Ok, we needed that clarification. I believe Steenkh was referring to unexposed & undeveloped photo paper, which would be white, as opposed to exposed & developed photo paper, which yes, would come out black. Will the process photos all be the same size? That would be necessary to insure that there is no way of picking up cues from the size or weight. I like the idea of using processed blank photos, as that makes them identical to the processed real photo. (no way to discern if it's real or blank by trace odors of the developing chemicals)
Information on Protocol for Pavel Ziborov Applicant
- Thread starter RemieV
- Start date
-
- Tags
- pavel ziborov
pavel_do
Critical Thinker
- Joined
- Aug 7, 2007
- Messages
- 363
All the photos I have printed are the same size 6x5 inches.
Since you cant really order plain photo to make it exposed to the light and become black.. What I done is.. made a “photo” that I filled with black color in photos shop
and it worked.. I have a few black photos.. at first at the lab they thought that it is image came bad ( as I ordered 5 of them) and there is something wrong with it.. So I had to tell them that the way I wanted them… I could read on there face like “WTF!?” 
guy orders black photos…
Last edited:
Jackalgirl
Graduate Poster
- Joined
- Aug 25, 2004
- Messages
- 1,801
I don't think there's any functional difference between a piece of photo paper that's been covered with black ink and a blank piece of photo paper.
Pavel, I highly recommend that you wrap each photo (black or actual photo) in opaque black paper (like construction paper), then put it into an opaque envelope. That ensures that there can be absolutely no information leakage (i.e., the black photo being more apparent through the envelope).
If 100% accuracy (correctly identifying which photo of each pair is the "black blank" and which one is the photo every time for ten pairs of photos) is unreasonable, just repeat the test with multiple sets. Let's say you claim to be able to correctly identify four pairs out of the ten pairs. That's a 1:16 chance of success. But if you do that ten-pair trial three times, your chances of identifying four in ten correctly (each time for three total trials) is 1:4096.
(if I'm not doing my math incorrectly.)
Incidentally, thanks everyone for confirming my probability calculation. I've written down the date! ; )
Pavel, I highly recommend that you wrap each photo (black or actual photo) in opaque black paper (like construction paper), then put it into an opaque envelope. That ensures that there can be absolutely no information leakage (i.e., the black photo being more apparent through the envelope).
If 100% accuracy (correctly identifying which photo of each pair is the "black blank" and which one is the photo every time for ten pairs of photos) is unreasonable, just repeat the test with multiple sets. Let's say you claim to be able to correctly identify four pairs out of the ten pairs. That's a 1:16 chance of success. But if you do that ten-pair trial three times, your chances of identifying four in ten correctly (each time for three total trials) is 1:4096.
(if I'm not doing my math incorrectly.)
Incidentally, thanks everyone for confirming my probability calculation. I've written down the date! ; )
pavel_do
Critical Thinker
- Joined
- Aug 7, 2007
- Messages
- 363
Thank you for suggestions. I am placing photos in double envelopes so there is nothing that you can see through as second envelope is manila and thicker than the first white one.. so even if u put it against the light u cant see what photo is inside.. unless u put it close to the bulb
that I don’t do obviously.. Beside that my eyes covered and see no point in leaving any possible clues to myself.. that would be cheating of myself at first place and I don’t have time to waist with such games..Regarding the odds.. I never claimed 100% success and being able to identify it with 100% accuracy… For that exact reason I wanted to have a test with 30 pairs in trial.. Where I need to get 24-25 correct.. still there is a room for mistake that we all can do even by random chance..
last time I tested my self my result was 23 out of 30. I know it is not 1.000 odds but if I am not mistaken significant enough that it is over chance odds…Any ways.. I have done some test today, and will do the rest tomorrow during a day, my friend will help me out. I will post my results and thoughts
and we`ll see Have a nice weekend everybody.Jackalgirl
Graduate Poster
- Joined
- Aug 25, 2004
- Messages
- 1,801
Well, if my math is correct (and I warn you, my math is always suspect), you'd only need to get 4 out of 10 correct three times. So that's a total of 12 right out of 30, but you'd be doing it in three steps, or trials, of 10 pairs each.
