RFC: Rating Conversations With the Dead

[Beleth]

Please comment. It's dry in parts, and the formatting's not to my liking, so for those things I apologize.

An Objective Measurement of the Truthfulness of Statements Made During Conversations With The Dead

Abstract
There are few areas of paranormal abilities today that are as popular, or as controversial, as speaking to the dead. Many believers listen to a reading and only hear the “hits,” or true statements, the reader makes about the deceased. At the same time, skeptics tend to dismiss the hits as statistically expected and assign no value whatsoever to psychic readings. Such subjective analysis by both sides leads to endless, fruitless, debates on the subject. This article attempts to provide a way to objectively measure the effectiveness of a psychic reading by assigning weighted numerical values to statements the reader makes about the deceased.


Definitions
There are three groups of people involved in a psychic reading:
- The person acting as a conduit between the world of the dead and the world of the living, also known as the reader.
- The person or persons who are alive and listening to the reader, known as the petitioner pool.
- The group of dead people who could potentially use the reader to contact someone in the petitioner pool, known as the contact pool.

A statement is an unambiguous declaration made by the reader about someone in the contact pool. Statements are never questions, and they do not contain ambiguous words like “maybe” or “could”.

Assumptions
All of the numerical values here are subject to change as more precise data becomes known.
- One out of every 23 people who has ever been alive is alive now.
- There are approximately 6 billion people alive now.
- Each person will know approximately 8,000 people reasonably well, either because they have personally met them or because they know of them through other means.
- As one grows older, the number of people one knows who have died increases.


Determining The Truth Value Of A Statement
Statements have two qualities:
- Their veracity, or whether they are true statements or false statements.
- Their precision, or the number of people they could be true for.
For instance, the statement “The person who is contacting me is male” has a 50-50 chance of being true, but is not very precise since half of all the people who have ever died are male. The statement “He wrote a 3-page letter to his dead sister and placed it in her coffin at her funeral” has a low probability of veracity, but is very precise.

The veracity of a statement can be determined in many ways. The easiest but potentially least accurate is to verify the statement with the petitioner pool. A better way is to independently verify the statement through other methods – checking the dead sister’s coffin, for instance.

The precision of a statement is harder to determine. This article proposes using a scale based on the ratio of the size of the contact pool to the number of people it can be reasonably believed the statement would be true for. The first thing we have to do is determine the size of the contact pool, and to do that, we have to determine the size of the petitioner pool.

The Petitioner Pool
It is important to realize that the petitioner pool may not be just one person. The petitioner pool consists of everyone who is listening to the reader at the time the reader is making statements. This could be as small as one person, in the case of a personal reading, or as large as many thousands of people. In the case of media productions, such as television shows, it includes the entire audience, the cameramen, the people in the control room, and all the other people in the television studio at the time the reading is taking place. This is usually an unambiguous number of people.

The Contact Pool
The size of the contact pool, however, is much more open to interpretation, so this paper will only address the maximum, minimum, and reasonable values.

The maximum contact pool (CPmax) is the total number of people who have ever lived and who are currently dead. Given the assumptions above, there have been 23 x 6 = 138 billion people who have ever lived. Accounting for the 6 billion people who are still alive, 132 billion people are currently dead.

The minimum contact pool (CPmin) is the total number of people the petitioner pool knows who are dead now, and who would possibly be contacting someone in the petitioner pool. This could be as small as zero, for a newborn baby, or as large as all the dead friends and relatives of a popular but very elderly person.

The reasonable contact pool (CP) is the total number of dead people who would be reasonably trying to contact someone in the petitioner pool. It is reasonable to assume that Grog the Caveman would not try to contact anyone living today. It is also reasonable to assume that a famous figure from the distant past – Cleopatra, or George Washington, for instance – would also not reasonably be contacting anyone living today. Only relatives or good friends would reasonably try to contact someone in the petitioner pool.

That’s where the 8,000 estimate in the Assumptions section comes in. A poor but simple approximation is that 1% of everyone you know dies each year. That comes to 80 people a year. This is way too high early on in life but gets more realistic the older one gets. It might not be too unrealistic early on in life if one considers that recently deceased relatives may also be included in the contact pool.

So an adequate estimate for the contact pool is (80 x the person’s age) for every person in the petitioner pool. If there’s only one person in the petitioner pool, and he is 50, the contact pool is 80 x 50 = 4000. If there are 100 people in the petitioner pool, and they vary in age from 14 to 80, the contact pool could reasonably be around
((14 + 80) / 2) x 80 x 100 = 432,000.

The Contact Pool Constant
The contact pool constant, or P, is the logarithm (base 2) of the contact pool. For a contact pool of 8,000, P is approximately 13 (since log2 (8,000) = 12.965…). For a contact pool of 432,000, P is approximately 18.7.

