from it by ignoring it or simply using the "ignore " button, which is exactly the same thing..

I will, go over the past few days posts and address those that are not too silly or plain dumb. <- (polite)

As for my statement in post #1120 in the 1,000,000 series , no-one seems to have understood what I was saying apparently from the recent posts.

Carl very clearly wrote "Of course I can dowse! It's quite easy, just not very useful. - carl". In any court of Law that I have ever attended or practiced that is a complete, legally binding, statement with no modifiers or ambiguities.

He flatly states that dowsing exists, so all that the incredibly flawed statistic tests, or any of the others can show is only what that "particular dowsers ABILITIES are", nothing more.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Operational definitions are inherently difficult — arguably, even impossible — to apply to mental entities, because these latter are generally understood to be accessible only to the individual who experiences them and are therefore not independently verifiable."
=================
Of course I can dowse! It's quite easy, just not very useful. - Carl
\\\\\\\\\\\\\\\\\\\\\\\\\\


Tropical Tramp

 

[Carl-NC]

Test update...

I have the coins selected, sealed, and packaged for delivery. To recap...

I collected 40 different dates of Lincoln cents from 1960 to 1999. I put all 40 cents in container and, without looking, randomly drew 3 coins. Those 3 coins became the practice pieces. The dates are 1976, 1970, and 1987 (that's the order I drew them) and are labeled "A", "B", and "C".

I replaced those 3 dates so that I once again had 40 coins. I then randomly drew 10 coins for the test pieces. The dates are... errr, obviously, that's a secret!

Each coin was placed in an envelope made of 2 pieces of tri-folded card stock. That is, the coin is first placed in a piece of card stock that is tri-folded around it, then another piece of card stock is tri-folded around that at 90 degrees, so the coin cannot fall out. A pic below shows this. The result is that there are 3 layers of card stock on both sides of the coin.

One side of the envelope is numbered (or lettered, for the practice pieces), and the other side has a fingerprint. (See other pics.) The envelope is then laminated in clear tape. The purpose of all this is to make it tamper-proof. Though I have no reason to believe that Sandsted would cheat, no test should rely on trust to be valid. I also do not expect Sandsted to trust me, and that will be dealt with as well.

All 13 coins will be mailed to Sandsted. The 3 lettered coins have known dates, and are packaged exactly like the 10 test pieces. This will allow him to try dowsing the dates of known coins, to see if the packaging has any effect. If he is satisfied with his results on known coins, he can move on to the test pieces.

In order to keep the overall test relaxed, there is no time-frame for completion. It is wide open. Once Sandsted has dowsed the dates, he is to mail the 10 sealed test pieces back to me. He is free to place a unique identifying mark on each piece or to take high-res photos of each fingerprint. He will also send me his dowsed dates. As soon as I receive his results, I will immediately mail the still-sealed* test pieces back to him, and he can open them himself and verify the dates.

Results of the test will NOT be posted by me, nor will I share them with anyone privately. Only Sandsted can reveal the results, if he chooses.

Therefore, the remainder of my activity in this thread will be to discuss the test protocol, the statistics of the expected results, and whatever Sandsted might want to discuss.

- Carl

*Tamper-proofing is a 2-way street.

 

[Oroblanco]

COOL test!

 

[Carl-NC]

Hi Carl.....this looks to be an interesting exercise. I hope you don't mind if I ask a couple of questions....just to clarify a few points for everyone.

Of course not. I'll even answer your questions.

Question #1
Has there been a date "range" agreed upon that will be used to determine his level of accuracy, and a pre-determined number of accurate "guesses/divinations" to qualify as a Significant Finding?

Sandsted has stated 6 or more correct to within +/- 3years. Therefore, for a coin dated 1970, anything from 1967-1973 would be accepted.

Question #2
You say that each coin is placed into a tri-folded piece of cardstock then placed into another tri-folded piece of card stock, illustrating this with an accompanying pic. From the photo it appears to me that the cardstock is actually folded 4 times....side...side...over...and under.

The photo is not clear. It shows 2 pieces of card stock but it's hard to see how they get folded. The first piece is tri-folded around the coin so there is one layer of card on one side (let's call it the back side) and 2 layers on the front side. But the coin can just slide out, so I tri-fold another piece of card stock at 90 degrees, with the 2 flaps to the back side. This gives 3 layers for each side, and no way for the coin to slide out.

And BTW....I actually like this test Carl.
( Even if it is anecdotal and doesn't actually prove anything....heehee. )

It would be anecdotal, if Sandsted simply showed up and said, "I successfully dowsed the dates on some coins." But this is a controlled randomized-blind test that is being publically reviewed, and proctored by someone who is independent of the claimant (and a skeptic, no less), so it ceases to be merely anecdotal.

As far as "proof," if the test is properly completed, it can provide evidence for (or against) the individual claim that Sandsted can dowse the dates on coins. It does not necessarily prove or disprove his ability, as he might have had a bad day, or he might have gotten Powerball-lucky. Also, it cannot prove that no one on Earth is able to dowse the dates on coins, or that no one on Earth can dowse at all, but I make no claim that it is intended to do that, because it is not.

So, does this test "prove" anything? Strictly speaking, no. Can it provide compelling evidence? Yes. That's the purpose of testing.

