Case Study: The IIG Protocol for VFF

[Uncayimmy]

She does’t just claim to be able to see kidneys, she also claims to be able to do so with 100% accuracy. What’s unfair about a test that requires 100% accuracy to discontinues if she fails to achieve 100% accuracy? What reason is there to continue with a test that has already failed? Given the IIG test required 100% accuracy why did it continue after she had failed? If the answer is because it was in the protocol them my next question is why was it in the protocol of a test that was requiring 100% accuracy?

This is straight out of the VFF playbook: repetition. I've already addressed the bulk of the above.

If she knew that half the test people had a missing kidney and she correctly guessed the first half then she would automatically know what the second half had (no x-ray vision required). The test I’m suggesting is - “does this person have a missing kidney or not?”. It’s not - “one of these people has a missing kidney which one is it?”. If the particular person she is required to “scan” may or may not have a missing kidney then the odds for that single guess are 50/50. The person either does or doesn’t have a missing kidney.

You're not understand the full depth of this issue, which is the difference in odds between any answer and a specific answer. I'll lay this out from the very beginning. Correct me where you think I'm wrong.

You have two buckets, each containing a red ball and a blue ball. Blindfolded, you remove one ball from each bucket. What are the chances they match? 1 in 2. The two balls drawn could be R-R, R-B, B-B, B-R.

Now you only have one bucket. You draw a ball and ask me to tell you what you pulled because I say I am psychic. What are the odds that I guess correctly? 50-50. Suppose I say red. What are the chances I was right? Looking at the above paragraph, 50-50. Suppose I say blue? Again, 50-50.

If I'm right, you have no confidence whether I am psychic or not because half the non-psychic population would also get that one answer correct. So you decide to repeat this 10 times. That makes 2^10 or 1,024 possible sets of 10 answers. What are the odds of me getting all 10 right? 1 in 1,024.

If I answered all red? 1 in 1,024. If I answered all blue? 1 in 1,024. Alternating red/blue? 1 in 1,024. Every possible answer I could give has an equal probability of being right. If I give the correct answer, whatever it may be, you have to figure either I got very lucky or maybe I was somehow able to determine the answer through other means. This is why we run tests like this.

Now we dump 1,000,000,000 red balls into your bucket and repeat the experiment. Without knowing what I answered, what are my chances of being right? 1 in 1,024.

Suppose I answered all red. What are the odds this answer was correct? Virtually 100%. Look at any other possible answer I could give. What are the odds of it being right? Virtually 0%.

Do you understand what's happening here? Before we dumped the billion red balls into your bucket, the probability was evenly distributed for all answers. Afterwards, we effectively shoved 99.99% of the probability into one answer and distributed the remaining 0.01% among the other 1,023 answers.

So, while my chances of being correct are the same, the meaning behind me being correct is lost because the answer I gave had virtually no chance of being wrong. Likewise, if I had answered anything else, that answer had virtually no chance of being correct.

But you're going to argue, "You're only guessing. What are the odds of you choosing all red?" I argue that the goal of the test is to find out if I'm guessing or not. My correct answer has a different meaning depending on your distribution of red and blue balls. One is a 1 in 1,024 event and the other is a 1 in 1.00001 event (close enough). You have no way of distinguishing a lucky guess from "ability" if I'm right.

There’s also no reason for her to know how many possible test people there are or that the same test person may be being used more than once. In fact the test could be done with identical twins with one having a missing kidney. She wouldn’t know which twin was being used for any test and they could be used any amount of times.

If you want to put people back in the pool, you really only need one target and one decoy. Of course, you greatly increase the need for perfect blinding, especially in your test scenario where you give her immediate feedback. If she's right about the first one, then the test simply becomes about her ability to find that person again.

If you have a larger pool, you still have this issue but only to a lesser degree. In effect you have added a new blinding issue, which is her ability determine if she has seen a person before. And like I said, this is a problem because if she was correct once by chance she can be correct again based on her ability to recognize a person.

She would look silly if she passed on too many test people because she has never done this in the past

Never done what in the past? Passed on lots of people? She attends a university. She encounters probably hundreds of people per day. How do we know how many people she says she can't read?

Obviously she isn’t a lab rat, but neither are the people that agree to test her. Seems to me she expects the testers to navigate through more mazes than she is prepared to do.

So what? The bottom line is to get her to agree to a test where she is confident that her paranormal abilities will work and the IIG is confident they have eliminated all known means that pose a major risk. You're all hung up testing a particular subset of claims and conditions that you deem important.

