• Quick note - the problem with Youtube videos not embedding on the forum appears to have been fixed, thanks to ZiprHead. If you do still see problems let me know.

challenge procedure questions

latent aaaack

Muse
latent aaaack
Joined
Aug 4, 2006
Messages
926
I'm sending in an application after I find a notary but I first want to know if my claim is considered to be testable. I would like to know if there are some types of claims that could be tested over a distance instead of in person. My claim only involves correctly guessing a two digit number before it is revealed but there are limitations on the claimed ability that restrict me from being able to take the test in person in an acceptable time frame.

This is the site I've been using for probabilities. Unless I'm really bad at statistics what I've observed of my claim passes the one out of a million requirement. http://www.automeasure.com/chance.html

I don't know to what extent test protocol can be established here instead of by mail but, if something like a reoccuring number from a lottery website were chosen as a test target, in your organization's view, could any secure means of communication of guesses between me and the organization be established, such as by e-mail or phone?

The reason I ask is that I don't claim to be able to correctly guess numbers on command, but I claim to know when a guess made by me can be accurate, which may not happen except for one or two times a week and I can't make those guesses more than a few hours before the numbers are revealed. My claim is that over time those guesses would beat random chance odds. My estimate is that I'd be accurate between 10%-50% of tries. If the lower threshold of 10% is tested and I make two guesses per week then it would take about a year to build up enough tries copmared with five weeks for a 50% threshold and I'd like to know if that is a claim that your organization can test. Thank you for your help.
 
The people on this forum are just members of the general public, like you. We do not speak for JREF.
 
What makes you think you have this ability?

IXP
 
I know it's a stupid question, but have you read everything at this link?

http://www.randi.org/research/index.html

Rule 5 of the application states: "In all cases, applicant will be required to perform the preliminary test either before an appointed representative, if distance and time dictate that need, or in a location where a member of the JREF staff can attend."

This seems to indicate that the tests does not have to be taken at JREF HQ, only before an appointed JREF representative or staff member.

The statement at the end of the application states: "... conduct proper double-blind tests of any ability you believe you can demonstrate, before attempting to undergo a testing for this prize." and "... do not choose to ignore the need for such a precaution."

This seems to be a strongly-worded plea that you should make certain that you are not merely deluded about a known and well-established scientific principle before you send in your application.

I am also under the impression that a series of negotiations may take place before undertaking the actual test / demonstration.

Finally, please understand that I am not a foundation member or representative, nor am I a member of the JREF staff. I have no authority at all in the tests or application process. I am only another poster, trying to help.

Please make absolutely certain that you have a complete and accurate application, that you fully understand the rules, and that you are fully prepared to demonstrate your abilities.

-Fnord of Dyscordia-
 
Latent,

It would seem quite possible to make a protocol for conducting the experiment over the internet (or any network).

One possibility would be to write two pieces of software: server and client.
The server, as run by a referee, generates the random number and informs the client (you) that "a number has been selected" and is currently residing in a variable. (The actual value of the random number would of course not be passed on).
You, using the client software would return your predicted number, and the server software compares the two, and returns "true" or "false".

Repeat X number of times. Easy.
 
Tetrahymenix said:
Latent,

It would seem quite possible to make a protocol for conducting the experiment over the internet (or any network).

One possibility would be to write two pieces of software: server and client.
The server, as run by a referee, generates the random number and informs the client (you) that "a number has been selected" and is currently residing in a variable. (The actual value of the random number would of course not be passed on).
You, using the client software would return your predicted number, and the server software compares the two, and returns "true" or "false".

Repeat X number of times. Easy.
Click to expand...

Where is the "paranormality"?

M.
 
latent aaaack said:
...
The reason I ask is that I don't claim to be able to correctly guess numbers on command, but I claim to know when a guess made by me can be accurate, which may not happen except for one or two times a week and I can't make those guesses more than a few hours before the numbers are revealed. My claim is that over time those guesses would beat random chance odds. My estimate is that I'd be accurate between 10%-50% of tries. If the lower threshold of 10% is tested and I make two guesses per week then it would take about a year to build up enough tries copmared with five weeks for a 50% threshold and I'd like to know if that is a claim that your organization can test. Thank you for your help.
Click to expand...

How often have you tried this before to come with the numbers you gave, latent aaaack?

What exactly were your results, at what occasions (lotto, sports, etc.) how did you document them and what controls did you use?
 
latent aaaack said:
...My claim only involves correctly guessing a two digit number before it is revealed but there are limitations on the claimed ability that restrict me from being able to take the test in person in an acceptable time frame.

...