The Precision Value
The precision value VP of a statement is a number that is based on how many people the statement would be reasonably true for. It’s the logarithm (base 2 again) of ratio of the number of people in the contact pool to the number of people in the contact pool the statement could reasonably be true for.

That sounds more complicated than it is. Say the statement is “The person contacting me is male.” Half the people in the contact pool are male, so the precision value of this statement is log2 (1/.5) = 1. If a statement is true for everyone in the contact pool, the precision value is log2 (1/1) = 0

If the statement is “The person contacting me put a 3-page letter in his sister’s coffin at her funeral,” the determination of how many people in the contact pool actually did that will be more open to conjecture. How many people do things like that? Not many. Perhaps it is one in a million. So the precision value of this statement is log2 (1/.000001), or approximately 20.

The Truth Value
The truth value V of a statement is computed by determining whether the statement is true or not and applying the following formula:
- If the statement is true, V = VP with a maximum of P.
- If the statement is false, V = VP – P with a maximum of –1.


For instance, if the statement is “The person contacting me is male” and the contact pool is 8,000, VP = 1 and P = 13. If the statement is true, the truth value is 1. If the statement is false, the truth value is –12. The same statement made with a contact pool of 432,000 (P = 18.7) would still have a V of 1 if it were true, but would have a V of –17.7 if it were false.

On the other hand, if the statement is “The person contacting me put a 3-page letter in his sister’s coffin at her funeral,” and the contact pool is 432,000, VP = 20 and P = 18.7. If the statement is true, the truth value is 18.7 (because it can never be higher than P, even if VP > P), and if it is false, the truth value is –1 (because it can never be higher than –1).

This system rewards precision and penalizes guessing. A reader who is merely guessing and just guesses easy things (gender, age, cause of death being “heart related”, etc.) is bound to get one wrong eventually and end up with an overall negative score. A person who can truly talk to the dead should have no problem getting some very precise details correct and score in the double-digits.

Any overall score of 20 or more would be a one-in-a-million event, and would certainly warrant submission to Randi’s Million Dollar Challenge.

 

[Ladewig]

I guess I am a bit thick. I had to go all the way down to The Truth Value before I realized it was a joke.

 

[apoger]

1- Too much of this is subjective
2- Without a controlled enviornment this is all worthless
3- This is not needed to determine the value of readers for potential in Randi's test. All they need to do is sign up and be tested.

 

[Ersby]

I may be thicker than a bus sandwich, but there’s things here I don’t get at all. It may be, as Ladewig said, a joke but in case not, I have trouble with:

The idea that everyone, more or less, knows about 8,000 people. This is crazy, and even if true wouldn’t make a difference to After Death Communication. The people referenced in a reading would need to be from family or close friends, alive or dead. For me that numbers around 40 people. If the reading included everyone I’d ever met or heard of, it’d be pretty pointless.

Then there’s the Contact Pool Constant. Is there any reason why they took the log of their (over-estimated) Contact Pool? They don’t give a reason, and I honestly can’t think of one.

 

[Beleth]

Ladewig:
Nope, not a joke.


Apoger:
The only parts that are currently subjective are the contact pool and the precision value of typical statements, and that's only because we don't know what the correct values are yet. Obviously the VP of the statement "The person who is contacting me is male" is an objective, not subjective, 1. We can objectively determine things like how many people's names begin with an M, or how many people own pieces of jewelry that match specific parameters, or how many people put letters in their sister's coffins... or at least gather enough data on it so that the VP does not remain a subjective SWAG.

Yes, of course a controlled environment is necessary. It's necessary for any scientific endeavor. I'm not sure how that is a criticism of this method specifically.

Once the reader signs up to get tested for the Challenge, an objective test needs to be administered. The scoring for such a test is what this RFC is about.


Ersby:
The numbers are, as I said, subject to change. It might very well be 40 people in any particular person's contact pool (making the Contact Pool Constant around 5.3).

Logarithms of large numbers are easier to work with than the large numbers themselves. For instance, the Contact Pool Constant for CPmax is very close to 37, which is a lot easier number to work with then CPmax itself, which is 132,000,000,000.

 

[Ersby]

I'm still not sure about this whole logarithm thing. Small numbers are easier than big numbers (although I posit that 40 is easier to work with than 5.3), but why take the log?

And if I'm reading it right, CPmax would only come into play if a medium was doing a psychic reading whereby everyone in the world is a potential sitter. That's not too likely.

 

[Beleth]

It's the other way around - CPmax only comes into play when everyone who has ever died is a potential contact.

It's just the absolute maximum. I only included it to set boundaries. I'm not expecting any reasonable person to ever use it.


Logs don't just make the numbers smaller. They also make the penalty for guessing, and the reward for getting precise statements correct, more pronounced.

Let's take a typical John Edward reading as an example.
50 people in the audience.
Another 14 people on the crew.
So petitioner pool is 64.

