Slashdot Mirror


Netbook-Run Dice Robot Can Rack Up 1.3 Million Rolls a Day

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."

280 comments

  1. More Like Color Recognition by eldavojohn · · 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.

    --
    My work here is dung.
    1. Re:More Like Color Recognition by SQLGuru · · Score: 5, Insightful

      In the "good old days," we called that a clever hack. Solving a problem is about simplifying the problem space in any and every way possible. I've made similar "OCR" hacks when everything was going to be in a known font and size.

    2. Re:More Like Color Recognition by HEbGb · · Score: 1

      So what's your problem with this? It sounds like you think he didn't make the problem hard enough for your taste.

    3. Re:More Like Color Recognition by badboy_tw2002 · · Score: 1

      How is that not still optical recognition?

      - Well done on the designer's part but surely reduces the computational work load.

      You say it like that's a _bad_ thing!

    4. Re:More Like Color Recognition by ivan256 · · Score: 1

      Welcome to the world of computer vision.

      If you're expecting magical software with human like intelligence (many people do...), I suggest you don't look behind the curtain.

    5. Re:More Like Color Recognition by Patch86 · · Score: 2, Insightful

      Why bother with numerical "dots" at all?

      If it can do colour recognition (obviously it can) why not just have a single coloured circle on each side of the dice. When it's a blue circle, it's a "1". A yellow circle, it's a "2". Two yellows and three blues, 7 total.

      Surely the need for a symbolic representation of the number is only necessary for us feeble humans, with our tendency to forget abstractions. For a computer, which need never forget that green means "6", actually drawing a picture or making a pattern seems pointless.

    6. Re:More Like Color Recognition by cecille · · Score: 3, Informative

      That's actually a clever idea. The only real problem I can see with it is that real dice are correctly balanced so that every side has an equal chance of coming up (or they are supposed to be - real ones are). Well balanced generic cubes without the dots or with single colour dots are probably pretty hard to find, and probably more expensive given the relative cheapness of generic dice.

      --
      ...no two people are not on fire.
    7. Re:More Like Color Recognition by SQLGuru · · Score: 1

      I agree that a die that has solid colored sides would be easier to read, however, the dice with colored pips are off-the-shelf parts. Even though it adds a little to the complexity (complexity which has already been solved, might I add), reduced on-going maintenance cost is probably worth it in this case. At some point, one or more of the 200 dice will wear (crack, chip, have a worn edge, etc.) and being able to go to Joe's Gaming Shop and buying replacement dice instead of ordering a custom set from Bob's Custom Die Works will be easier and cheaper.

    8. Re:More Like Color Recognition by Patch86 · · Score: 1

      I've never seen, in gaming shops, dice with different coloured pips on each side. They've always been a uniform colour on all sides (white, red, black, whatever).

      I have, however, seen dice with a solid colour side before. They're used in some board games, and even one RP game I once played (couldn't tell you what it was though, was only the once).

    9. Re:More Like Color Recognition by extremescholar · · Score: 0

      A little bit of paint on the indentations would be a trivial customizing activiyt.

      --
      Using the Freedom of Speech while I still have it.
    10. Re:More Like Color Recognition by Patch86 · · Score: 1

      GP said that the benefit of pipped dice is that they are off the shelf. Colouring in the dots of 200 dice seems fairly non-trivial. That's 4200 dots you'd be painting, you know.

      If you're going to customise your equipment, there are better solutions.

    11. Re:More Like Color Recognition by SQLGuru · · Score: 1

      I actually read the article. He refers to http://dicegamers.com/catalog/44/michigan_red_eye these (but later learns they are knock offs). Either way, they are available for purchase as I had indicated. If you can find some off the shelf that are solid colors, sure it would a good way to go. But again, he's already solved the problem with the colored pips, there is no benefit (to him) to pursue a different type of die.

  2. Excellent... by ben0207 · · Score: 5, Funny

    Finally a sensible way to play a 3000 pt Imperial Guard list!

    --
    cmd-q.co.uk - some sort of stupid fucking internet bullshit
    1. Re:Excellent... by Anonymous Coward · · Score: 0

      I'll take 2

    2. Re:Excellent... by corbettw · · Score: 1

      Or just use one Baneblade and *bam!* you have half the number of figs you did before.

