Anyone heard about this dude? (pavelprorok on youtube)

[Startz]

Okay, you guys both agree , it must be that 66 does beat the 1000 to 1 odds, so that will be the number required. We'll see what JREF says, RemieV didn't seem prepared to accept the 50/50 protocol.

One question, do your calculations show that 63 yields a probability less .01? Just curious as to whether I should trust those tables. That huge spread between 63 and 66 is counterintuitive.

Your intuition is pretty good, the table doesn't mean quite what you think it does. (I had to read the online explanation five times to understand the table.)

The probability of getting getting k or more right out of 100 is
k prob
63 0.0060
64 0.0033
65 0.0018
66 0.0009

The tables use the two-tailed probability rather than the one-tailed probability. For example, they are based on the probability of getting 63 or more or 37 or fewer. That's not Pavel's claim.

And, as you've said, Remie and others at JREF may not like this part of the protocol anyway.

 

[pavel_do]

Pavel, could you do it three pictures at a time instead of two? This would greatly change the odds and therefore the time taken to perform the test.

Ok, I think I could try it, just need to redo sets of two in sets of 3, I just have to make sure that all three are different. In order not to get confused and again I would need to know what are the 3 “options” for my final decision same as with pairs… I would make my decision based on my vision and descriptions that I would do, while I hold the envelopes. I could try a few runs to train if it worth it. Can anyone calculate, how many you have to get right to beat 1.000 odds… If i wold have say, 10 sets by 3 per "pair" (that would make 30 photos in all) and have to identefy 1 from each set..

 

[pavel_do]

Pavel, let me reiterate: You will need to focus on the protocol. The JREF - and most forum members - do not care at all for theories or explanations. Save that for interviews in the months after you won.

Your verbosity paired with your (understandable) difficulties with english grammar lead to an exponantial increase of possible misunderstandings. Often, people have to assume what you mean, instead of clearly understanding what you mean.

To avoid pre-protocol negotiation fatigue (PPNF), I suggest you postpone protocol discussion until after TAM and discuss it solely with a JREF Representative.

I still stand at: "No test will happen."

I understand that, just pity that till TAM ends, time will be “wasted”, cause as I said before, I would like to have a test while I am here in US and I cant stay here for a year. And whatever we decide with the final protocol I will take probably 2 weeks to practice the “decided way” before the actual test. So if for now I would know at least more or less what can it be, I would not waste my time and would practice if it is pair of 2 then with pairs of 2 if it is with 3 then with 3 etc… Cause for example all the months that I am here I practiced in a different ways and each new idea and “way” I would try to make sure it is better… as I explained before I started with zenner cards, than switched to photos and tried different sets and different amounts to be drawn out and “guess”

 

[Startz]

Ok, I think I could try it, just need to redo sets of two in sets of 3, I just have to make sure that all three are different. In order not to get confused and again I would need to know what are the 3 “options” for my final decision same as with pairs… I would make my decision based on my vision and descriptions that I would do, while I hold the envelopes. I could try a few runs to train if it worth it. Can anyone calculate, how many you have to get right to beat 1.000 odds… If i wold have say, 10 sets by 3 per "pair" (that would make 30 photos in all) and have to identefy 1 from each set..

If you are identifying the photo from a group of three, then 16 out of 24 would do the trick. If you are comfortable with that, my guess is it would allay some of JREFs concerns.

Alternatively if you feel you could do 20 out of 30, then the odds work out that you could have three separate tries to hit the 20 out of 30 mark and win the prize if any of them succeed!

 

[pavel_do]

If you are identifying the photo from a group of three, then 16 out of 24 would do the trick. If you are comfortable with that, my guess is it would allay some of JREFs concerns.

Alternatively if you feel you could do 20 out of 30, then the odds work out that you could have three separate tries to hit the 20 out of 30 mark and win the prize if any of them succeed!

What you calculated sounds reasonable and I guess I can possible try and practice with it some runs and see what will happen.. But as I explained the way I see, (if you read my reply to Remiev) I would need to know what are the 3 options before my final decision with the answer of what is the 1 photo out of 3… I see no affection of the odds in example with 50/50 and I guess here also it does not make deferens if I know what are the options or not as it is still 1 out of 3. Or does it?

 

[Startz]

What you calculated sounds reasonable and I guess I can possible try and practice with it some runs and see what will happen.. But as I explained the way I see, (if you read my reply to Remiev) I would need to know what are the 3 options before my final decision with the answer of what is the 1 photo out of 3… I see no affection of the odds in example with 50/50 and I guess here also it does not make deferens if I know what are the options or not as it is still 1 out of 3. Or does it?

From a statistical point of view there is no problem with your being told what the three pictures are before you make a final decision, so long as there is a fresh set of three for each try.

Of course, JREF may have concerns that I wouldn't know about.

 

[pavel_do]

From a statistical point of view there is no problem with your being told what the three pictures are before you make a final decision, so long as there is a fresh set of three for each try.

Of course, JREF may have concerns that I wouldn't know about.