If the contact pool for each person is 40 then the contact pool for this reading is 64 x 40 = 2,560. Even this is overkill, as many petitioners go with other members of their family and their individual contact pools will overlap because of that. But it will suffice for this example.

If you don't take the log of things, and use the raw (contact pool) / (people in contact pool for which the statement is true) number, then a 50-50 statement like "The person I am contacting is male" would have a V of +1,280 if true and
-1,280 if false. You could guess and guess and guess 50-50 statements all day and still come out with a total of about 0.

Taking the logs, however, adds more information to this number. A 50-50 guess in this case has a V of 1 if true and a V of ((log2(2560/1280)) - log2(2560) = 1 - 11.3 = -10.3. Asking two 50-50 questions, one true and one false, will yield a net truth value of -9.3 when you use logs, where it yields a net truth value of 0 if you don't.

 

[Ladewig]

If there are 100 people in the petitioner pool, and they vary in age from 14 to 80, the contact pool could reasonably be around ((14 + 80) / 2) x 80 x 100 = 432,000.

I need clarification on this point. Are you saying that the reader doesn't have to identify the sitter when making these statements? Each of the (in this example) 100 people witnessing the reading can evaluate the statements?

Truth Values.
For instance, if the statement is “The person contacting me is male” and the contact pool is 8,000, VP = 1 and P = 13. If the statement is true, the truth value is 1. If the statement is false, the truth value is –12. The same statement made with a contact pool of 432,000 (P = 18.7) would still have a V of 1 if it were true, but would have a V of –17.7 if it were false.

I'm not sure how one determines whether this type of statement is true or not. Statements such as, "when you were sixteen you made a homemade rocket and set the garage on fire," can clearly be determined to be true or not by any sitter. Statements such as, "the person contacting me is wearing hat," have meaningless truth values because they can never be determined to be true or not.

The biggest problem with reading a large pool in this manner is it takes only one accomplice, one overeager sitter willing to exagerate/lie to prove the worth of ADC, or one easily convinceable person (planted memories are easier to plant among people who want to believe) to render the experiment useless. For instance:
Reader - you put a three-page letter begging for forgiveness and two photos of an albino squirrel that ate from your hand in your sister's coffin when she was buried.
Accomplice - yes, I did.
Statistician - wow, that's a one-in-one hundred-million shot; we now have proof.
Skeptic - well, to be completely accurate, we should dig up the coffin and look.
Audience - Boooo. Get that ghoul out of here.

A much simpler approach has been put forward many times, yet has never been accepted. Divide a small room in half with a screen. Let the reader use ADC to determine if anyone at all is on the other side of the screen. That's it. No subjective precision values; just a simple binary result. Repeat the process an agreed-upon number of times (perhaps 20), use a random method (coin-flipping) for determining whether of not to place a person behind the screen in each trial, and calculate the p-value of the results.

 

[Beleth]

I need clarification on this point. Are you saying that the reader doesn't have to identify the sitter when making these statements? Each of the (in this example) 100 people witnessing the reading can evaluate the statements?

Exactly. Think about how John Edward does it. He's got an audience of maybe 50 people, and his first few statements are just cast out before the whole audience. His audience is the petitioner pool, and it stays that size throughout the entire reading, even if he zeroes in on one audience member after getting hits on the first few questions.

I am reminded of one episode of John Edward where a private reading he was doing for a couple actually ended up being a reading for one of the cameramen. Seriously. The cameraman came out from behind the camera and came up on stage to be read, instead of the people that were supposed to be being read. If he wants to consider that a "hit", fine; it's only fair to include such people in the pool when there's a miss, too.

I'm not sure how one determines whether this type of statement is true or not. Statements such as, "when you were sixteen you made a homemade rocket and set the garage on fire," can clearly be determined to be true or not by any sitter. Statements such as, "the person contacting me is wearing hat," have meaningless truth values because they can never be determined to be true or not.

Well, yes. Questions about the contact's current state like that are pretty undecidable and are best left unanalyzed. Sounds like I should tighten up the definition of "statement". Thank you for pointing that out.

But a statement like "the person contacting me often wore a hat when he was alive" is something we could assign a precision value to.

The biggest problem with reading a large pool in this manner is it takes only one accomplice, one overeager sitter willing to exagerate/lie to prove the worth of ADC, or one easily convinceable person (planted memories are easier to plant among people who want to believe) to render the experiment useless. For instance:
Reader - you put a three-page letter begging for forgiveness and two photos of an albino squirrel that ate from your hand in your sister's coffin when she was buried.
Accomplice - yes, I did.
Statistician - wow, that's a one-in-one hundred-million shot; we now have proof.
Skeptic - well, to be completely accurate, we should dig up the coffin and look.
Audience - Boooo. Get that ghoul out of here.