I don't know the probability for 23 pairs out of 30 (assuming you want to do a straight run of 30 pairs), but I'll bet it's okay for the test.
I think you're pretty close to a workable protocol. JREF is going to want to ensure that the envelope setup you've got is actually opaque, so make sure you describe very exactly how you're setting it up so that they can duplicate it and try it out on their end.
Can't wait to see how your self-test goes. Good job! : )
I don't know the probability for 23 pairs out of 30 (assuming you want to do a straight run of 30 pairs), but I'll bet it's okay for the test.
I think you're pretty close to a workable protocol. JREF is going to want to ensure that the envelope setup you've got is actually opaque, so make sure you describe very exactly how you're setting it up so that they can duplicate it and try it out on their end.
Can't wait to see how your self-test goes. Good job! : )
You must be careful here, because many modern photo labs actually use a printing process, and the old developing method is fast disappearing with the advent of digital pictures - if that is important for your abilities.
OK. If you prefer black pictures, it is fine with me. They are just not "blank". You once claimed that you did not "see" the picture itself, but some emotional content that was associated with the picture, and I would have thought that actual blanks would have zero emotional content, and you would then no more need to to distinguish between pictures with different emotional content, but merely between "no content" and "content".
It is your claim, and you must know what you can do.
Akhenaten
Heretic Pharaoh
All the photos I have printed are the same size 6x5 inches.
Since you cant really order plain photo to make it exposed to the light and become black.. What I done is.. made a “photo” that I filled with black color in photos shop
Hi Pavel!
You tricked me with this a while ago, when I thought you were joking about printing blank photos. Thank you for sorting that out
I can imagine the folks at the photo lab.
It's a shame that home-made prints don't work properly; this must be getting expensive for you.Things seem to be progressing well, and you continue to impress with your willingness to take advice from the people helping you. Good on ya mate.
Cheers,
Dave
Gr8wight
red-shirted crewman
- Joined
- Jul 5, 2004
- Messages
- 1,661
Except, Gmonster, if Pavel can get seven out of ten correct on three, consecutive, properly controlled trials, he probably satisfies the requirement. Can someone who understands stats confirm that for me?
pavel_do
Critical Thinker
- Joined
- Aug 7, 2007
- Messages
- 363
If that correct.. and JREF would agree with that.. ( that I doubt) I am ready for test tomorrow
as I never got less than 4 correct out of 10 pairs.. ( up till today at least)Any ways.. here is my results of the tests.
First the test by Ramiev suggested method where 10 and 10 of each photos was mixed up and I had to separate them in 2 pails.
Here is the results of 3 trials.
1. 9 correct and 11 incorrect.
2. 11 correct and 9 incorrect.
3. 11 correct and 9 incorrect.
Second test is where I had 3 trials by 10 pairs.. where out of each pair I would need to identify 1 photo correct. After each pair was named the envelopes was open and answer revealed as usually I use to do.
1. 8 correct and 2 incorrect.
2. 7 correct and 3 incorrect.
3. 6 correct and 4 incorrect.
Third test where it is all done the same as in second one, but the results was unknown till the last pair of 10 was identified.
1. 6 correct and 4 incorrect.
2. 5 correct and 5 incorrect.
3. 6 correct and 4 incorrect.
and the last one is I done 2 trials of 10 pairs each, where one envelope contained photo and the other one photo card that is just black “plane”. One trial was in the way where the results was revealed immediately and the second it was unknown till the end of the test.
1. 8 correct and 2 incorrect.
2. 6 correct and 4 incorrect.
SO based on my results as you can see, the best test for me so far would be still the one where..
I name one of the photo from the pair and it is revealed immediately.