Well, every time there will be a new set of 3 photos. ( 1 photo actually from the set of 3) but still from the set of let say 10 by 3 or 15 or 20 if Ill manage to make 20 sets by 3 that would be all deferent… If not I guess set of 10 sets could be used 2 times that would make 20 pairs per run any way. But again, every time 1 photo from a new set of 3 will be described by me and than based on my descriptions from 3 possible options I will name 1 that would feet to my visions. You by replacing set every time, you mean, that the set of 3 that will be “guessed” first will have to return in all sets and then 1 of 10 will be chouse and one envelope from 3 possible will be drawn and offered to me for ‘prediction” ?

 

[Startz]

Well, every time there will be a new set of 3 photos. ( 1 photo actually from the set of 3) but still from the set of let say 10 by 3 or 15 or 20 if Ill manage to make 20 sets by 3 that would be all deferent… If not I guess set of 10 sets could be used 2 times that would make 20 pairs per run any way. But again, every time 1 photo from a new set of 3 will be described by me and than based on my descriptions from 3 possible options I will name 1 that would feet to my visions. You by replacing set every time, you mean, that the set of 3 that will be “guessed” first will have to return in all sets and then 1 of 10 will be chouse and one envelope from 3 possible will be drawn and offered to me for ‘prediction” ?

Statistically it doesn't matter how many different sets you provide, so long as each identification attempt begins with a fresh set. If it works for your talent, you could have 30 (or however many) identical sets of 3. You could identify the photo in set 1, then either have that photo opened or have it set aside. Then do set 2. So long as nothing gets re-used and all the sets are prepared in advance, it doesn't matter if the sets are all alike or all different. Go for whatever you find easy and whatever works best for you.

 

[pavel_do]

Statistically it doesn't matter how many different sets you provide, so long as each identification attempt begins with a fresh set. If it works for your talent, you could have 30 (or however many) identical sets of 3. You could identify the photo in set 1, then either have that photo opened or have it set aside. Then do set 2. So long as nothing gets re-used and all the sets are prepared in advance, it doesn't matter if the sets are all alike or all different. Go for whatever you find easy and whatever works best for you.

Sounds great, I would need it to be open straight after I named it.. with the rest I am fine, if the JREF would be also fine with that.. I will practise a bit with some sets of 3 later today and see what will I get...

 

[RemieV]

Pavel,

What if they were instead divided into sets of ten photos, wherein each set held 9 pyramids and one apple, and the only goal for each run was to find the apple? Of course, the envelopes couldn't be opened one at a time unless you identified the apple... For instance, if you held the first three and said each was a pyramid, we would not open them to verify, but wait for the apple to be identified.

When thinking this over, I'm also factoring in the amount of time it would take to complete such an experiment. This would require significantly fewer reps.

~Remie

 

[pavel_do]

Pavel,

What if they were instead divided into sets of ten photos, wherein each set held 9 pyramids and one apple, and the only goal for each run was to find the apple? Of course, the envelopes couldn't be opened one at a time unless you identified the apple... For instance, if you held the first three and said each was a pyramid, we would not open them to verify, but wait for the apple to be identified.

When thinking this over, I'm also factoring in the amount of time it would take to complete such an experiment. This would require significantly fewer reps.

~Remie

I can try this one to see what will I get… just need to print out 8 more of the same photos.. ( just that same pictures confusing as I explained to you before unless there will be some different photos but only one of It say APPLE or something elts).
How many runs would JREF be satisfy with how many do I need to get right to pass the preliminary test?? With 1 out of 10?

p.s. The test where there would be 1 out of 3 that was described above, does it sound at least a bit OK? Still the photo that I name would be opened straight a way and I would need to know what are the options before I name it as I would be able to match my description to one of it..

 

[Startz]

I can try this one to see what will I get… just need to print out 8 more of the same photos.. ( just that same pictures confusing as I explained to you before unless there will be some different photos but only one of It say APPLE or something elts).
How many runs would JREF be satisfy with how many do I need to get right to pass the preliminary test?? With 1 out of 10?

p.s. The test where there would be 1 out of 3 that was described above, does it sound at least a bit OK? Still the photo that I name would be opened straight a way and I would need to know what are the options before I name it as I would be able to match my description to one of it..

With 1 out of 10, getting 5 out of 7 should be good enough for the prelim. Or 6 out of 9 if you prefer.

 

[GzuzKryzt]

With 1 out of 10, getting 5 out of 7 should be good enough for the prelim. Or 6 out of 9 if you prefer.

I would prefer Seven of Nine.

 

[Crundy]

I would prefer Seven of Nine.

Wouldn't we all

 

[pavel_do]

Wouldn't we all

COMON… lets make it 9 of 9… as 6 of 9 it is easy.. everybody can do that..

 

[GzuzKryzt]

I would prefer Seven of Nine.

Everyone except Crundy please flip your humor switch: Seven of Nine

I expect a hefty suspension - eight to seventeen exa-eons - for this unprovoked derail.

 

[Crundy]

Everyone except Crundy please flip your humor switch: Seven of Nine

I expect a hefty suspension - eight to seventeen exa-eons - for this unprovoked derail.

Two questions:
1) Why just me, and
2) Why would you make a link to Seven of Nine which did not include a good image?

I'm off to the bathroom to "flip my humor switch".

 

[Pope130]

Crundy,
Here ya' go:

 

[Crundy]

Rockin'

 

[chillzero]

Guys, come on... whenever there's a discussion on statistics you get all sidetracked and derailed on this. Can we keep it on topic please?

Thanks.