Cold Reading Demos at TAM2

[CFLarsen]

T'ai Chi,

So you did read the papers! After all this time, finally you come clean. I have no idea why it is so difficult for you to simply say what you have read or what you mean, but obviously, you must have some agenda. Which you also keep hidden from others.

At any rate, tell us what they contain.

 

[Interesting Ian]

If you read the post of mine that you quoted you will see that I have already stated why I don't feel any obligation to discuss the point you want me to discuss.

Obviously I missed it. Repeat it please. What I want to know is why you refuse to answer my question.

 

[T'ai Chi]

T'ai Chi,

So you did read the papers! After all this time, finally you come clean. I have no idea why it is so difficult for you to simply say what you have read or what you mean, but obviously, you must have some agenda. Which you also keep hidden from others.

At any rate, tell us what they contain.

You know how to find them. Read them for yourself.

I don't want you to argue against my understanding of the paper; I want you to argue against the actual contents of the paper.

You refuse to do this for some odd reason.

 

[CFLarsen]

I don't want you to argue against my understanding of the paper; I want you to argue against the actual contents of the paper.

But you don't know what they contain.

 

[T'ai Chi]

But you don't know what they contain.

You know how to find them. Read them for yourself.

I don't want you to argue against my understanding of the paper; I want you to argue against the actual contents of the paper.

You refuse to do this for some odd reason.

 

[CFLarsen]

You know how to find them. Read them for yourself.

I don't want you to argue against my understanding of the paper; I want you to argue against the actual contents of the paper.

You refuse to do this for some odd reason.

I don't refuse at all. I just find it...well, interesting that you are seemingly unable to explain what they contain. You either don't know what they contain (in which case you are a liar) or you think it somehow serves a purpose (in which case you are deliberately obstructing a discussion (in which case you are a waste of time).

Why should we discuss with someone who either lies or does everything he can to obstruct a discussion? Perhaps both?

 

[T'ai Chi]

You know how to find them. Read them for yourself.

I don't want you to argue against my understanding of the paper; I want you to argue against the actual contents of the paper.

You refuse to do this for some odd reason.

 

[T'ai Chi]

Bumped for BillHoyt and Claus to comment on

BUMPED for BillHoyt and Claus to comment on.

Again, Claus, Bill, anyone; please show why the letter/name counts are independent. ie. show that:

P(first name is a J-name)*P(second name is a J-name) =
P(first name is a J-name AND second name is a J-name).

(where P(blah) means the 'probability of event blah')

Or, equivalently show:

P(second name is a J-name|first name is a J-name) = P(second name is a J-name)

(where P(A|B) is read as 'the probability of event A given that event B has occured')

That's all those who claim the letter/name counts to be independent have to show. I'll wait...



But see, I'm thinking that:

P(second name is a J-name|first name is a J-name) = 100%.

which does not equal P(second name is a J-name), because this event occurs only when the event 'first name is a J-name' occurs, and that event does not occur 100% of the time because there are names guessed that start with other letters. For P(second name is a J-name) to equal 100%, JE or a medium would have to always guess J-names, which clearly isn't reality.

Therefore the events are dependent, and we cannot take any analysis that treats them independent seriously.

However, Thanz counts are different, because he (my take on it, please correct if I am wrong) considers the letter/name counts to be independent BETWEEN readings reading for individuals, not within readings for individuals, whereby 'individuals' I mean the subjects in things like 'I see a grandpa', 'I am sensing a male figure', 'I am seeing a old female'; the individuals are the grandpa, the male figure, and the old female. That is, Thanz's probabilities are:

P(first name in second reading is a J-name|first name in first reading is a J-name) = P(first name in second reading is a J-name), an equality which is more likely to be satisfied and thus have the letter/name counts be independent.

 

[CFLarsen]

Nothing new....

 

[Darat]

Obviously I missed it. Repeat it please. What I want to know is why you refuse to answer my question.

Happy to oblige you on this issue for one last time:

To T'ai Chi:

"Hmm...

Let me help you out. I have answered the questions about what I actually said. I haven't answered Ian's question about what I didn't say...

Perhaps I am strange but I don't feel any sort of obligations to answer a question about something I didn't say. "

From you:

"Now of course this implies such a definition does set alarm bells ringing for Darat. But whatever causes such alarm bells ringing, he confirmed that it was not because he feels AC does not exist. So it is for some other reason. But Darat will not specify what this reason is. Why will Darat not specify this reason?"

And in return I've said:

"If you read the post of mine that you quoted you will see that I have already stated why I don't feel any obligation to discuss the point you want me to discuss."

So to summarise

I ask T'ai Chi a specific question about part of one of his posts that I would like an answer to, he answered my question totally and directly and asked me to elaborate. As he had answered my question directly and posted more information I responded to his request for me to elaborate.

And once again let me emphasis that I asked a very specific question to T'ai Chi that I wanted the answer to.

