Newsgroups: sci.math From: boll@CS.ColoState.EDU (dave boll) Subject: 3 sided coin - experimental data Date: Tue, 24 Nov 1992 21:07:55 GMT Re: the 3 sided coin problem Screw the theory, we've got some experimental data here. I went down to the local hardware store and bought (or attempted to buy) a metal cylinder with height=radius. What I got was a metal cylinder with height = 5.5 units, and radius=4.5 units (as measured with a micrometer). So, it's a bit off. I gave it a total of 40 coin-type flips (I was going to do more, but the results are so lopsided it's not worth it), and got 33 'sides' and only 7 heads/tails. Heads and tails were more common if the flipped object lands in a small area; initially I was flipping it into an old cardboard pizza box and getting maybe 70% sides. The 33/40 data came about when I discarded the box and flipped over my kitchen floor (lineoleum (sp???)). Even tho my object was a bit tall for its radius, I now think the height must be a fair amount less than the radius in order to get the % of 'sides' down to 33% Maybe H=2/3R or H=1/2R or some such. Any comments (other than "Dave, get a life!") are welcome. -- ------------------------------------------------------------ Dave Boll boll@handel.cs.colostate.edu "The speed of time is 1 second per second" ------------------------------------------------------------ ============================================================================== From: dwboll@gr.hp.com (David Boll) Newsgroups: sci.math Subject: 3-sided coins, some experimental results Date: 17 Jul 1998 15:10:56 GMT Hi all: I've been playing with real 3-sided coins lately (cylinders thick enough that they land on heads, tails, and sides with equal probability). Check out http://www.frii.com/~dboll/coin3.htm for some interesting results. It turns out that Thickness=Radius isn't far from the truth, although there are a host of unanswered questions and concerns, some of which are discussed on the URL above. Enjoy! ps: If you want to generate some data, I'll send you my disks if you promise to send me results of at least 2000 trials -- Dave Boll http://www.frii.com ============================================================================== From: dwboll@gr.hp.com (David Boll) Newsgroups: sci.math,sci.physics Subject: 3-sided coin, update Date: 31 Jul 1998 17:12:37 GMT Hi all: Re: the 3-sided coin problem (A coin thick enough that it lands on heads, tails, and sides with equal probability), check out http://www.frii.com/~dboll/coin3.htm for a update on trying to find the critical thickness empirically. We now have the results of over 10,000 trials, thanks mostly to some dedicated flipping by Justin Smith. It looks like thickness=radius is close to the answer. I have machined some disks to experiment with, and I'll send them to you if you send them back along with the results of at least 2000 trials. --- Dave Boll ============================================================================== [For some theoretical considerations see also 99/3sided