The 10000 sided die is about the size of a tennis ball and looks round. It is round. Numbers are very small and read by a laser device which is pushed into position over the dice after it stops rolling. The (self-levelling) laser emitting device shines straight down, and detects only reflection directly back upwards - numbers not really "on top" reflect slightly off to one side.
-- bungston, Apr 24 2005

(?) No myriagons... http://jcrystal.com...POLYHEDRA/p_00.html
but still a link cool enough to keep in favorites. [2 fries shy of a happy meal, Apr 24 2005]

(?) Mathworld: Polyhedron http://mathworld.wo...com/Polyhedron.html
Neat pictures [Detly, Apr 24 2005]

(?) Johnson Solids http://web.ukonline...ohnson/johnson.html
Tabulates the Johnson solids [Detly, Apr 24 2005]

Wikipedia entry on pseudorandomness http://en.wikipedia...seudorandom_numbers
a much easier way [apocalyps956, Jun 27 2006]

Site that uses atmospheric noise to generate random numbers http://www.random.org/
"randomer" than pseudorandom [apocalyps956, Jun 27 2006]

tennis quantum physics?
-- benfrost, Apr 24 2005


There isn't one chance in 10,000 that anyone would ever use this! Bun.
-- Ichthus, Apr 24 2005


How about rolling the hyperdie (two dice, one die, by the way. Although oddly the same rule does not apply to mice) on a painstakingly-levelled glass flat-bottomed bowl filled with an optically-dense black liquid? Then, if the liquid is opaque enough, you'll be able to look up from under the bowl and see only the single micro-face which is resting on the glass; the others will be obscured by the intervening liquid.

This would work on the same principle as those "prediction balls" where an icosohedral (?) die floats up to the top of a glass jar full of black liquid.

Also, is it possible to make a convex object with exactly 10,000 identical faces? The rules of geometry might restrict you to, say, 12,436 or 9,084.
-- Basepair, Apr 24 2005


Darn! I rolled a three.
-- ato_de, Apr 24 2005


An easier way to get to 10,000 would be to roll 4 ten-sided dice. Or flip 13 and a half coins.
-- hippo, Apr 24 2005


Unless you have a supercomputer nearby to work out a random number and transmit it to the bluetooth transponder during the time the dice is rolling. This number can then be transmitted as the result of the dice roll.
-- hippo, Apr 24 2005


I like this idea but how about an enhancement?

Make a hollow sphere out of crystal glass and engrave numbers on the inside. The sphere contains water and a floating, but snugly fitting, second sphere. The second sphere is free to move but contains a laser pointer. The weight distribution in the second sphere causes the laser always to point upwards.

Roll the die and wait. The number will be projected onto the ceiling.

I should point out that each number is cunningly engraved with a suitable lens to cause a magnified image to be formed.
-- DenholmRicshaw, Apr 24 2005


Hmm. Eight hundred and three. Hope I don't end up on a snake.

[hippo] - shouldn't that be five ten-sided dice?
-- moomintroll, Apr 24 2005


To answer [Basepair]'s question:

From Mathworld ("Polyhedron" (linked), and others from that):

There are a total of nine regular polyhedra: five Platonic solids, with a maximum of 20 faces for the icosahedron and four Kepler-Poinsot solids, with a maximum - sort of - of 53 faces for the great icosahedron. You might find it hard to roll that one, though.

Getting into the semiregular polyhedra - or Archimedean solids - of which there are 13, the snub dodecahedron has the most faces with 92.

Johnson solids, which have regular faces but are not regular polyhedra, max out at 62 faces (J71 to J75, my favourite of which is J73, the parabigyrate rhombicosidodecahedron).

Of course, you could also consider that the numbers could be placed at the vertices or the midpoints of the edges.
-- Detly, Apr 24 2005


I dunno. Counting the dots would get really tedious...
-- DrCurry, Apr 24 2005


moomintrol is right, except the first die should just be a coin with 0 or 1. Otherwise you'd have 99,999 combinations.
-- Size_Mick, Apr 24 2005


A D100, is two tensided dice with the one going from 1 to 00 and the other has 1 to 0. Roll 00/1=one. Roll 00/0=one hundred. A similar trick could be used with the five dice.
-- zeno, Apr 24 2005


And I know I just couldn't resist counting how many times I have to roll this super die before I get 10,000. And then I would have to break that record, and then, etc.+
-- zeno, Apr 24 2005


I like the "magic 8 ball" adaptations by [DenholmRicshaw] and [Basepair]. [Hippo], there will be no more of your futuristic supercomputer random number generating flimflam. These are dice we are talking about.

[Basepair] I too suspected that there might only be certain numbers which lend themselves to even distribution over the surface of a sphere. This is the sort of thing those math people figure out for a living.
-- bungston, Apr 24 2005


Surely to all intents and purposes this would be round. meaning that it would take a long amount of time to stop if you were to give it anything more than a gentle tap.
-- hidden truths, Apr 24 2005


I wondered how long it would take to align the reading device so that it got a read.
[hippo] Did you mention a supercomputer because cheap calculators' random number generators are not truely random? No wonder I haven't yet won the lotto. I'm still looking for someone to invest around $8 million USD on a 50% chance of winning scheme. I think I have charts & graphs on my old computer.
-- Zimmy, Apr 24 2005


I think aligning the device would be a problem. Also chasing the thing all over the table top. Perhaps it should have a special saucer with a tiny level sweet spot in the very center. Eventually it would settle there. The reader would be ready.
-- bungston, Apr 25 2005


A few modifications would make the die more interesting AND more readable. 100-sided die aren't difficult to read and they are maybe 1.5 inches diameter. I believe a 6 inch diameter die would provide the right balance between readability and size.

Some can see the number with the naked eye, some need a magnifying glass.

But, to reduce the amount of numbers you have to put on that tiny little face, a color code for the final digit can be used. I suggest a modified ROY G. BIV type thing.

Dark, blood red = 0 Bright red = 1 Reddish Orange = 2 Yellowish Orange = 3 Yellow = 4 Yellowish Green = 5 Green = 6 Bluish Green = 7 Blue = 8 Violet = 9

the side would be numbered 0-9999, but each red,blue, etc pane would only be numbered 0-999. Thus, you would only need to fit 3 digits into that tiny space, and the color would be an indicator of your general luck.

Plus it would look so pretty.
-- Madai, Apr 25 2005


Not as cute, but more practical (slightly):

A hollow crystal sphere (but with thick walls to avoid breakage), with one side flattened, mathematically calculated to stop the sphere from rolling after so many revolutions (on the average) after a toss, so that the thing tends to stay on the table. When the electronic innards detect that the thing has come to a stop, it generates the number pseudorandomly between 1 and 10,000 and displays it with one of those cool wand display animators they use now on some alarm-clocks/message centers, so that the number animates in blue digits, rotating around so that everyone at the table can see it. Make it as small as the electronic innards can cheaply fabricate (and so that the number generated is legible without squinting). Power to run the thing is generated by motion--when you shake it up in your hand, you run a little counterweighted device of the sort they used to use in watches to charge up the microcell a little.
-- Soterios, Apr 25 2005


(to hippo) //thirteen and a half coins// ?
-- apocalyps956, Jun 27 2006


I think It'd be a lot easier to just use pseudorandom numbers (see link)
-- apocalyps956, Jun 27 2006


Or atmospheric noise (see other link).
-- apocalyps956, Jun 27 2006


I made one of these when the idea was first posted, and I'm still waiting for it to stop rolling.
-- phundug, Jun 27 2006



random, halfbakery