Merged New telepathy test: which number did I write ?

[SezMe]

Suppose we let the testee discard up to X% of all responses, based on any criteria he sees in the clear text, ....

That might work if the acceptance/rejection criteria were specified as part of the protocol and were clearly spelled out in such a manner that any observer could ascertain whether the protocol was being followed. Michel doesn't want to do that. He wants to accept or reject data based on the attitude of the source of the data. That's a complete no-no.

 

[xtifr]

That might work if the acceptance/rejection criteria were specified as part of the protocol and were clearly spelled out in such a manner that any observer could ascertain whether the protocol was being followed. Michel doesn't want to do that. He wants to accept or reject data based on the attitude of the source of the data. That's a complete no-no.

I don't think it's a complete no-no as long as all rejection takes place before Michel learns what answer was provided. That's where the last test went wrong, and even more so, the one before it. Once he knows what answer a person gave, we can no longer trust him to be objective about their trustworthiness. Before that, though, he should have no particular reason to be biased for or against any given person.

That said, the fact that we're on the third go-round, and Michel *still* isn't offering an decent protocol—one where the results would be clear and unambiguous—suggests to me that he's scared of what a proper test might show.

I think he's suffering from cognitive dissonance. He desperately wants to prove he's right, but is still terrified that he might find out he's wrong, and as a result, is unable to consider proposing an unambiguous test. It doesn't matter as much what the rest of us think (though he'd like to convince us), as long as he's able to somehow convince himself that the results were what he wanted. Which is basically what happened in the last test. And each time he manages to convince himself that he's won, he comes back suggesting the addition of the smallest bit of extra rigor he can get away with, in the desperate hope that maybe it'll still work and maybe this time he'll convince someone else too.

It's honestly all a little sad. And I don't mean that in a condescending way. I genuinely feel sorry for him.

 

[Innocuous]

He wants to accept or reject data based on the attitude of the source of the data. That's a complete no-no.

But why is it a no-no? I am probably expressing myself badly, so let me try again.

Yes, the credibility rating lets the testee reject some data, and subjectively that troubles me. But objectively, he can't use that to improve his results. He can't game the system. His rejections aren't based on the actual numbers he needs to match - he doesn't know those when he makes the rejections. If he doesn't have any special powers, he is just randomly throwing away some of the data. As long as we have enough data left, he isn't going to improve his success rate - he doesn't know if he is throwing away hits or misses. As a percentage, he will be throwing away same-as-chance successes and same-as-chance failures, and the net result will still be same-as-chance.

Except for the forum-related limitations I mentioned in my last post, the only reason I can understand for disallowing the credibility test is to control for some skill or power that lets him predict with better-than-chance accuracy someone's number based on their "attitude text". Is that the reason to disallow the credibility test? Because without such a skill, he can't improve his success rate, and his rejections will not change the outcome.

Look at it another way. Suppose someone claimed to be able to psychically view in detail the deaths of the deceased parents of a random stranger, but he needed to see the stranger first to know if he could connect with him psychically. Even with that limitation, it is an extraordinary power. Surely you wouldn't refuse to test his claim because he is affecting the data selection? I hope not. Instead, that limitation of his power would be accounted for in establishing the testing parameters. One of those parameters might be this: at least X% of Y candidates must be selected for testing, otherwise the test fails.

 

[Innocuous]

I don't think it's a complete no-no as long as all rejection takes place before Michel learns what answer was provided.

Right. I have been assuming this is part of the revised protocol, but it should be stated explicitly. Once credibility is assigned, he can throw away the "bad" responses, but then they are no longer part of the test. We don't even need to see what the chosen numbers were for those responses. In fact, we should NOT look at them and tempt post-test rationalizations.

 

[xtifr]

Right. I have been assuming this is part of the revised protocol, but it should be stated explicitly. Once credibility is assigned, he can throw away the "bad" responses, but then they are no longer part of the test. We don't even need to see what the chosen numbers were for those responses. In fact, we should NOT look at them and tempt post-test rationalizations.

That's what happened last time, but he didn't like the results, so he rejected several people after the fact (after he'd seen the results) on the flimsy excuse that they hadn't followed optional parts of the protocol (and, moreover, parts that logically couldn't have affected the results in any case). That post-reveal rejection was a much bigger violation of the agreed-upon protocol than anything anyone else did, but he refuses to see it that way.

Bottom line, if he proposed a reasonable test instead of all this nonsense, none of this niggling would be necessary. Have us pick a number between 1 and 1000, and even a single hit would be impressive enough to warrant further study, no matter how many people supposedly lie about receiving his transmissions. (Unless there were a whole lot more participants than in any of the previous tests.) Doing anything less is simply a waste of everyone's time.

 

[Ladewig]

It is perhaps part of a skeptics' strategy, to try to impose a mediocre protocol. Then, the test fails, and skepticism seems victorious. Very smart, isn't ]

If you really believe this, then you would abandon the idea of doing a test here. There is not a single poster in this entire thread that believes your application of the protocol is worth anything. You are convinced that several people in this thread tell direct and indirect lies in order to minimize the extent of your powers. In fact, from you previous evaluations of our answers, you believe we take turns lying to you (a single poster may get a negative credibility rating in one guessing game and end up with a positive credibility rating in another - but according to you someone is always lying).