- Carl

 

[Real de Tayopa Tropical Tramp]

=Carl-NCI have the coins selected, sealed, and packaged for delivery. To recap...
I
*Tamper-proofing is a 2-way street.
***************
Put a series of marks such as our familiar xoxoxoxoxoxo covering both sides so that later, if hes successful, no-one can say that he used strong lights to peek.
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Operational definitions are inherently difficult — arguably, even impossible — to apply to mental entities, because these latter are generally understood to be accessible only to the individual who experiences them and are therefore not independently verifiable."
=================
Of course I can dowse! It's quite easy, just not very useful. - Carl
\\\\\\\\\\\\\\\\\\\\\\\\\\


Tropical Tramp

 

[Carl-NC]

So let's take a look at the statistics behind this test, starting with the questions I posted earlier:

There are 10 coins, randomly drawn from a batch of 40 coins, each with a different date from 1960 to 1999. The date on each coin is randomly guessed, knowing that it can be in the given date range and there are no duplicate dates.

What are the odds that 6 (or more) guesses will fall within +/- 3 years (or less) of the actual dates?

Warning: this is a very difficult problem, so here are a couple of simpler problems...

What are the odds that 6 (or more) guesses will be EXACT?

What are the odds that 1 guess will be within +/- 3 years?

Add'l info: The actual dates are not revealed until all guessing is done. It is assumed that all guesses will be unique.

Let's start with the easy problems first. In fact, let's start even easier:

Q1: If only 1 guess is made, what are the odds that it will be correct?

A1: That's pretty easy... 1-in-40, or 2.5%.

Q2: Next, in 10 guesses, what are the odds that only 1 guess will be exact?

A2: Let's say the first coin is guessed correctly, and the other 9 incorrectly. The odds of this happening are 1/40 * (38/39)^9 = 1.98%. But you have 10 chances to get one right, so overall it's 19.8%.

Q3: In 10 guesses, what are the odds that only 2 guesses will be exact?

A3: You would think that guessing any 2 particular coins coins correctly has odds of 1/40 * 1/39 * (37/38)^8 = 0.05%, and with 45 combinations to get 2 right, an overall 2.33%. But this is not quite right.

With this kind of test, there are multiple quirks to deal with because the results are only known at the very end. Let's say, ferinstance, that on the first selection 1970 is guessed incorrectly, but that 1970 is an actual date that is selected later on. Since 1970 was guessed earlier, it is now impossible to get this one correct.

This skews the odds slightly. However, the effect is not large, especially when the pool (40) is much larger than the number of selections (10). And calculations of the exact odds are very difficult*. So I will leave the exact calculations as an exercise to the reader, and use the approximated odds.

Q4: In 10 guesses, what are the odds that only 6 guesses will be exact?

A4: 1/40 * 1/39 * 1/38 * 1/37 * 1/36 * 1/35 * (33/34)^4 = 0.000000032%, and with 210 combinations to get 6 right, an overall 0.0000067%.

One of the earlier questions was the odds that 6 (or more) guesses will be EXACT. For this, take A4 and add to it the odds for 7 correct guesses, 8 guesses, 9, and 10... it turns out they are exponentially smaller, so A4 is approximately the right answer.

Q5: What are the odds that 1 guess will be within +/- 3 years?

A5: The approximate answer is 7/40 * (32/39)^9 * 10 = 29.5%. Again, this is not exact, because the odds of a single success is not always 7/40. Ferinstance, 1970 has a correct range of 1967-1973 which is 7 years, but 1960 has a correct range of 1960-1963, which is only 4 years. But I'll avoid the excruciating calculations and go with the approximations, which are plenty close enough.

More later....

- Carl

*Very very difficult. To the point where I would need to write a computer program, which I'm not going to do.

 

[Jean310]

The actual Probabilities for this scenario makes for a very interesting problem, and yes it does have some quirks.

For instance, here is one I thought of that may or may not be significant:

Suppose Sandsted dowses one of the "early" packets and comes up with the date of 1972. Now does he consciously decide that 1972 is not a possible answer for any of the following dowsing attempts? If he does, then the field has been reduced by one for the next dowse. Or, might he dowse one of the later packets and get the distinct impression that IT is 1972, and in which case he would go back and re-dowse the earlier pick for 1972, and then actually reassign a new date to the earlier packet?

As you say, these quirks will probably not skew the gross Pb values by that much --but they are interesting to consider, and I am putting them through a little number crunching exercise.

But here is another interesting thing. I setup this very test, and executed it 5 times. Each test was the 10 trials and dowsing was not involved, just random picks. Also, the guesses did not have to be exact, but had a range of + or - 3.

The results were: 0 correct, 1 correct, 2 correct, 0 correct and 3 correct. (in that order) Remember, those results did not involve dowsing. Honestly, I was a little surprised to see 3 correct, just from random guessing.

So.... to show a significance where dowsing might have merit over random guessing, I would think probably 6 or more correct would make some compelling evidence in favor of a real dowsing talent.

 

[Carl-NC]

In my statistical analyses, I assumed that no date would be chosen twice. Yes, Sandsted is free to dowse and re-dowse as often as he'd like. And he is free to chose the same date twice. In fact, if his dowsing skills are real, it might be to his advantage to do so. Let's suppose that 1972 is dowsed for a selection. Regardless of right-or-wrong, let's say that 1972 is dowsed for a second time, and that the actual date is 1969. If he sticks with 1972, it's a "hit". If he bumps it to 1973 because 1972 has been chosen already, it becomes a "miss". (Ad hoc claims that he initially dowsed 1972 will be ignored.)

Also, you might be surprised at how likely 3 hits are. It's around 16%.

- Carl