      --
      God invented whiskey so the Irish would not rule the world.
    3. Re:Excellent... by N1AK · · Score: 1

      Finally a sensible way to play a 3000 pt Imperial Guard list!

      I keep threatening to use models knocked over by dice to decide casualties, but as I now regularly have to roll around 150 it'd be a little over-powered (yes really a single unit shooting once can lead to 150 to-hit rolls...)

    4. Re:Excellent... by 800DeadCCs · · Score: 1

      gods, what is it?

    5. Re:Excellent... by ben0207 · · Score: 1

      There are two approaches to Imperial Strategy:

      Half a million men, who all die terrible, terrible deaths.
      About 3 very large tanks, who all cause terrible, terrible deaths.

      You could say I side more with the traditional mindset.

      --
      cmd-q.co.uk - some sort of stupid fucking internet bullshit
    6. Re:Excellent... by Anonymous Coward · · Score: 0

      Even with this thing if my orks assault you I STILL won't have enough dice being rolled ;)

  3. Need more stats by Junior+J.+Junior+III · · Score: 3, Funny

    What's its AC and THAC0? :)

    --
    You see? You see? Your stupid minds! Stupid! Stupid!
    1. Re:Need more stats by Anonymous Coward · · Score: 3, Funny

      This is the 21st century sir.

      Your roll of Wis vs Fort is a failure. You suffer 50pts of Necrotic damage; save ends.

    2. Re:Need more stats by Anonymous Coward · · Score: 2, Funny

      It's a gazebo Eric! A gazebo!

    3. Re:Need more stats by the+unbeliever · · Score: 3, Funny

      THAC0 is deprecated as of the latest release. Please upgrade your packages.

    4. Re:Need more stats by Anonymous Coward · · Score: 0

      Screw THAC0, BAB is where it's at.

    5. Re:Need more stats by Anonymous Coward · · Score: 0

      Well hello there Junior J. Junior III, welcome back to the land of the awake! I see you've been comatose for the last decade... So it pleases me to to be the first to inform you that "To hit armor class 0" has been replaced with "difficulty class."

    6. Re:Need more stats by Jurily · · Score: 1

      Please upgrade your packages.

      Must be the side effects of Gentoo, but I parse that as "ICEBERG AHEAD!".

    7. Re:Need more stats by bughunter · · Score: 2, Interesting

      Sorry - those are all six-sided dice. Clearly, this machine plays GURPS.

      Or maybe Traveller.

      (We need a percentile version of one of these bots for our Rolemaster/Spacemaster games.)

      --
      I can see the fnords!
    8. Re:Need more stats by Bigbutt · · Score: 2, Informative

      It has 200 dice. Clearly it's made for Shadowrun. :)

      [John]

      --
      Shit better not happen!
    9. Re:Need more stats by zeropointburn · · Score: 1

      [bittervitriol]Burning iceburgs made of fail. (for this 'package' GP references) At least Wizards released the d20 rules for public use while they were simultaneously backstabbing, disemboweling and decapitating the system. But hey, in these 'trying economic times', I guess everyone needs to rush out and buy hundreds of dollars in books for a system that has almost nothing to do with D&D.[/bittervitriol]

      My limited experience with Gentoo has been very smooth. Updating glibc and gcc took a while, but otherwise I've had no upgrade problems.

      On topic, this is excellent. I don't trust PRNG dicerollers. Capturing the actual rolls of real, physical dice is a great idea. Now they just need to build 5 more, and that should cover nearly every game involving dice.

      --
      -1 raving lunatic; +6 subGenius... Things even out...
    10. Re:Need more stats by SCPRedMage · · Score: 1

      Or maybe Traveller.

      I'm sorry, but most of us would rather play a game where Death must wait until AFTER character generation.

      --
      My sig can beat up your sig.
    11. Re:Need more stats by Anonymous Coward · · Score: 0

      Tell this guy.

    12. Re:Need more stats by iocat · · Score: 1

      weakling.

      --

      Dude, I think I can see my house from here.

    13. Re:Need more stats by SCPRedMage · · Score: 1

      I'm sorry, but I prefer my games to NOT waste my time while I'm not even PLAYING them.