In practice, though, the analysis and assigning of truth values would typically not happen in real-time, as the reading was taking place. And the reader in any test like this should not be getting real-time feedback from anyone in the petitioner pool like that.

A much simpler approach has been put forward many times, yet has never been accepted. Divide a small room in half with a screen. Let the reader use ADC to determine if anyone at all is on the other side of the screen. That's it. No subjective precision values; just a simple binary result. Repeat the process an agreed-upon number of times (perhaps 20), use a random method (coin-flipping) for determining whether of not to place a person behind the screen in each trial, and calculate the p-value of the results.

Maybe contacts don't "work like that". That does not sound like any ADC demonstration I have ever seen. There's no petitioner pool, for example - no one asking to be contacted.


I have been trying to think about how you could test this on live people - have the contact pool be people who have not yet died. You could take the petitioner pool's address book(s), pick someone at random, and phone them. Then you could ask the contact to tell you various facts about someone in the petitioner pool. Think how quickly the total truth value would skyrocket. Simple things like "The contact currently lives at 123 Main Street, Springfield, Arkansas, and is the oldest person living there" uniquely identifies a person out of all the people who have ever lived. Think about how quickly you could rack up unique statements about a living person. If it were possible to contact the dead, unique statements like that should be just as easy to rack up and get correct.


I want to end this by thanking you, Ladewig, for taking this seriously. I really do mean for it to be taken seriously. It's an off-the-wall subject but if there's no scale to measure statements on, then believers and skeptics are just going to keep talking past each other.

 

[Ladewig]

But a statement like "the person contacting me often wore a hat when he was alive" is something we could assign a precision value to.

I agree. We could assign a precision value to that statement. I am asserting that it is impossible to assign a truth value to that statement, because the sitter may never recognize who is contacting him or her.

 

[Ladewig]

Maybe contacts don't "work like that". That does not sound like any ADC demonstration I have ever seen. There's no petitioner pool, for example - no one asking to be contacted.

Perhaps the reason you have never seen any ADC demonsteration like that is that there really is no ADC and that type of demonstration would prove it.

All the people volunteering to be sitters could have dead parents. You could then tell the ADCer that if there is someone on the other side of the screen, that person wants to contact his or her mother.

 

[Beleth]

All the people volunteering to be sitters could have dead parents. You could then tell the ADCer that if there is someone on the other side of the screen, that person wants to contact his or her mother.

I have never seen a reader claim to be able to do this.

Besides, isn't this more of a test of remote viewing than of ADC?





(edited for formatting)

 

[Ladewig]

So spirits of dead people can communicate ( at various times) the following items to mediums:

their sex,
their cause of death,
their relation to the sitter,
names of relatives of the sitter,
initials of relatives of the sitter,
diseases the sitter had or has,
a description of the sitter's current or previous homes,
names of the sitter's current and previous pets,
the current or previous occupation of the sitter,
how much the spirit loves the sitter,
current or previous hobbies of the sitter, and
whether or not the sitter has possession of any of the spirits jewelry.

But. Spirits cannot communicate to mediums whether or not the sitter is the only other person in the room.

Also, as for the proposed test being more a test of remote viewing than ADC, how does your test distinguish between ADC and mind reading? If the subject knows all these things, perhaps the medium is simply reading the subject's mind.

 

[Beleth]

So spirits of dead people can communicate ( at various times) the following items to mediums:

their sex,
their cause of death,
[...]
But. Spirits cannot communicate to mediums whether or not the sitter is the only other person in the room.

I don't make the rules. I'm certainly not an apologist for readers. All I know is that there are certain things readers say they can do, and I have never seen a reader say they can use their ADC power to determine whether there's a living person in a room or not.

Also, as for the proposed test being more a test of remote viewing than ADC, how does your test distinguish between ADC and mind reading? If the subject knows all these things, perhaps the medium is simply reading the subject's mind.

That's true. ADC can look remarkably like mind reading. But ADC and RV look less like each other. They overlap at a few points, like this one, but other than at those points they are totally distinguishable.

 

[SteveGrenard]

Beleth:

I asked a researcher to have a look at your proposal. Here's what she wrote back. If you want to discuss this with her personally, I will give you her e-mail by Private Message.

"I think that you are all making it very difficult for yourselves. In the R/R MIA experiments -only statements are evaluated. The weighting of that statement is simply deduced as follows. If there are 30 participants in an audience, and everyone accepts a statement such as " your grandfather is in spirit" the weight of that statement is 1 - 30/30 = 0. If one person accepts a statement such as "your mother's name was Freda Forsesque Smythe" - the weight of that statement will be 1-1/30, which is very close to 1. The weight of a statement can never be more than 1. If a medium should offer a statement such as " I have a man with an m sound in his name" get rid of that medium from your experimentation, as this is rubbish."