SCEPCOP Forum

You are the one who said that we were doing a few trials and then we would develop a protocol.

I'm not backing down from anything. You want to do a test, let's do a test. But doing one trial proves absolutely nothing- whether you get them all right, or none of them right. You've been doing this for so long and haven't yet realised that? No wonder you have so many problems when you deal with skeptics.

If you want to draw any ACTUAL conclusions I'm game. But let's do it right. I'm not going to do some piecemeal thing where we have no real idea of what the results mean.

You did a subjective demo but when I want to make a guess from 4 options you want a water tight legal contract with all the statistical stuff and so on?

What changed? I have written about this flaw in the skeptics protocol numerous times.

There is a protocol to score a single hit. And there is a protocol to repeat trials to break certain odds.

Because skeptics demand both at once, they never get around to testing if he can do it even once.

Too much red tape... I guess I'll never get to have a single guess, why? .... you have to write a legal contract and a brief description of what you can do simultaneously.

Anyway you keep ridiculing my protocol and my stats with me doing everything and everybody else just abusing me.

I think this is going nowhere, I've seen these signs on 50 skeptics before, once you have your back up there is no test happening ever.

RESULTS:

1 in 4 right = average.

Doing a lot better than that = it's working

Number of trials = when Arouet says enough

You don't know any maths, you want me to go through the binomical distribution step by step another 5 times?

I can tell you the number of trials and hits required to defeat 1000:1 or 1000000:1 but you just wont listen.

Attempting to guess the 4 words in order IS THE SAME AS having 1 guess from 4 options.

I'm allowed to repeat my guesses.

e.g.

FRIDGE
MAGNET
TIRE
MAGNET

each has 1 chance in 4 of being right.

I can't write a protocol that is clear and straighforward as possible
AND prove paranormal to set objective set odds in the one document
that any of you would understand.

If you say the protocol you just quoted is untestable then we are done.

If you want to try my protocol then post 4 words.

Unless you have a change or addition to the protocol, stop whining about it
in general terms.

I'm trying to guess a word Arouet, not program the space shuttle computer.

Are you proposing to do all four guesses at once? Or one at a time?

If its 4 at once then I don't know what the odds are of getting one or two in the right order. I don't know that its as simple as 1/4.

If it's one at a time, then previous guesses can give you information that changes the odds for the remaining guesses.

Are you in such a hurry to get going that you don't want to spend some time figuring out a good protocol?

G. Adam, with all due respect you are going at this all wrong - at least from the perspective of trying to figure out if something real is going on. The only real way we can draw conclusions is over many trials - not one or two. If you want to just try one or two to see how it works in practice, that's fine, but we can't draw conclusions from it.

Also, the stats here aren't that simple - at least not to me. For example: you are allowing yourself to double up words. If you were to double up two words I would think that would greatly alter you odds (favourably) of getting two right. Doubling up words changes the stats in ways that I don't fully understand - the thing is, I don't get the sense you fully understand it either? What happens if you triple up one word? then you have huge odds of getting one word right, and 25% chance of getting the other one right.What does this do to the overall odds? I don't know. But again, I don't think you do either, or how these multiple guesses should be scored, or combined.

I know these are just informal tests, and we're not doing parapsychology here, but there are enough stat freaks hanging around the skeptical forums that it really doesn't make sense to do this without at least casual involvement from one. Otherwise what possible good could this do other than as a mildly interesting time passer?

I also don't think you have any clear idea of exactly what would be required to achieve statistical significance here. If you don't double up words and guess randomly, what are the odds of getting 1 or 2 hits? Or 3? I don't think you know. Or at least, you haven't demonstrated that you know.

I am actually trying to help you here. This seems to be something you have obsessed over for a long time. Part of the problem may be the haphazard manner that you go about this. For someone who has done 20,000 trials, you seem to be going about this in a rather unfocussed manner.

uhhh right! I did 2nd year stats at university and I gave you a complete list of the probabilities from the binomial distribution.

Let's give it a rest and when you've consulted with all the highly competent skeptics statisticians (0) give me a hoy.

analogy of Michael Jordan getting grilled for not knowing how to shoot hoops from girl scouts comes to mind..

On another topic I PROGRAMMED the test at http://paranormaltest.com

A computer can randomly select the word and page and line numbers, need another hour to do the 'select from 3 words' part then I'll do a HIGH SCORE board. Check it out!

This entire protocol is about 50 lines of code.

You didn't respond to any of the points I raised. Are you saying it makes no difference if you double up words?

Again, Mr. I-did-two-stat-courses, set out the complete protocol, including how many trials you think are necessary to get a good confidence level, how the scores are to be tabulated, etc. and I'll get someone to look it over. I'm not saying that you don't know the stats, I've said that you haven't shown that you know the stats. The fact that you think two trials can ever mean anything significant is concerning to me about your stat knowledge. Stats are hard stuff, you've got PHD scientists who completely bungle stats in their experiments. I am not pretending to be competent in stats either.

So, if you just want to do a couple trials so you can get a feel for how it works, then fine. But we're not going to draw conclusions from it. I'll pick four words (to keep it simple we'll liimit it to nouns) and PM them to derrida and Ninja, then send you the numbers with the four words in random order.

If you want to do something more meaningful I'm up for that. I'm even up for doing some legwork to verify, but I'm not going to try and come up with a protocol myself.

I think you missed what I said about the 2 trials, hint I capitalised THEN shortly afterwards.

I probably should have clarified a guessing-protocol not a final proof-protocol.

You seem to be going towards the JREF contract route, there is only one shot at that in a probably foriegn setup for the psychic.

My circle of friends are not in on any of this... I try to be a regular dude in person so I need the practice.

Yeh shoot, 4 nouns, then I'll give 4 quotes.

Patch to the protocol.
Say something to identify the words in the test, that you PM'd them, concentrated etc.

------ SAMPLE POST -----

I PM'd the 4 words to NP and Derida, ...

Word 1
page 123 line 22

Word 2
page 100 line 10

Word 3
page 44 line 1

Word 4
page 200 line 20

---- END SAMPLE POST ----

Ok, I've pm'd 4 words to Ninja and derrida chosen by a random word chooser. Here are the 4 words ordered by random.org.

1. rattle
2. motorway
3. thesis
4. facility



Here are the 4 sets of numbers, also chosen by random.org:

1. 185/2
2. 52/11
3. 186/13
4. 157/8

#4 works too, GAVE HIM A ... RATTLE

4 Good channels, let's hope the protocol works!