      --
      My sig can beat up your sig.
  4. only useful if ... by petes_PoV · · Score: 1

    ... you can program in what you want the dice to come up with - and get it into a Vegas casino

    --
    politicians are like babies' nappies: they should both be changed regularly and for the same reasons
  5. A good first step by Drakkenmensch · · Score: 4, Funny

    Now that we've built a dice-rolling gambling robot, we just need to create a leg-breaking loanshark bookie-bot and we'll be all set to fully automize Vegas!

    1. Re:A good first step by N3Roaster · · Score: 1

      Yes, I saw that episode of Lost in Space. The robot self-destructed (or something like that) after losing at the cup game.

      --
      Remember RFC 873!
    2. Re:A good first step by actionbastard · · Score: 1

      Now that we've built a dice-rolling gambling robot...

      You just need some hookerbots and a casino for blackjack.

      --
      Sig this!
    3. Re:A good first step by rishistar · · Score: 1

      Yes, I saw that episode of Lost in Space. The robot self-destructed (or something like that) after losing at the cup game.

      Did that involve Judy and Maureen Robinson?!!!

      --
      Professor Karmadillo Songs of Science
    4. Re:A good first step by SQLGuru · · Score: 2, Funny

      Are you sure it wasn't the two girls one cup game?

    5. Re:A good first step by th0mas_g · · Score: 1

      Awwww... I see what you did there. Gross, dude.


      :-P

    6. Re:A good first step by sheetsda · · Score: 4, Funny

      fully automize Vegas!

      Someone already tried this. The machine took your debit card number, and generated a random number: if (N <= 45) { card->cash *= 2; } else { card->cash *= 0.5; }. The end result was the same, but for some reason it just didn't have the same appeal. My theory is it has something to do with the tangible dice.

    7. Re:A good first step by Anonymous Coward · · Score: 0

      And Pimpbot 5000 to keep all the robo-hos in check.

    8. Re:A good first step by clintp · · Score: 2, Funny

      Ahhh, forget the casino!

      --
      Get off my lawn.
    9. Re:A good first step by Maxo-Texas · · Score: 1

      Assuming your number was 1 to 100,

      I think best odds you can get in craps is a little over .48 so that would be one reason people wouldn't accept it.
      This seems to support that: http://scoblete.casinocitytimes.com/articles/30.html

      From my two experiences actually playing (i.e. "throwing money into a hole"), some of the "good odds" bets are apparently socially unpopular. There is a social aspect to the table where you join a shared unreality where dice have memory and so on. I didn't totally get it but my EQ bud had done it a lot so he kept me from making a terrible faux pas.

      The electronic slot machines used to be thought of as having worse odds than the mechanical machines.

      --
      She was like chocolate when she drank... semi-sweet at first and then increasingly bitter.
    10. Re:A good first step by Razalhague · · Score: 1

      Like a robot could lose to a human in that game...

    11. Re:A good first step by maino82 · · Score: 1

      human... robot... i think everyone loses in that game

    12. Re:A good first step by Amazing+Quantum+Man · · Score: 1

      Don't forget the hookers! In fact, forget the dice...

      --
      Fascism starts when the efficiency of the government becomes more important than the rights of the people.
    13. Re:A good first step by LandDolphin · · Score: 3, Interesting

      They are socially unpopular because yo uare betting against everyone. Craps, unlike poker, you are all on the same if. If one win, you al lwin type situation. You all want the roller to keep rolling and winning. When someone comes and bets the rolelr will lose, he is at odds with the interests of everyone else at the table.

      That beign said, I love craps. It's a lot of fun. Do I play to win money? No, but it is nice when it happens. I play for the entertainment value. The excitement of when the dice hit my number, the cheering and yelling. What other game at the casino can I stand around with complete strangers (and friends) and yell and scream and generally make an ass of myself? Sure, you get the occational noise from a winnign group at another game, but if you spend any time in a casino you will soon find out that all the noise is coming from the craps table.

      Also, if you bet wise and dont get carried away its pretty easy to play for a long while and break even. But its hard not to get swept up in the game and start betting all crazy. Dont win big if you dont bet big.

      And as for "throwing money into a hole", most entertainment is just that. You spend money on thigns that are enjoyable: A meal, a concert, camping, hookers coke and craps. At the end of each you've lost your money and gained entertment (and possibly a rash)

      --
      Spelling and Grammar errors have been added to this post for your enjoyment
    14. Re:A good first step by maxume · · Score: 1

      That was the prototype. The succeeded with Wall Street, but left out the winning.

