Netbook-Run Dice Robot Can Rack Up 1.3 Million Rolls a Day 280
stevel writes "The owner of games site GamesByEmail.com created Dice-O-Matic, 'a machine that can belch a continuous river of dice down a spiraling ramp, then elevate, photograph, process and upload almost a million and a half rolls to the server a day. ... The Dice-O-Matic is 7 feet tall, 18 inches wide and 18 inches deep. It has an aluminum frame covered with Plexiglas panels. A 6x4 inch square Plexiglas tube runs vertically up the middle almost the entire height. Inside this tube a bucket elevator carries dice from a hopper at the bottom, past a camera, and tosses them onto a ramp at the top. The ramp spirals down between the tube and the outer walls. The camera and synchronizing disk are near the top, the computer, relay board, elevator motor and power supplies are at the bottom.' While not called out in the article, the pictures clearly show a Dell Mini 9 running the show (and performing the optical recognition of the dice values.) No, it's not running Linux."
More Like Color Recognition (Score:5, Informative)
While not called out in the article, the pictures clearly show a Dell Mini 9 running the show (and performing the optical recognition of the dice values.)
Yes but there's not a lot of "optical recognition" going on. From the article:
The dice are "Michigan Red Eyes", which have different colored pips for each value. The different colors make it pretty easy to count rolls. For example, if 6 yellow dots are found in the image, there were three 2s rolled, no need to worry about determining the proper grouping or orientation of pips.
If you control the background as being black or shades of grey (which is what it appears on that dirty dirty Windows XP screen) then your task is a lot easier and less error prone. Well done on the designer's part but surely reduces the computational work load.
Not running Linux? (Score:1, Informative)
No, it's not running Linux.
That's because a Linux user would do it properly. "Windows XP, I see your Dice-O-Matic machine and raise you a /dev/random powered by a noise diode."
Re:Why? (Score:0, Informative)
FTFA: "Currently, GamesByEmail.com uses some 80,000+ dice rolls for play in games like Backgammon, Gambit (a RISK clone), W.W.II (an Axis & Allies clone) and others. To generate the dice rolls, I have used Math.random, Random.org and other sources, but have always received numerous complaints that the dice are not random enough. Some players have put more effort into statistical analysis of the rolls than they put into their doctoral dissertation...There is no doubt that I will still receive complaints about the rolls, but now I can honestly say I have done all that I can possibly do: the rolls you get are exactly as random as those you would get throwing by hand."
In other words, the software RNGs used may have been sufficiently random, but players don't believe that they were truly random. Who is right? Who knows. A lot of RNGs are not very good. But it is just as likely that the people complaining are seeing non-existant patterns in randomness (just like stock traders worldwide).
Throwing dice eliminates the question of whether the RNG algorithm used is truly random.
From TFA (Score:2, Informative)
Re:Why? (Score:1, Informative)
Well, who (apart from Risk players) needs a source of truly random numbers? Who would be interested in measuring randomness?
At one time in history, top cryptographers used mechanical systems like this. But they were abandoned in favour of noise diodes and radioactive decay, because those were not affected by mechanical imperfections in the machine and the dice, which turned out to be quite measurable (Shannon entropy). Using a poor source of random numbers is a disaster in cryptography terms, so the dice machines were replaced with something smaller, cheaper and better.
You can find this in many books about cryptography, but it also turns up in the Neal Stephenson book "Cryptonomicon", where the scientists actually calculate where the dice imperfections are found inside the enemy's "random" number generator based on analysis of ciphertext.
What we see here is an example of people being stupid. The demand for a dice machine is based on a misunderstanding of randomness. If it stops the users of the service crying out for "more random" numbers, then that's good, but at the same time those users need beating with a clue stick.
Re:Why? (Score:2, Informative)
The developer says that his users have complained that his software RNGs were not random enough. His aim in building this is to build a machine that is as random as if users were throwing their own dice. At the end, he promises (light-heartedly, I presume) to punish the dice if a user shows that they are not random.
All your dice suck - Testable! (Score:2, Informative)
Re:Why? (Score:1, Informative)
Agreed by whom? Those ignoring 90+ years worth of work on Quantum Mechanics?
You wouldn't need to know the entire state of the universe for such macro effects. There are limits to how quickly a change can be effected by a particle of a certain size.
Not being able to know the state has nothing to do with size - it's an inherent property of Quantum Mechanics (see Heisenberg Uncertainty Principle [wikipedia.org])
None of this has to do with "atomic landscapes". The main actors will be the material the die is made from, the shape of the die, the material of the surface, the shape of the surface, gravity, the angle at which the die falls, the position of the die when it is released. Other factors such as temperature & air pressure can be assumed to be normal (room tempature & air pressure, no significant air movement, etc). The other forces are simply too small to play any role (even combined). Their orders of magnitude weaker.
Re:Why? (Score:3, Informative)
http://en.wikipedia.org/wiki/Hardware_random_number_generator [wikipedia.org]
Re:What a waste, (Score:2, Informative)
Re:Why? (Score:3, Informative)
It depends on the dice being not biased, and the mechanics not exerting any influence.
One should test it by letting the machine rip for a couple of days, and then analyse the produced numbers for randomness, e.g. with the diehard [wikipedia.org] test suite.
Re:What a waste, (Score:1, Informative)
No problem. You can generate any die roll you like from D6's, just do a little math.
For a D8, just roll two D6's, add them together, and then take the result modulo 8 and add 1. Poof! A random number between 1 and 8!
Math Fail!
Obviously you don't play craps. When you throw 2D6, there is a 6 in 36 chance of comming up seven, not the 1 in 12 chance you seem to assume. Try this thought experiement: What are all the combinations for each potential number between 2 and 12 when you throw 2D6? I'll give you a hint, there is only one way each to get 2 or 12. Now, if you take mod 8 and add 1 of the result, you have heavily weighted the middle numbers
Re:Need more stats (Score:3, Informative)
It has 200 dice. Clearly it's made for Shadowrun. :)
[John]
Re:Why? (Score:3, Informative)
Shannon entropy isn't the same thing as randomness. It's a measure of balance of the distribution. You could have a high entropy generator that is very unrandom, say, one that basically alternates between outputs. Or you could have a truly random distribution that favors some outputs over others, but completely unpredictably.
Re:What a waste, (Score:5, Informative)
Wow. Fishing for nerds is so damned easy, they take the bait even when you put a sign on it saying "WARNING: CONTAINS FISHHOOKS".
For a good nerd time, try working out the probability distribution table for Mod8(D6+D6)+1. I suggested it as a joke, but it's less horrible thank you might think.
Re:What's so clever? (Score:5, Informative)
Why not use traditional white dice with black dots on a white background. Then it is as simple as
Calculating the full image is far more work than necessary. This also easily allows for an arbitrary number of dice to be used, assuming the don't start piling on top of each other.
Because if you count 36 dots, was that six dice with six rolled on each, or twelve threes, or ...
Even if you can count exactly how many dice you rolled, was the total of six dots from four dice three ones and a three, or was it one plus two plus two plus one?
I'm guessing that knowing the exact number of dice rolled and what each die rolled is important.
Re:More Like Color Recognition (Score:4, Informative)
Re:A good first step (Score:1, Informative)
Your number seems to say that 45% of players will double their money, and 55% will half it. Assuming eventual random distribution of affluence of players, that would leave Auto-Vegas with a deficit:
Vegas.cash = (.5 * 55% of gambled cash) - (1 * 45%) = 27.5% - 45% = -17.5%
So, for every period of gambling, Auto-Vegas will lose 17.5% of the total amount gambled during that period.