First it means that the participant will inevitable "guess" about the hidden number of "burned in" within the 10 cables.
If he knows in advance about the number of "burned in" this factor will be taken out.
Yes the statistics varies, but it is a well known problem and follows the hypergeometric distribution.
Unfortunately i was wrong about the number of correct answers needed; it has to be 5 if 5 "burned in" were in the lot.
Wishes
Still not my understanding. It's not knowing that one of three doors has a prize, it's the host's act of revealing a door which doesn't that skews the remaining odds. Apologies for the Wikipedia reference. As long as the subject isn't informed during the trials of any cable's true identity it shouldn't be an issue.
I would contend that the setting be a "living room" type room, with the appropriate paraphenalia(?) be scattered about(Standard sound absorptive materials)..the system set-up be created by a studio engineer & corrected for room acoustics. The "tester" be confined to a "standard" sofa.be restricted to a 'zone' nearby said sofa.
The entire "system" be hidden from view by a floor to ceiling acoustically transparent curtain, stretching wall to wall.
While we are debating numbers of samples and statistics we also must come up with music samples.
This is going to be very hard as our resident "golden ear" will fight accepted scientific method tooth and nail...all the way.
________________________________________________________Rick....
The entire "system" be hidden from view by a floor to ceiling acoustically transparent curtain, stretching wall to wall.
While we are debating numbers of samples and statistics we also must come up with music samples.
This is going to be very hard as our resident "golden ear" will fight accepted scientific method tooth and nail...all the way.
________________________________________________________Rick....
Please read the thread! That is totally irrelevant and counterproductive.
Why? The guy who claims he can hear break-in (well, one of the guys, the other ran away) says he hears it on his system, in his room, with his music. Throwing in a whole lot of variables and complications doesn't yield more information.
If you have many variables and you do not know if they are inter-related and have a limited time/budget for testing, may I recommend you investigate the Taguchi method for experimental design.
Of course, the problem is still to identify a measurable difference anywhere, but that has been covered in 959 posts about phantoms. Myself, I will continue to keep my cables short and connections tight, so all this is moot.
Of course, the problem is still to identify a measurable difference anywhere, but that has been covered in 959 posts about phantoms. Myself, I will continue to keep my cables short and connections tight, so all this is moot.
'Knowledge" isn't the same as knowledge. Monty, when he's not drinking, is always 100% correct in his choice of doors. There is no element of chance in his sample elimination. It's my understanding of the core of the paradox, by always correctly eliminating a variable he provides information that unambiguously alters the odds.
Your cable subject doesn't have that luxury unless the proposal is to reveal cables before the process completes. Still sounds like an unnecessary hurdle.
Last edited:
I did a number of 10 coin flips at the random site SY linked too. Most of them were 5/5 or 4/6. In fact enough were that I just don 't think it matters.
Maybe we can agree that the random sequence should have at least 3 of one type to cover objections like Mr. Curl's that they may all be of the same flavor. Kind of a compromise between just taking the first random sequence that comes along and dividing it strictly 50/50.
Maybe we can agree that the random sequence should have at least 3 of one type to cover objections like Mr. Curl's that they may all be of the same flavor. Kind of a compromise between just taking the first random sequence that comes along and dividing it strictly 50/50.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.