      --
      Nerd rage is the funniest rage.
    15. Re:A good first step by Anonymous Coward · · 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.

    16. Re:A good first step by Maxo-Texas · · Score: 1

      I know.. gambling is something I've never understood. It feels worse than actually taking a $20 bill and burning it to me.

      Meanwhile, dumping $1000 on a computer item is completely entertaining and makes sense.

      OTH, remodeling my kitchen made logical sense but emotionally put me under huge stress. It's taken months of the new kitchen to start enjoying it.

      OTH, I trade in the stock market. It is very close to gambling (esp. in today's highly manipulated markets). But it feels better than gambling to me emotionally. My winnings are mixed. They were good for 5 years but the last year has been brutal.

      --
      She was like chocolate when she drank... semi-sweet at first and then increasingly bitter.
    17. Re:A good first step by LandDolphin · · Score: 1

      I certainly feel you. Prior to a friend showing me craps, I never liked gambling. I had the same "throwing money away" feeling about it. Games like Black jack still feel like throwing money away to me, just because there is a gambling excitment, but certainly not the same crowd excitement as craps. Guess the energy level isn't high enough on Black Jack for the price tag to play for me.

      --
      Spelling and Grammar errors have been added to this post for your enjoyment
  6. Why? by Publikwerks · · Score: 2, Insightful

    Why would you need this? And how is this better than a RNG?

    1. Re:Why? by Domint · · Score: 5, Insightful

      Because any developer worth their weight in salt will tell you that RNGs are not truly random.

    2. Re:Why? by gnick · · Score: 2, Insightful

      Trust.

      People who know better will trust a good RNG just as well (as long as it's open source) - They're not perfectly random, but probably just as random as the dice roll. But if you're dropping $$ on the roll of a couple of dice (especially if you're remote), people will put more faith in a couple of pieces of bouncing plastic than they will a computer telling you that you just lost your $100 with no explanation.

      Of course, that's purely speculation - Why RTFA when you can just glean through the comments.

      --
      He's getting rather old, but he's a good mouse.
    3. Re:Why? by CheddarHead · · Score: 5, Insightful

      From TFA:

      "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."

      So, basically it was to quiet complaints about the randomness of the computer generated dice rolls. I question whether it's really better, but the players think it's better and in this context I guess that's all that matters.

    4. Re:Why? by Anonymous Coward · · Score: 0

      And yet, there is probably some bias in the way that the dice are thrown. This project is cool, but as a source of random numbers it is quite silly.

    5. Re:Why? by Anonymous Coward · · Score: 0

      Like if the dices were truly random...

    6. Re:Why? by Anonymous Coward · · Score: 0, Informative

      Why would you need this? And how is this better than a RNG?

      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.

    7. Re:Why? by gnick · · Score: 2, Insightful

      That's true RNGs are not truly random. But, then again, neither is anything else. Just sufficiently random to be indistinguishable from an actual random event.

      --
      He's getting rather old, but he's a good mouse.
    8. Re:Why? by eldavojohn · · Score: 5, Interesting

      Because any developer worth their weight in salt will tell you that RNGs are not truly random.

      No, they are not truly random. Nor is his dice machine, as the dice are possibly imperfect and subject to gravity or the way it reloads them into the hopper. Influences could be anywhere.

      I would be interested in seeing him run this machine for 30 days and then compute the Shannon entropy on the results and then compare this to popular RNGs out there.

      Although I would expect the RNGs (however flawed) to perform better, it would be interesting nonetheless.

      --
      My work here is dung.
    9. Re:Why? by aurb · · Score: 5, Funny

      ...neither is anything else. (random), ...an actual random event..

      It's a good thing I have my paradox absorbing crumple zones installed...

    10. Re:Why? by OglinTatas · · Score: 1

      I like the lava lamp rng
      http://www.lavarnd.org/

    11. Re:Why? by canajin56 · · Score: 1

      Ummm, a PRNG in your computer, open source or not, is purely deterministic, and so the complete and polar opposite of "Random." A hardware RNG could have open source drivers or not, but it's a hardware device. They tend to be skewed and also to have some dependence between bits, when what you really want is for each bit to be uniformly and independently random. Either way, even with your own drivers, you have to trust the device. Cue the Dilbert comic: "Seven Seven Seven Seven Seven Seven..." "Are you sure it's random?" "That's the thing with randomness, you can never be certain." It's actually quite difficult to look at the output from a RNG and determine if there's any dependence between bits, and if there's any skew towards either 1 or 0.