Is this some new guideline that we all have to give extensive reasons to every Tom, Dick or Harry as to "why" we asked our questions in the first place... ? LOL!

 

[T'ai Chi]

Nothing new....

Please, Claus, your comments, not your avoidance! Show us how rigorous you can be with your reasoning, and we'll see if it holds water... or if it is a sieve.

BUMPED, again, for BillHoyt and Claus to comment on.

Again, Claus & Bill; please show why the letter/name counts are independent. ie. show that:

P(first name is a J-name)*P(second name is a J-name) =
P(first name is a J-name AND second name is a J-name).

(where P(blah) means the 'probability of event blah')

Or, equivalently show:

P(second name is a J-name|first name is a J-name) = P(second name is a J-name)

(where P(A|B) is read as 'the probability of event A given that event B has occured')

That's all those who claim the letter/name counts to be independent have to show. I'll wait...



But see, I'm thinking that:

P(second name is a J-name|first name is a J-name) = 100%.

which does not equal P(second name is a J-name), because this event occurs only when the event 'first name is a J-name' occurs, and that event does not occur 100% of the time because there are names guessed that start with other letters. For P(second name is a J-name) to equal 100%, JE or a medium would have to always guess J-names, which clearly isn't reality.

Therefore the events are dependent, and we cannot take any analysis that treats them independent seriously.

However, Thanz counts are different, because he (my take on it, please correct if I am wrong) considers the letter/name counts to be independent BETWEEN readings reading for individuals, not within readings for individuals, whereby 'individuals' I mean the subjects in things like 'I see a grandpa', 'I am sensing a male figure', 'I am seeing a old female'; the individuals are the grandpa, the male figure, and the old female. That is, Thanz's probabilities are:

P(first name in second reading is a J-name|first name in first reading is a J-name) = P(first name in second reading is a J-name), an equality which is more likely to be satisfied and thus have the letter/name counts be independent.

 

[BillHoyt]

BUMPED, again, for BillHoyt and Claus to comment on.

N'it Wit,

I have already answered this question in several ways. You either didn't comprehend the answers or choose to continue playing churlish games. Here is a summary of the answers once again. Use your "vast" knowledge of statistics to figure them out. Or google till your fingers drop, you tr'oll:

1. Radioactive decay
2. Doubly stochastic process
3. Markov process
4. MMPP

The most salient answer, however, was given in my discussion with 69dodge about what the null hypothesis actually is. That one also appears to have gone right over your puny head.

 

[T'ai Chi]

N'it Wit,

I have already answered this question in several ways. You either didn't comprehend the answers or choose to continue playing churlish games. Here is a summary of the answers once again. Use your "vast" knowledge of statistics to figure them out. Or google till your fingers drop, you tr'oll:

1. Radioactive decay
2. Doubly stochastic process
3. Markov process
4. MMPP

The most salient answer, however, was given in my discussion with 69dodge about what the null hypothesis actually is. That one also appears to have gone right over your puny head.

(Bill's posterior posturings saved for posterity)

Jr,

Don't side track with your pathetic dodges of radioactive decay, stochastic processes, markov chains, blah blah. JUST ADDRESS the letter/name counts independence, and preferably, do this mathematically using the actual definition of independent events.

Listing things you believe are similar to letter/name counts is not providing evidence, Beavis.

I'll use my knowledge of statistics to predict you'll not provide evidence again. Don't prove me wrong.. you surely haven't in the past.

YAWN!

Sir, "statistics on the table, please".

 

[BillHoyt]

Don't side track with your pathetic dodges of radioactive decay, stochastic processes, markov chains, blah blah. JUST ADDRESS the letter/name counts independence, and preferably, do this mathematically using the actual definition of independent events.

Listing things you believe are similar to letter/name counts is not providing evidence, Beavis.

It provides more than ample evidence that your questions are faulty to begin with. I gave you the answers days, nay, weeks ago, and you've been whining ever since because you simply don't know what you are talking about.

I presented radioactive decay for a reason. You've had weeks to figure it out, but have failed. I presented Markov for a reason. You still don't get it. But your error is even worse, sir. You fail to understand that the question is completely wrong-headed.

Any sampling of names of real people would fit the Poisson model we've been dealing with for months in this question about JE. The null hypothesis tests for differences against this model. It is very simple, sir. If JE were calling out names from any random, representative list of names, any single initial letter would fit Poisson. When we look at JE's distribution, we compare it against this Poisson model. It doesn't fit. That is all we need to know, sir. If the null hypothesis were correct, it would fit. It doesn't. Null hypothesis rejected.

Now your specious question actually goes to the next question: why doesn't JE fit. You point to the multiple JE ejaculations and the fact that they are not independent. No, they are Markov, sir. I wrote that way back, but you ignored it. They are Markov in the same way radioactive decay is Markov. And, in the same, way, they should still result in a Poisson distribution, sir, just as radioactive decay can be modelled as Poisson. The fact is, of course, that both are doubly stochastic processes. But the fact is, of course, both should still fit a simple, single Poisson model.