This forum, therefore is the most useless place to run a test.

In addition, every third poster urges you to see a doctor.

.............

Please see a doctor.

 

[fromdownunder]

There is not a single poster in this entire thread that believes your application of the protocol is worth anything.

To me one of the major issues is that Michel H does not even comprehend that even if a poster sounds "nice" that 99% of the people on his threads are simply taking the pith, and do not give a flying **** about his failed claim. If a poster sounds "nice" as already shown several times on this latest iteration of the BIG TEST Michel assumes sincerity, and honesty, which shows that Michel simply not only cannot read emotions from the written word (well, who can?) and negates his assumption that he can conclude who is genuine and who is not.

(Subtle) sarcasm and facetious commentary are simply beyond his comprehension. Which places further errors into an already failed protocol.

Norm

 

[AdMan]

(Subtle) sarcasm and facetious commentary are simply beyond his comprehension. Which places further errors into an already failed protocol.

^This.

 

[fromdownunder]

Once credibility is assigned, he can throw away the "bad" responses, but then they are no longer part of the test. We don't even need to see what the chosen numbers were for those responses. In fact, we should NOT look at them and tempt post-test rationalizations.

The other problem with Michel is he can claim to throw away "bad" responses, then six months later on a completely different thread and totally out of the context of the thread, claim that one of the answers he threw away in the original test was actually evidence for his telepathy. Bad "thrown away stuff" suddenly became good "proof" with the passage of time, and he changed the context of the events described, and probably had the idea that no one would actually investigate true facts of this particular lie:

http://www.internationalskeptics.com/forums/showthread.php?t=275465

So even when he loses, he wins in his own mind.

Norm

 

[Akhenaten]

But why is it a no-no? I am probably expressing myself badly, so let me try again.

Yes, the credibility rating lets the testee reject some data, and subjectively that troubles me. But objectively, he can't use that to improve his results.

He certainly can, by deciding whether to reject or retain data after he knows the answers.

He can't game the system.

He can and does.

If he'd accept a properly blinded test he couldn't, however, and that's why he refuses such a test.

His rejections aren't based on the actual numbers he needs to match - he doesn't know those when he makes the rejections.

Yes he does.

If he doesn't have any special powers, he is just randomly throwing away some of the data.

No, he throws away data that doesn't match the answers he wants, after the fact.

As long as we have enough data left, he isn't going to improve his success rate - he doesn't know if he is throwing away hits or misses.

Why do you keep insisting this is the case when it's obvious that you haven't looked at the results of the previous iterations of this test?

<snip>

You really should either read this whole thread or start taking notice of what people are telling you about the post hoc nature of Michel's rejection of certain data.

Michel makes the final decision on whether a response is acceptable after he knows the number which the respondent guessed.


Do you understand?

 

[Innocuous]

You really should either read this whole thread or start taking notice of what people are telling you about the post hoc nature of Michel's rejection of certain data.

Michel makes the final decision on whether a response is acceptable after he knows the number which the respondent guessed.


Do you understand?

No. I understand the first iteration had this problem. But we are talking about a revised protocol:

...In this thread, like in my previous test on this forum, I want to evaluate credibilities without any knowledge of the number you picked
...After a reasonable number of forum members have validly answered (if this "reasonable number" is ever achieved ), I should post my "credibility ratings" for all valid answers to this test (and also the number I wrote and circled). When this is done, you should post your numerical answer

The credibility ratings are evaluated and posted before the answers are revealed. This is why I don't understand the objections. As long as we set up the success or failure criteria before the test, the credibility ratings can't be retroactively reinterpreted. We don't need to even see the rejected responses.

I understand that previous tests have involved post hoc rationalizations. I thought we were trying to come up with a protocol to avoid those. That is all. My apologies if I have offended anyone.

 

[Michel H]

No. I understand the first iteration had this problem. But we are talking about a revised protocol:
...

The credibility ratings are evaluated and posted before the answers are revealed. This is why I don't understand the objections. ... the credibility ratings can't be retroactively reinterpreted. ...
... I thought we were trying to come up with a protocol to avoid those...

Good, Innocuous. I acknowledge that the protocol I use in this test has been substantially infuenced by the feedback and suggestions made on this forum.


Newcomers in the thread, please go to post #1031 to find the opening post of this test. Please, do answer! I need more answers (only two valid so far)

 

[Pixel42]

The credibility ratings are evaluated and posted before the answers are revealed. This is why I don't understand the objections.

Because that was the agreed protocol last time too (the second iteration). And yet Michel revisited and discarded or changed his pre-reveal ratings after the answers were revealed. There is no reason to believe he won't do the same thing again this time if, once again, the initial results aren't to his liking. It's clear he will do whatever it takes to maintain his mistaken beliefs, which is why hardly anyone is indulging him this time.

 

[Innocuous]

Bad "thrown away stuff" suddenly became good "proof" with the passage of time, and he changed the context of the events described...So even when he loses, he wins in his own mind.