The test when they all mixed up as Remiev proposed and when I don’t know the results…
I think what affects it is would explained the great example that was written by Remiev.. about rullet.. So when I am identified let say first 4 as let say Titanic and by not knowing either I am right or wrong.. I am starting doubting myself.. maybe I am wrong.. and maybe now should come up the Earth photo… etc.. that affect the results..
SO I am don’t think that test where all photos mixed up and I have to separate them would work for me. For the reason that I wrote.
When I chouse one of two photos and name it each time the chance for that is 50/50 and knowing the result immediately does not affect odds of the next pair.. But that make a difference for me and for my results. I would be happy to be tested with the test with pairs same as it was done by professors for Affidavit letters.. just with the difference that I would hold both photos before naming 1. Even to have 3 trials by 10 pairs would need only 60 photos that is not that much.. and it would not take allot of time either. And if odds would work out in some way as Jackalgirl proposed that would make things even more easy…
Any comments, suggestions, questions?
Regards,
Pavel
pavel_do
Critical Thinker
- Joined
- Aug 7, 2007
- Messages
- 363
Hi Pavel!
You tricked me with this a while ago, when I thought you were joking about printing blank photos. Thank you for sorting that out
I can imagine the folks at the photo lab.
Things seem to be progressing well, and you continue to impress with your willingness to take advice from the people helping you. Good on ya mate.
Cheers,
Dave
Hello Mate,
I am doing my best.. and while I have time.. trying to use it.. as when I leave US on the last week of August, then for a few weeks I don’t think I will have much time for it.. as I will need to sort out some things that “accumulated” for months that I have been a way from Europe. But I will do my best to get back to it as soon as I will be able to. And will be continuing with all this work at every opportunity and free time that I will have.
Cheers,
Pavel
Last edited:
Startz
Muse
If that correct.. and JREF would agree with that.. ( that I doubt) I am ready for test tomorrow
Any ways.. here is my results of the tests.
First the test by Ramiev suggested method where 10 and 10 of each photos was mixed up and I had to separate them in 2 pails.
Here is the results of 3 trials.
1. 9 correct and 11 incorrect.
2. 11 correct and 9 incorrect.
3. 11 correct and 9 incorrect.
Second test is where I had 3 trials by 10 pairs.. where out of each pair I would need to identify 1 photo correct. After each pair was named the envelopes was open and answer revealed as usually I use to do.
1. 8 correct and 2 incorrect.
2. 7 correct and 3 incorrect.
3. 6 correct and 4 incorrect.
Third test where it is all done the same as in second one, but the results was unknown till the last pair of 10 was identified.
1. 6 correct and 4 incorrect.
2. 5 correct and 5 incorrect.
3. 6 correct and 4 incorrect.
and the last one is I done 2 trials of 10 pairs each, where one envelope contained photo and the other one photo card that is just black “plane”. One trial was in the way where the results was revealed immediately and the second it was unknown till the end of the test.
1. 8 correct and 2 incorrect.
2. 6 correct and 4 incorrect.
SO based on my results as you can see, the best test for me so far would be still the one where..
I name one of the photo from the pair and it is revealed immediately.
The test when they all mixed up as Remiev proposed and when I don’t know the results…
I think what affects it is would explained the great example that was written by Remiev.. about rullet.. So when I am identified let say first 4 as let say Titanic and by not knowing either I am right or wrong.. I am starting doubting myself.. maybe I am wrong.. and maybe now should come up the Earth photo… etc.. that affect the results..
SO I am don’t think that test where all photos mixed up and I have to separate them would work for me. For the reason that I wrote.
When I chouse one of two photos and name it each time the chance for that is 50/50 and knowing the result immediately does not affect odds of the next pair.. But that make a difference for me and for my results. I would be happy to be tested with the test with pairs same as it was done by professors for Affidavit letters.. just with the difference that I would hold both photos before naming 1. Even to have 3 trials by 10 pairs would need only 60 photos that is not that much.. and it would not take allot of time either. And if odds would work out in some way as Jackalgirl proposed that would make things even more easy…
Any comments, suggestions, questions?