As someone else said, if a guy claims he can shoot a fly off a sandwich at 1,000 yards while drunk, you don't actually need a sandwich, a fly and a bottle of whiskey to issue a challenge.

 

[Uncayimmy]

If she passed on so many people that the test couldn’t be completed then she would have failed to complete the test. She wouldn’t have failed the test as the test wasn’t completed.

Dear Anita,

Please send us a check for $750 so we may begin preparations for a test that may end up being inconclusive. Don't worry, though, we figure we'll only need to do this five or ten times. You've got plenty of time and money, right?

Love,

The IIG

 

[ynot]

UncaYimmy - "You have two buckets, each containing a red ball and a blue ball. Blindfolded, you remove one ball from each bucket. What are the chances they match? 1 in 2."

No it would 100% certain that they wouldn’t match because you said that there’s only A red ball and A blue ball in each bucket (only two balls of different colours). They wouldn't match even if they were in the same bucket. Even if you had a hundred of each coloured ball in each bucket the odds would still be 100% certain that the colour wouldn’t match if you took a ball from each bucket. One bucket = red balls and the other = blue balls.

Perhaps you meant she didn't know which bucket she was taking which ball from? In which case why complicate things by having two buckets? just have one bucket containing an equal number of red and blue balls.

 

[ynot]

Dear Anita,

Please send us a check for $750 so we may begin preparations for a test that may end up being inconclusive. Don't worry, though, we figure we'll only need to do this five or ten times. You've got plenty of time and money, right?

Love,

The IIG

"There only needs to be two test people involved. One with two kidneys and the other with one kidney. As long as she didn’t know which person she was testing at any time these two people could be used for any number of tests. She could confirm the suitability of these two people before testing began and no “free pass” option would be required."

 

[AvalonXQ]

UncaYimmy - "You have two buckets, each containing a red ball and a blue ball. Blindfolded, you remove one ball from each bucket. What are the chances they match? 1 in 2."

No it would 100% certain that they wouldn’t match because you said that there’s only A red ball and A blue ball in each bucket (only two balls of different colours).

A red ball and a blue ball in each bucket. As in,
Bucket 1: red ball, blue ball
Bucket 2: red ball, blue ball

 

[AvalonXQ]

UY's discussion on probabilities is very important, because someone who is "playing the odds" can potentially do better than someone who is randomly guessing.
Assume, for example, that we showed Anita 20 subjects, of which 4 were missing kidneys. Now assume that she passed on 10 and guessed "both kidneys present" on the other 10. The odds of the people matching her guesses is far higher than 1 in 2^20.

ETA: I think the odds would actually be (10*9*8*7)/(20*19*18*17).

 

[ynot]

A red ball and a blue ball in each bucket. As in,
Bucket 1: red ball, blue ball
Bucket 2: red ball, blue ball

Why make it so unnecessarily complicated?

Why not just provide a ball that could be either red or blue and get her to say which colour it is? Simple 50/50 odds that everyone understands and can't be disputed.

 

[ynot]

UY's discussion on probabilities is very important, because someone who is "playing the odds" can potentially do better than someone who is randomly guessing.
Assume, for example, that we showed Anita 20 subjects, of which 4 were missing kidneys. Now assume that she passed on 10 and guessed "both kidneys present" on the other 10. The odds of the people matching her guesses is far higher than 1 in 2^20.

ETA: I think the odds would actually be (10*9*8*7)/(20*19*18*17).

Why make it so unnecessarily complicated?

Why not just provide a person that has either one or two kidneys and get her to say how many kidneys the person has? Simple 50/50 odds that everyone understands and can't be disputed.

 

[Uncayimmy]

UncaYimmy - "You have two buckets, each containing a red ball and a blue ball. Blindfolded, you remove one ball from each bucket. What are the chances they match? 1 in 2."

No it would 100% certain that they wouldn’t match because you said that there’s only A red ball and A blue ball in each bucket (only two balls of different colours). They wouldn't match even if they were in the same bucket. Even if you had a hundred of each coloured ball in each bucket the odds would still be 100% certain that the colour wouldn’t match if you took a ball from each bucket. One bucket = red balls and the other = blue balls.

You are growing tiresome. You ignore the questions put to you. You repeat your arguments rather than defend them. And now this. I have no idea what you are trying to say except perhaps that you think I somehow incorrectly set up the problem with unclear language. I'll address that in a moment, but even if I had simply misstated something, you knew what I was driving at and chose to ignore it, just like you keep ignoring other points I and others make. You're making VFF look good.