I don't know to what extent test protocol can be established here instead of by mail but, if something like a reoccuring number from a lottery website were chosen as a test target, in your organization's view, could any secure means of communication of guesses between me and the organization be established, such as by e-mail or phone?
Click to expand...

As has been pointed out, we're not JREF employees, so we can only give you suggestions. I suggest that it needn't be this complicated. You could easily solve the problem this way:

Equipment needed:
2 (two) 10-sided dice.
1 person in addition to yourself: a number generator (NG)
Online communication capability.

Protocol:
NG will roll the two 10-siders to determine a number between 00 and 99. He/she will write this down and email/post/IM you: "go for attempt #1". When the answer comes to you, you write it down and email/post/IM back: "attempt #1 complete." NG will roll another number, email(etc) you, etc. Continue 10 times. When this is complete, the results will be forwarded (via registered snail-mail) to a trusted third party and opened on camera. The test is successful if you determine 8 out of 10 numbers.

Note that there are some serious problems with this protocol: for example, there isn't really any way to make sure that the NG doesn't communicate with you outside of the test protocol. Plus, you'll be wanting to ensure that the NG is actually writing down the number he/she rolled. There are probably other problems that other folks here will be pointing out shortly. But hopefully this will give you an idea of what the JREF will probably be looking for.

BTW, I have two 10-siders. I'd be happy to do this with you, here, if you like. I'll just post "go for attempt #x" and you respond with "attempt #x complete" and when we're completely done, we'll both PM our results to a trusted third party.

What do you say?

ETA: P.S. This would NOT constitute a successful demonstration of the Challenge test. It'd just be a test run of a proposed protocol that you and I would try out to work out the bugs so that when you apply for the Challenge, you have a protocol ready to go. -- J
 
I'm sitting by my computer with my dice: one is a 10-sider marked "10s" (e.g., 00, 10, 20, etc) and the other is simply marked 0-9. I'm totally ready to go. I figure it would be best if we each designated a trusted third party (that is, not the same person) and, when everything's done with, they'll post the results here. Any problems with this very basic experimental protocol?

Again, this isn't part of the JREF challenge at all since I'm not a JREF person and you haven't (to the best of my knowledge) applied. It's just a test-run to work out all of the possible objections and solutions, so that when you DO apply, you have a completely solid protocol to submit.
 
Jackalgirl said:
I'm sitting by my computer with my dice: one is a 10-sider marked "10s" (e.g., 00, 10, 20, etc) and the other is simply marked 0-9. I'm totally ready to go. I figure it would be best if we each designated a trusted third party (that is, not the same person) and, when everything's done with, they'll post the results here. Any problems with this very basic experimental protocol?

Again, this isn't part of the JREF challenge at all since I'm not a JREF person and you haven't (to the best of my knowledge) applied. It's just a test-run to work out all of the possible objections and solutions, so that when you DO apply, you have a completely solid protocol to submit.
Click to expand...

Thanks for giving that example of what a test might consist of, but my qualm with that protocol is primarily that it leaves everything in the hands of us without any way to verify that we're not tampering. I think it would be simpler and provide more proof if a number that can be found online were used.

I've been trying to figure out what the best protocol would be for getting past the following hurdles and I'm hoping someone can help me think of ideas. The requirements I want to satisfy are:

A means of leaving a message on any part of the web, who's date can be completely trusted to be accurate and not, at a later date by someone just coming accross it, be accused of having been hacked by me after the fact. Posting a message on a message board is one way a time stamped record can be left, along with sending an email and those are the only two I can think of (not including instant messages). In your (anyone's) view are either of these an absolutely fool proof method of having a definitely authentic time-stamp left? I know that there are modifications to vbulletin message boards that allow for time stamps to be changed.

The second requirement is that whatever info I'm guessing about has to be testable itself for not being personally connected to me. For instance, if it were a lottery site then the testers would, upon success, have to interview the lottery site and dig up everything they can about it to make sure I didn't pull a hoax off with them. Alternatively the lottery site could be chosen by someone other than me. If using a lottery site as a source of random numbers isn't acceptable I'd like to know what is.

Those are the only requirements I have and if those can be satisfied I'm ready to go, and I'd like to do it outside of the JREF challenge actually. I'd like to have a protocol that leaves conclusive proof that anyone can access over the net, if anyone can think of one and this will allow there to be minimal effort expended in the testing of the claim.
 
latent aaaack said:
Thanks for giving that example of what a test might consist of, but my qualm with that protocol is primarily that it leaves everything in the hands of us without any way to verify that we're not tampering. I think it would be simpler and provide more proof if a number that can be found online were used.
Click to expand...