Regards,
Pavel
One main comment: Your honesty in reporting both trials that work and ones that don't work is most admirable!
Startz
Muse
Pavel:
The chance of getting 8 or more right out of 10 is about 5 percent. The probability of getting 8 or more right at least once with three chances is over 15 percent. So while your scores are quite a bit better than chance, they aren't likely to be acceptable to JREF.
I would imagine (and I obviously don't speak for JREF) that they might let you have three shots at getting a perfect score out of ten.
There is also the issue of whether JREF will agree to a test in which cards are revealed before the very end of the test.
Startz
Muse
if you like, suggest a protocol based on this. I'm happy to rework it for clarity. Then it can be sent in to Remie to see if there are any remaining barriers.
pavel_do
Critical Thinker
- Joined
- Aug 7, 2007
- Messages
- 363
There was one before, I will find it and review maybe change something and send you to rework it.
Startz
Muse
Great! Post it here so others can take a look too.
Wow! What a difference from another applicant's thread I've been following. Pavel is a gentleman, and he is genuinely committed to determining a test protocol. What a pleasant surprise!
Those of you who are statistically savant, is opening the envelopes before all are revealed getting into the Monty Hall issue at all? Would it make any difference if the test was run as, he completes 3 pairs, and is told how many he got right BUT not which ones he got right??
Just thinking, MK
Those of you who are statistically savant, is opening the envelopes before all are revealed getting into the Monty Hall issue at all? Would it make any difference if the test was run as, he completes 3 pairs, and is told how many he got right BUT not which ones he got right??
Just thinking, MK
Jackalgirl
Graduate Poster
- Joined
- Aug 25, 2004
- Messages
- 1,801
Miss Kitt --
I have no idea re: the Monty Hall issue as I just barely understand the edges of that, and I barely understand that it has edges, if that makes sense. But I see what you're saying and I think that it is highly, highly unlikely that the JREF will agree to the confirmation of any success until the entire run is complete. That is to say, none of the envelopes will be opened until they have /all/ been sorted into "contains black photo", "contains regular photo" piles.
I agree with you re: Pavel, too. I'm really impressed and pleased with how this is progressing, and with Pavel's obviously sincere efforts to come up with a good test. Way to go, Pavel!
I have no idea re: the Monty Hall issue as I just barely understand the edges of that, and I barely understand that it has edges, if that makes sense. But I see what you're saying and I think that it is highly, highly unlikely that the JREF will agree to the confirmation of any success until the entire run is complete. That is to say, none of the envelopes will be opened until they have /all/ been sorted into "contains black photo", "contains regular photo" piles.
I agree with you re: Pavel, too. I'm really impressed and pleased with how this is progressing, and with Pavel's obviously sincere efforts to come up with a good test. Way to go, Pavel!
ddt
Mafia Penguin
I'd be glad to comment, but I'm a bit at a loss now how the draft test protocol looks like. Which post(s) should I look at?
Do I understand it right that Pavel's basic claim is that, when presented with two sealed envelopes, one containing a completely black piece of paper, and one containing a photo, he can say which one is which?
Startz
Muse
This does not raise the Monte Hall issue as each set of pairs is separate from those that precede them.
From a statistical point of view, there is no reason not to tell Pavel what happens as he goes along. However, JREF may not go along (as JackaGirl and others have pointed out) for two reasons.
1) If the results are bad, the challenger could declare midway through that some condition of the test is being violated.
2) It may be generally bad procedure for undefined reasons.
However, JREF hasn't spoken about this issue in this challenge yet.
IXP
Graduate Poster
- Joined
- Oct 19, 2004
- Messages
- 1,395
No more than looking at the results of a string of coin flips would change the statistics. The Monty Hall paradox results from Monty having knowledge which is partially revealed when he opens one of the doors. Now if we allowed him to change his guess after opening an envelope, that would be a problem.
IXP