      --
      ASCII stupid question, get a stupid ANSI
    12. Re:Why? by Anonymous Coward · · Score: 5, Insightful

      And yet, there is probably some bias in the way that the dice are thrown. This project is cool, but as a source of random numbers it is quite silly.

      If this does not qualify as random, what does?

      A dicerolling machine external to and slightly more complex than the universe itself?

      I normally don't mind the arguments or objections that this or that is psuedo-random rather than truly random, especially when it comes to computer generated numbers designed to do X or Y.

      But really these are physical dice. If this doesn't cut it for you, what will?

    13. Re:Why? by goodmanj · · Score: 1

      If you think software RNGs aren't random enough to play board games with, you'd better delete your Webmail accounts and close your online bank account, because Web security absolutely relies on them.

      On the other hand, this machine is frickin' awesome. But I do worry about the fact that it makes die rolls in large batches, and stores them for hours before using them. A fine opportunity for cheating if you can get access to any of the dice-roller's controller software.

    14. Re:Why? by Domint · · Score: 3, Interesting

      True. I'd like to see someone put this machine up against a currently accepted RNG routines and see which one produces more 'believably random' results.

    15. Re:Why? by ArsonSmith · · Score: 1

      I remember an article many years ago about how dice stack up to being fully random. They had some interesting pictures of the dice in 6 stacks something like 50 high. Each stack with the dice arranged with a different number up. The height of the stacks were different by several inches. There was also several rolls that were then statistically analyzed to find that imperfections in the dice made them favor certain rolls over others.

      --
      Paying taxes to buy civilization is like paying a hooker to buy love.
    16. Re:Why? by PMuse · · Score: 1

      To improve the randomness of this robot, it would be useful to identify and remove all those dice that are insufficiently random. For a nominal fee, I am willing to recycle this byproduct . . .

      --
      "We reject as false the choice between our safety and our ideals." --The American President (20.1.2009)
    17. Re:Why? by SCPRedMage · · Score: 1, Insightful

      "Dices"? Really? "Dice" is the plural, "die" is the singular.

      --
      My sig can beat up your sig.
    18. Re:Why? by SCPRedMage · · Score: 1

      But I do worry about the fact that it makes die rolls in large batches, and stores them for hours before using them. A fine opportunity for cheating if you can get access to any of the dice-roller's controller software.

      An interesting point, but if the site goes through THOUSANDS of rolls a day, then there really isn't any way to be able to exploit that. You could take a look at the order and such, but with 80,000+ rolls being used a day, there's no way of knowing which ones you'll get in advance.

      --
      My sig can beat up your sig.
    19. Re:Why? by Anonymous Coward · · 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.

    20. Re:Why? by goodmanj · · Score: 1

      80,000 rolls a day is one every 10 seconds. You can't be sure exactly which roll is going to be yours vs another gamer's, but you can use the average of say, the next 50 rolls to give yourself an edge the same way a card-counting Blackjack player would.

    21. Re:Why? by yuna49 · · Score: 1

      Because people can see dice falling, but they can't see a random-number generating algorithm at work. In cases like this, computers are inherently untrustworthy because they can be programmed to produce a desired result. It's all about credibility, not whether a computer algorithm is better or worse at randomization than this device.

    22. Re:Why? by Anonymous Coward · · Score: 0

      not truly random. But, then again, neither is anything else.

      I would like to direct you to the website "4chan", then click on "b"

    23. Re:Why? by AlecC · · 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.

      --
      Consciousness is an illusion caused by an excess of self consciousness.
    24. Re:Why? by CarpetShark · · Score: 1

      Because any developer worth their weight in salt will tell you that RNGs are not truly random.

      And neither are robots using RNGs to decide how to throw a die.

    25. Re:Why? by Anonymous Coward · · Score: 0

      You're thinking of GameScience. His two youtube videos are posted at http://www.gamescience.com/

      The man is clearly a fanatic about dice, and it's lovely to watch his clearly honed-over-time rant about inferior dice. Heck, I may get some casino dice to play Settlers with after watching them again.