Now I also flew way over your pea-brain when I mentioned MMPP, or Markov-moderated Poisson Processes. The ability to model MMPP as simple Poisson has been demonstrated yet again in papers on network traffic models. Network traffic is Poisson. Even though, clearly, the ability of one message to get through is dependent on the prior message's having been delivered. Network traffic is, in fact, doubly stochastic.

So, did the method distort the data because of its MMPP nature? No. The same MMPP nature applies to the other twenty-five letters of the alphabet. Something about JE's approach to mediumship causes him to keep double-ejaculating the Js more than any other letter. But, then, you never understood that argument from months ago? One wonders from what barn you garnered your "stat degree."

Now go google your little fingers off, pupdog. Learn a bit about doubly stochastic processes. MMPP. Radioactive decay. And stop playing games, you insufferable idiot troll.

 

[Thanz]

Now your specious question actually goes to the next question: why doesn't JE fit. You point to the multiple JE ejaculations and the fact that they are not independent. No, they are Markov, sir. I wrote that way back, but you ignored it. They are Markov in the same way radioactive decay is Markov. And, in the same, way, they should still result in a Poisson distribution, sir, just as radioactive decay can be modelled as Poisson. The fact is, of course, that both are doubly stochastic processes. But the fact is, of course, both should still fit a simple, single Poisson model.

I think that the problem still stems from equating the specific name "Joe" with the initial "J" when comparing it to the general population. The distribution of "Joe" in the population is obviously smaller than the distribution of all J names combined. I don't think that the same can be said for radioactive decay. Each decaying atom is much like any other. That is why I don't think that your analogy is correct.

 

[T'ai Chi]

(snip)

Mathematics Bill, mathematics. Not words, but symbols.

Thanks.

I presented radioactive decay for a reason. You've had weeks to figure it out, but have failed. I presented Markov for a reason. You still don't get it. But your error is even worse, sir. You fail to understand that the question is completely wrong-headed.

Don't list things which you believe are related. I'm not interested in you trying to convince me that a medium naming names is like radioactive decay.

Any sampling of names of real people would fit the Poisson model we've been dealing with for months in this question about JE.

Perhaps only if that nagging issue of independence actually held.

 

[BillHoyt]

I think that the problem still stems from equating the specific name "Joe" with the initial "J" when comparing it to the general population. The distribution of "Joe" in the population is obviously smaller than the distribution of all J names combined. I don't think that the same can be said for radioactive decay. Each decaying atom is much like any other. That is why I don't think that your analogy is correct.

Absolute rubbish, which has already been addressed.

 

[BillHoyt]

Mathematics Bill, mathematics. Not words, but symbols.

Thanks.

[/b]

Don't list things which you believe are related. I'm not interested in you trying to convince me that a medium naming names is like radioactive decay.

[/b]

Perhaps only if that nagging issue of independence actually held. [/B]

Th'is is the la'st ti'me thr'ough th'is, a'** h'ole. S'o l'isten up. Th'e nu'll hyp'othesis must f'it th'e mod'el, no't th'e mea'sured da'ta. W'e are tes'ting th'e mea'sured da'ta, du'fus. Th'e nu'll hyp'othesis ab'solutely f'its the mod'el. The da'ta fr'om th'e nu'll hyp'othesis are ind'ependent. Th'en w'e s'ee if the mea'sured da'ta f'its the mod'el. Th'ey didn't, d'ork. Th'at is ho'w sc'ience wo'rks, bo'zo.

Edited by Pyrrho: 

Edited to apply asterisks.

 

[michaellee]

THANZ and ERSBY---

Thanks to you both for your honest and thoughtful posts re:the JE independent vs. dependent count argument. I, as both of you, think I see both sides of the argument, and believe it to come down to a matter of semantics and/or personal belief on what side to look at and attack the problem from. For me, without trying to sound shrill, I believe in the long run, the question of independent vs. dependent "J", "Joe" etc.. counts to be toward the bottom of the entire investigatory scale when it comes to analyzing JE's techniques or expertise. A few less or more guesses does not deter my belief that JE is an expert cold reader, after reading many transcripts and other writings concerning cold reading of this fashion. But, of course, opinions do and always will differ, and arguing over trivial (yet important) aspects just seem to deter from the overall picture, encouraging back and forth repetitive argument while preventing further, meaningful analysis. Such is the reality when discussing this subject matter.

 

[T'ai Chi]

Th'is is the la'st ti'me thr'ough th'is, a'ss h'ole. S'o l'isten up.

Saved.

Bill,

You don't find it the slightest bit odd that you chose a model that has independence in for data that doesn't?

It seems to me you need to chose another model, or count differently.

But what do I know..