As to how you're trying to twist my scenario around, I'm actually at a loss. I said that each bucket contains a red ball and a blue ball. How do you then conclude that "One bucket = red balls and the other = blue balls?"

Each - every one of two or more considered individually or one by one.

I don't know how to be any more clear than that.

 

[ynot]

You are growing tiresome. You ignore the questions put to you. You repeat your arguments rather than defend them. And now this. I have no idea what you are trying to say except perhaps that you think I somehow incorrectly set up the problem with unclear language. I'll address that in a moment, but even if I had simply misstated something, you knew what I was driving at and chose to ignore it, just like you keep ignoring other points I and others make. You're making VFF look good.

As to how you're trying to twist my scenario around, I'm actually at a loss. I said that each bucket contains a red ball and a blue ball. How do you then conclude that "One bucket = red balls and the other = blue balls?"

Each - every one of two or more considered individually or one by one.

I don't know how to be any more clear than that.

Yes I can tell from your body language that you’re becoming somewhat annoyed.

Sorry I read your balls wrongly. I now fully understand what you meant and admit my mistake (rushing).

 

[Uncayimmy]

"There only needs to be two test people involved. One with two kidneys and the other with one kidney. As long as she didn’t know which person she was testing at any time these two people could be used for any number of tests. She could confirm the suitability of these two people before testing began and no “free pass” option would be required."

You just finished explaining about how you would allow incomplete tests. Is this your backhanded way of saying you now see the err of your ways?

I also addressed the issue of using the same subject more than once. Did you even bother to read it? It doesn't seem like it because you are now telling her in advance who is who, which makes it even harder for you to blind her to the identity of each person. Smell, which was not an issue in the IIG, becomes an issue.

You have yet to propose anything that is simultaneously more effective, easier to accomplish and likely to be accepted the claimant. In the process you have refused to address objections presented. And yet you complain about how ridiculous claimants can be in agreeing to protocols.

Amazing.

 

[Uncayimmy]

Yes I can tell from your body language that you’re becoming somewhat annoyed.

Sorry I read your balls wrongly. I now fully understand what you meant and admit my mistake (rushing).

Fine. Then please continue through my entire explanation and tell me where you think I went wrong. If you cannot find anything wrong, then do you now understand how it's actually better to have an equal number of targets and decoys and how a claimant is not likely to accept anything else given your test scenario?

 

[ynot]

You just finished explaining about how you would allow incomplete tests. Is this your backhanded way of saying you now see the err of your ways?

I also addressed the issue of using the same subject more than once. Did you even bother to read it? It doesn't seem like it because you are now telling her in advance who is who, which makes it even harder for you to blind her to the identity of each person. Smell, which was not an issue in the IIG, becomes an issue.

You have yet to propose anything that is simultaneously more effective, easier to accomplish and likely to be accepted the claimant. In the process you have refused to address objections presented. And yet you complain about how ridiculous claimants can be in agreeing to protocols.

Amazing.

So a person has to stick with all original concepts and isn’t allowed to refine for improvement?

It’s not so much that “free passes” from using many test people that weren’t pre-confirmed as being suitable was an “err of my ways” as it is that no “free passes” from using only two people that are pre-confirmed as being suitable is better.

There’s no reason to suspect that Blondie’s powers of normal observation are any better than anyone else’s. Obviously if only using two people repeatedly it would have to be pre-tested that it wasn’t possible to distinguish between the two test people using normal observational means. I don’t see that this would be too hard to do.

 

[ynot]

Fine. Then please continue through my entire explanation and tell me where you think I went wrong. If you cannot find anything wrong, then do you now understand how it's actually better to have an equal number of targets and decoys and how a claimant is not likely to accept anything else given your test scenario?

It’s not a matter of right or wrong. It’s a matter of which testing method is best.

You seem to be taking this very personally.

 

[VisionFromFeeling]

I would not agree to a test that uses the same two subjects repeatedly. Since I do require to see their clothed backs, it would be impossible to prevent visual indication of who is which. Minor size and shape differences, but also breathing and movement. Humans are excellent observers and we are constantly reading subtle visual information and analyzing other people based on that. I would be unable to disable the automatic cold reading skill that we all have and always use. And the purpose of a test is to block out any possibility of finding the right answers by cold reading.

So a test involving the use of the same two, or few, subjects repeatedly simply would not work. I am not interested in obtaining false accuracy that would indicate towards a skill that I do not have.

Unfortunately, testing this claim does require new subjects in each trial. And that is what makes it difficult, but it is quite necessary.

 

[Gaspode]

Some posts moved to the general discussion thread since they are off-topic for this thread. ETA: A few moved to AAH.