    26. Re:Why? by AlecC · · Score: 1

      There are well known methods for getting unskewed random numbers from skewed inputs - by discarding data appropriately. And, while it is not intuitive, there are methods for evaluating the randomness of a random stream to some level. And there are methods of generating non-deterministic randomness in computers with mechanical components such as disk drives. The timing of packets from a large LAN can also be used as an effective source of randomness.

      --
      Consciousness is an illusion caused by an excess of self consciousness.
    27. Re:Why? by wjousts · · Score: 4, Interesting

      That's true RNGs are not truly random. But, then again, neither is anything else. Just sufficiently random to be indistinguishable from an actual random event.

      You fail at Quantum Mechanics.

      Hardware random number generators

    28. Re:Why? by CherniyVolk · · Score: 1, Insightful

      I would be interested in seeing him run this machine for 30 days and then compute the Shannon entropy [wikipedia.org] on the results and then compare this to popular RNGs out there.

      After reading the article, I think he's less focused on mathematical accuracy and more focused on appeasing customers. He only went to these lengths due to complaints against his RNGs he used to use.

      This way, those wannabe math majors can't so easily complain.

      Oh, and regardless of Shannon entropy, it is a bit more obvious that this approach will more satisfy a feeling of randomness. Unless the die are weighted, their effect will be random. Or as random as one will possibly get, or as random enough for those mathematicians who work for Casinos across the world who use real die for a variety of Casino games and are willing to shell out hundreds of millions in winnings should the roller win.

      It's already agreed that if we knew the entire state of the universe, that we could see past present and future (just as we can determine numbers either side of a target number within a known sequence or set). But, it's also agreed that it's impossible to know the entire state of the universe. While the universe is big, it's probably agreeable that you don't know the entire state of this machine, nor is it possible. So as each dice has vastly different atomic landscapes on it's edges, and as the surfaces vary greatly with wind currents randomly blowing through, variations in humidity, vibrations, electro magnetics and all sorts of subtle forces and their variations and effects on one another and the die... those dice rolls are as random as they can get.

      Please note, the machine can be manipulated. But, this has nothing to do with the fact that throwing dice upon a surface is sufficiently random.

    29. Re:Why? by TubeSteak · · Score: 1

      No, they are not truly random. Nor is his dice machine, as the dice are possibly imperfect and subject to gravity or the way it reloads them into the hopper. Influences could be anywhere.

      If the influence is consistent across all trials, then isn't the randomness maintained?

      --
      [Fuck Beta]
      o0t!
    30. Re:Why? by Anonymous Coward · · Score: 0

      If the dice have recessed dimples on their numbered faces, already there is a tiny bias due to the fact that the "1" face weighs slightly less than the "6" face. It would be interesting to see whether a slight bias shows up in the distribution of the 6 faces after running the thing for a month or so.

    31. Re:Why? by darkwing_bmf · · Score: 1
      Or as random as one will possibly get, or as random enough for those mathematicians who work for Casinos across the world who use real die for a variety of Casino games and are willing to shell out hundreds of millions in winnings should the roller win.

      Casino dice are of much higher quality than the normal dice shown in the video. Imperfections in one pair of dice can cost thousands of dollars. See this video at about 7:40 into it.

    32. Re:Why? by l00sr · · Score: 1

      Not to mention the fact that errors in the "OCR" system (and there will be errors) are likely to be biased. For example, it may be that the system is more likely to miss pips than to see phantom pips, biasing the RNG towards lower rolls.

    33. Re:Why? by nabsltd · · Score: 1

      You can't be sure exactly which roll is going to be yours vs another gamer's, but you can use the average of say, the next 50 rolls to give yourself an edge the same way a card-counting Blackjack player would.

      Only if "high rolls" were automatically better for you than for your opponent. For example, there might be a game where the roll result means the player number that gets a good thing, and you have the higher player number. Otherwise, it's completely useless.

      How much value would there be in knowing the next 100 rolls of the dice in a particular casino, but not knowing which table would get which roll? Even knowing that a vast majority are sevens would help only a little at craps, as a seven can be a bad thing based on when the roll happens.

      Likewise, for a game like "Risk", lots of high numbers would give a slight advantage to the defender, but since you don't get to choose when you defend, it's not much of a help.