The fact that exactly this happened, so then he went and further filtered the answers, after he knew the results, in order to get the percentage of hits he was looking for.

That's what happened last time, but he didn't like the results, so he rejected several people after the fact (after he'd seen the results)...

I see. So it isn't so much that the credibility ratings are necessarily bad in a proper test, but in this particular case it gives the testee an "out" if he chooses to ignore the test criteria previously agreed on. That helps to explain it - it isn't a mathematical objection, but more of a pragmatic one.

Bottom line, if he proposed a reasonable test instead of all this nonsense, none of this niggling would be necessary. Have us pick a number between 1 and 1000...

Agreed. That just works.

 

[Innocuous]

Good, Innocuous. I acknowledge that the protocol I use in this test has been substantially infuenced by the feedback and suggestions made on this forum.

Michel H, would you also agree that we must define (mathematically and precisely) the success and failure criteria for the entire test before the test begins, and that you have to hold yourself strictly to that interpretation and no other interpretation?

Would you also agree that if the test permits you to throw away some responses because of their credibility rating, that those responses will never be revealed to anyone? So no one will ever know what the numbers for those responses were (except the responders, individually).

 

[Michel H]

Michel5,

I am curious why you reject the "big random number" idea? It would eliminate the possibilities of cheating, poor statistics, and choosing a favorite number between 1-4 that have been brought up here. It is a lot more simple protocol and would eliminate the complexity and concerns about you choosing the "legitimate" guesses.

In terms of your current protocol: I think others have brought this up, but what if I tell you in my open post that I hear the number very clearly, but I am simply lying to you and I am randomly chosing a number? You would consider this a legitimate choice, but it is no better that if I was honest and told you I heard no number at all in my brain (or heard it correctly but chose to lie). You would accept my dishonest guess just as you would accept an honest guess. Both would throw off the statistics. You seem to believe people lie in telling you numbers- why wouldn't they lie in telling you how reliable a choice they made?. If you want a real test, just find a random number between 1 and 1,000,000 and accept all guesses. The lies would disappear in the noise.

At the beginning of my online telepathy tests, I was using a greater number of possibilities, I even used a three-digit target. But these tests generally did not produce good results. Here is an example (with a target between 1 and 100): answers.yahoo.com/question/index?qid=20091010125251AAOZnTS.
Here, you can see that the test was (superficially at least) a total failure, a real skeptical paradise . One of the difficulties in these tests is that people might think that they're doing a "good enough job", if they give an answer which is related (but not necessarily equal) to the target, from example, they might give just one digit right (or answer 53 instead of 55), and you have to try to figure out if the answer was (in a sense) "close" to the target (and good luck with that).

You said:
I think others have brought this up, but what if I tell you in my open post that I hear the number very clearly, but I am simply lying to you and I am randomly chosing a number? You would consider this a legitimate choice, but...
Well, this would be a cynical attitude. This may exist, but seems less common than "ordinary people" (I don't want to sound offending here), which are neither totally black, neither totally white. You have to realize that most people are not totally immoral, they usually give some information, even if the "purity" of "skeptics" is probably partly an illusion.


Newcomers in the thread, please go to post #1031 to find the opening post of this test. Please, do answer! I need more answers (only two valid so far).

 

[Akhenaten]

No. I understand the first iteration had this problem. But we are talking about a revised protocol:

Revised or not, Michel has a record of ignoring the protocol and refiltering the results after the numbers have been revealed.

The credibility ratings are evaluated and posted before the answers are revealed.

And then re-evaluated after the numbers are revealed in order to skew the result.

This is why I don't understand the objections. As long as we set up the success or failure criteria before the test, the credibility ratings can't be retroactively reinterpreted. We don't need to even see the rejected responses.

Tell that to Michel. He's the one who ignores this requirement.

I understand that previous tests have involved post hoc rationalizations. I thought we were trying to come up with a protocol to avoid those. That is all. My apologies if I have offended anyone.

I don't think you're offending anyone but you do seem to be completely unfamiliar with Michel's MO.

 

[Innocuous]

I don't think you're offending anyone but you do seem to be completely unfamiliar with Michel's MO.

I thought I had read the entire discussion, but I am thinking now I must have missed one iteration - I thought we were on the second.

 

[AdMan]

I thought I had read the entire discussion, but I am thinking now I must have missed one iteration - I thought we were on the second.

Michel H cheats when his tests don't go as he wishes. He has used his "Credibility Rating" criterion in order to do so. This is one main reason why many here object to having that in his protocol, aside from all the other reasons why the protocol is flawed.

 

[Akhenaten]

At the beginning of my online telepathy tests, I was using a greater number of possibilities, I even used a three-digit target. But these tests generally did not produce good results.

Only because you're defining "good" as "supportive of my claim".

Here is an example (with a target between 1 and 100): answers.yahoo.com/question/index?qid=20091010125251AAOZnTS.
Here, you can see that the test was (superficially at least) a total failure, a real skeptical paradise.

The test was successful. The only failure was your claim.

What you're referring to as a skeptical paradise is what's more commonly known as reality.