Please keep this thread on-topic i.e. the IIG protocol.

Replying to this modbox in thread will be off topic  Posted By: Gaspode

 

[AvalonXQ]

Fine. Then please continue through my entire explanation and tell me where you think I went wrong. If you cannot find anything wrong, then do you now understand how it's actually better to have an equal number of targets and decoys and how a claimant is not likely to accept anything else given your test scenario?

Why wouldn't the claimant accept an unidentified ratio of one- and two-kidney subjects? If she can tell when she's right and when she's not, the pass protocol should be enough.

 

[Uncayimmy]

So a person has to stick with all original concepts and isn’t allowed to refine for improvement?

Oh, the irony! She is thick.

There is a difference between jumping around from idea to idea and actually discussing the objections to your ideas when people take the time to address them.

It’s not so much that “free passes” from using many test people that weren’t pre-confirmed as being suitable was an “err of my ways” as it is that no “free passes” from using only two people that are pre-confirmed as being suitable is better.

Merely stating something is "better" with explaining why or addressing the concerns already present is like, I dunno, saying you can detect kidneys through a shirt but not through a screen.

There’s no reason to suspect that Blondie’s normal powers of observation are any better than anyone else’s.

Actually, there is a very good reason for suspecting that. If you're wrong, you might lose $50K and give the woos of the world their first ever success at a paranormal challenge.

Obviously if only using two people repeatedly it would have to be pre-tested that it wasn’t possible to distinguish between the two test people using normal observational means. I don’t see that this would be too hard to do.

Yet another unsupported assertion. You need to spend less time writing and more time thinking. Only one person needs to be identifiable to ruin the entire test. This means you have to have perfect blinding visually, you also need to blind other senses as well, to two most important of which are smell and hearing.

Smell: People smell differently. It's not just a matter of perfume or cologne. There's soap and shampoo. There's armpit stink. There's general body odor, which can vary from day to day (I have a friend who stinks the day after eating garlic). There's flatulence (who wants to lose $50K because a subject had Taco Bell for lunch?). There's the smell on the clothes. Short of glass cages, the only way to control for this is to get the claimant to agree to wear a nose plug. Good luck with that.

Hearing: People breathe at different paces. Some like to sigh. Sometimes people are sleepy and yawn. Some grind their teeth. I have the sniffles today, which is clearly audible. Short of glass cages the only way to blind for this is ear plugs or head phones with white noise. Again, good luck with that.

Vision: Even if you put each person in a burqa, you need to be 100% sure that there's not a single little errant stitch or other mark that distinguishes the two. If you accomplish that, you still have the issue of the pace at which people breathe, yawning, sighing, and other mannerisms which may differ by person.

Your proposal is the least likely to be accepted by a claimant. I wouldn't blame her for refusing to perform wearing headphones and a nose plug. I wouldn't blame her for refusing to be in a glass cage nor would I expect it to be easy to find people in glass cages.

Even if the claimant agrees and the subjects agree, you still have the issue of blinding visual cues. The IIG would be foolish to attempt such a test.

 

[Uncayimmy]

It’s not a matter of right or wrong. It’s a matter of which testing method is best.

You seem to be taking this very personally.

Whether I am taking it personally or not is irrelevant. Why would you even bring that up? It doesn't change the value of my arguments. Speaking of which, your arguments are sorely lacking in substance beyond mere assertions.

 

[ynot]

I would not agree to a test that uses the same two subjects repeatedly. Since I do require to see their clothed backs, it would be impossible to prevent visual indication of who is which. Minor size and shape differences, but also breathing and movement. Humans are excellent observers and we are constantly reading subtle visual information and analyzing other people based on that. I would be unable to disable the automatic cold reading skill that we all have and always use. And the purpose of a test is to block out any possibility of finding the right answers by cold reading.

So a test involving the use of the same two, or few, subjects repeatedly simply would not work. I am not interested in obtaining false accuracy that would indicate towards a skill that I do not have.

Unfortunately, testing this claim does require new subjects in each trial. And that is what makes it difficult, but it is quite necessary.

Whereas I might admire your honesty in not wanting to be able to cheat, I can’t help but also suspect that you merely want to make the test unnecessarily difficult for reasons best known to yourself and suspected by others

I would be very confident that I could present you with the clothed backs of two people and you and nobody else wouldn’t be able to tell the difference. This isn’t some thing that needs to be argued as it could be very easily tested before any real testing started. If it could be pre-test proven to you that you (and nobody else) couldn‘t distinguish between the clothed backs of two people, would you agree to this test? If not why not?