You're right in that you'd have to be sure that the number generated online was actually random. I personally think that a lottery declaration would be fine -- I am by no means a statistician, but I've seen the use of lottery numbers proposed here for other things, so I've got to assume it'd be okay.

I'm not qualified to tell you about time-stamps, either; insofar as I know, almost anything digitial can be tampered with. Perhaps there are some encryption programs that can be trusted to provide worthy data -- hopefully, some crypto or Internet expert will weigh in here.

What might be simpler, though, is you just sending a series of letters via registered snail-mail. That'd be a reliable time-stamp right there. Your letters would include a specific statement such as, "the following is my prediction of the first two (last two, whatever) numbers in the DAY-MONTH-YEAR lottery for Anytown, USA". Then your letters could be opened on camera and compared to the results for the dates indicated. You could take as long as you like to generate the letters -- the key thing is that the postmark would predate the lottery day.

How's that sound?
 
The biggest hurdle is that I can't make the guesses more than a few hours before the numbers are revealed so that limits it to telecommunications. I'm confident there's a way to do it though, even if it just takes putting predictions on a message board where there are a lot of witnesses, so as to preclude the posssibility of me having hacked the time stamp to read incorrectly afterwards.
 
latent aaaack said:
The biggest hurdle is that I can't make the guesses more than a few hours before the numbers are revealed so that limits it to telecommunications. I'm confident there's a way to do it though, even if it just takes putting predictions on a message board where there are a lot of witnesses, so as to preclude the posssibility of me having hacked the time stamp to read incorrectly afterwards.
Click to expand...

That would certainly be good enough for me, just as a start. If you posted something here, any attempt to edit it afterwards should be noted by the software (each edited message is marked with its edit time). Why don't you go ahead and give it a try here?

(When you apply for the JREF challenge, this will be something you can work out directly with Jeff Wagg, who handles the applications. He should be able to go into more detail about what's possible and acceptable during the protocol negotiation.)

Does anyone else have any knowledge of good crypto software that would help fix an "unbreakable" timestamp to something?
 
Test post. Pay no attention to this test post (I am playing with the edit function).

ETA: Urp, I just figured out a way to make an edited message appear without its "edited" marker. So posting to this Forum isn't going to be the most secure way to do your test. My question to the crypto people stands.

Another option would be to use, say, UPS. They attach tracking numbers to stuff you mail through them and -- if I recall correctly -- their tracking system tells you both the date and time something happened to that item (i.e., the date and time it was checked into the system). So it would be easy to see if, say, you checked a letter into them at 15:23 and the lottery prediction was made at 17:00 (adjusting for time zone differences, if any). That might be easier than trying to find a cryptographic answer.
 
Last edited:
I had another thought -- I loathe IM program, so I don't know anything about the popular clients. So does anyone else know whether you can get an IM program to log the conversation with a time-stamp for each person's "send"? If so, that should work fine (three parties could log, and then compare each other's log to make sure no tampering took place). And it'd be cheaper than UPS or FedEX.
 
Jackalgirl said:
I had another thought -- I loathe IM program, so I don't know anything about the popular clients. So does anyone else know whether you can get an IM program to log the conversation with a time-stamp for each person's "send"? If so, that should work fine (three parties could log, and then compare each other's log to make sure no tampering took place). And it'd be cheaper than UPS or FedEX.
Click to expand...

I like that idea but I still don't know if an IM time stamp or an e-mail is at all vulnerable to hacking so I await others' input on that. Someone could set up an IM account just for the purposes of testing and then the password can be given to anyone who wants to see it for themself later. My guess is that e-mail time stamps might be about as secure as a postage stamp's.
 
latent aaaack said:
I like that idea but I still don't know if an IM time stamp or an e-mail is at all vulnerable to hacking so I await others' input on that. Someone could set up an IM account just for the purposes of testing and then the password can be given to anyone who wants to see it for themself later. My guess is that e-mail time stamps might be about as secure as a postage stamp's.
Click to expand...

Well, anything on an email can be hacked. However, I think that if the receiving person's email program puts a timestamp on it (like Microsoft Outlook), that's independent of the message header (I think). As long as the receiver /received/ the message before the results were determined, you should be okay. You'd have to restart the test if Internet delays cause a problem in the reception time. Does anyone else think that might work?
 
Jackalgirl said:
Test post. Pay no attention to this test post (I am playing with the edit function).

ETA: Urp, I just figured out a way to make an edited message appear without its "edited" marker. So posting to this Forum isn't going to be the most secure way to do your test. My question to the crypto people stands.

<snip>.
Click to expand...
I do not think you know how to edit a post without its 'edited' marker. You edited this post 6 minutes after you made the post.
 
You must log in or register to reply here.

Back
Top Bottom