I love nearly all iconoclasm. A coin flip as indicator of true randomness is definitely a cultural icon (I know I use it as an shorthand in teaching about probability and related topics). This study does debunk that notion and remind us that it really is a simple physical process with deterministic outcomes if sufficient controls are put in place. It is only because of the lack of controls that the coin flip is random. The randomness is not a property of the coin, but of the flipping and catching (landing) mechanism.
I'd say people have a sense of this. Have you ever flipped a coin poorly so that the coin only turned over once or a very few times? Doesn't feel like a fair flip, does it? You gotta make it flip so much that when it is caught, or hits the floor, it is unlikely that the flipper could have controlled the outcome. So, maybe we shouldn't say 'coin flips' are 50/50 propositions but that 'a person flipping a coin' is a 50/50 proposition (assuming the person isn't a magician). Wow... I can't believe I just spent that many words on this topic. <grin> Paul C. Bernhardt Department of Psychology Frostburg State University Frostburg, Maryland -----Original Message----- From: Michael Smith [mailto:[email protected]] Sent: Tue 8/4/2009 3:21 PM To: Teaching in the Psychological Sciences (TIPS) Subject: Re: [tips] Flipping Out | The Big Money Yes. The longer any particular pattern continues the more improbable it will be (if it is supposed to be a random process) and one would then suspect a non-random influence. I was just laughing at the academics spending the time and money to construct such a machine. I would surmise that this is really academic tom-foolery (in cahoots with the university's mechanical department). Or at least it should be considered as such. I find the result entirely unsurprising (and uninteresting from a rational point of view) and wouldn't hail it as any kind of discovery (as the writer of the article alludes). I say this, because if all conditions are controlled then there IS no more random influences and one should not of course expect a random result. If there was, then I would really be surprised--totally shocked even! It would be similar to the kind of situation where I was baking a vanilla cake and when I took it out of the oven, behold it's a chocolate cake! Their coin-flipping machine is no different in principle from me repeatedly placing a coin heads-up on the table and saying "Oh look, the result isn't random!", "Oh look, the result isn't random!", "Oh look, the result isn't random!".... If I were actually surprised at my results, then I think I would need to check into a sanitorium (but of course, if I actually were surprised, I would be too far gone to realize....I digress) And, my method is a lot cheaper! Besides, I know I can influence the flip of a coin the way I want if I know what position I start from, and if I don't "flip it too much". ---One of your comments: "I think somebody, somewhere found something similar when carrying out Pi to an infinite number of digits but I can't remember which journal I read it in." -- Given that that's impossible, I would really like to meet that person! -- just joking around :) --Mike On Tue, Aug 4, 2009 at 6:50 AM, Mike Palij <[email protected]> wrote: > On Mon, 03 Aug 2009 20:30:27 -0700, Michael Smith wrote: > >"The machine could make the toss come out heads every time" > > > >Maybe it was just an extended pattern that we all know sometimes > >occurs in random sequences! lol > > Indeed but couldn't we calculate the probability of such an extended > run under the assumption that the process producing it has p=q=.50? > As the run becomes more improbably, alternative hypotheses would > have to be seriously considered as replacements for the original > assumptions. > > Then again, if such a coin tossing experiment was continued for > an infinite number of trials under constant conditions perhaps something > peculiar, something wonderful could happen, like a new pattern of > outcomes emerge under the constant conditions which, when interpreted, > turns out to be the plans for how construct a machine that would > transcend time and space, an intergalactic subway system, if you will. > > I think somebody, somewhere found something similar when carrying out > Pi to an infinite number of digits but I can't remember which journal I > read it > in. > > Send me something. > > ;-) > > -Mike Palij > New York University > [email protected] > > > --- > To make changes to your subscription contact: > > Bill Southerly ([email protected]) > --- To make changes to your subscription contact: Bill Southerly ([email protected]) --- To make changes to your subscription contact: Bill Southerly ([email protected])
<<winmail.dat>>