      For a game like "Monopoly" or backgammon, rolls are essentially unrelated, and no value is inherently better, but all values are situational. Doubles are good in "Monopoly" because you roll again, but too many and you go to jail. Doubles are very good in backgammon, unless you can't move that number of spaces with any piece, in which case they are the worst roll at that time.

    34. Re:Why? by Anonymous Coward · · Score: 1, Informative

      It's already agreed that if we knew the entire state of the universe, that we could see past present and future (just as we can determine numbers either side of a target number within a known sequence or set)

      Agreed by whom? Those ignoring 90+ years worth of work on Quantum Mechanics?

      But, it's also agreed that it's impossible to know the entire state of the universe.

      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.

      While the universe is big, it's probably agreeable that you don't know the entire state of this machine, nor is it possible.

      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)

      So as each dice has vastly different atomic landscapes on it's edges, and as the surfaces vary greatly with wind currents randomly blowing through, variations in humidity, vibrations, electro magnetics and all sorts of subtle forces and their variations and effects on one another and the die... those dice rolls are as random as they can get.

      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.

    35. Re:Why? by billcopc · · Score: 1

      It's mathematically "better" than the PRNGs used in almost all computers, but really the main reasons why this machine was built are:

      1. Why the hell not, it's geek cool!

      -and-

      2. The guy was being hounded by a bunch of witless nigglers (note the 'L' in that word)

      Myself, I would have built a full-auto dice-gun and shot the whiners in the groin. What's your testicular THAC0, smartass ?

      --
      -Billco, Fnarg.com
    36. Re:Why? by Maximum+Prophet · · Score: 1

      "God does not play dice with the Universe"
      - Albert Einstein


      "Albert, stop telling God what He can do."
      - Neils Bohr


      "God not only plays dice. He sometimes throws them where they can't be seen."
      - Stephen Hawking

      --
      All ideas^H^H^H^H^Hprocesses in this post are Patent Pending. (as well as the process of patenting all postings)
    37. Re:Why? by RobinH · · Score: 1

      Pseudo Random Number Generators (like linear feedback shift registers) are not truly random, but True Random Number Generators, like a reverse biased PN junction with an amplifier and an appropriate sampling algorithm can be truly random (i.e. unpredictable).

      --
      "I have never let my schooling interfere with my education." - Mark Twain
    38. Re:Why? by gnick · · Score: 1

      Still not random. Quantum events can be practically unpredictable and appear sufficiently random as to be indistinguishable from something truly random. But they're still not random. From the article you link to:

      These theories suggest that even though macroscopic phenomena are deterministic in theory under Newtonian mechanics, real-world systems evolve in ways that cannot be predicted in practice because one would need to know the micro-details of initial conditions and subsequent manipulation or change.

      There are a lot of examples of things that are 'random enough' (nuclear decay, thermal noise, atmospheric noise, etc) for anything you could need. But I submit that nothing in the world is really random, just unpredictable and close enough to being random for any practical use.

      --
      He's getting rather old, but he's a good mouse.
    39. Re:Why? by rbarreira · · Score: 2, Informative

      But really these are physical dice. If this doesn't cut it for you, what will?

      http://en.wikipedia.org/wiki/Hardware_random_number_generator

      --

      The AACS key is NOT 0xF606EEFD628B1CA427BEA93A9CA9773F
    40. Re:Why? by MattskEE · · Score: 1

      In TFA he states:
      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.

      It seems pretty reliable since his software is measuring both the color and number of pips at once. And I would be amazed if he didn't check it pretty thoroughly for accuracy. My main concern would be wear and tear on the dice themselves, but I'm sure he's aware of that eventuality. And the corners of dice will wear down first, not the faces.

    41. Re:Why? by Volante3192 · · Score: 1

      Well, in real life the dice are imperfect. They have flaws, pips shift the weight. People don't roll dice the same way twice. Gravity bounces the dice.
      .
      On the other hand, RNGs roll 'ideal dice' which in effect remove all these flaws from the rolling and reduce it to the simplest operation.
      .
      So we've got conflicting goals here. This machine reintroduces many of the original random elements of actually rolling dice. The operation may be imperfect, but those imperfections reintroduce the atmosphere of rolling dice, which may be more important than a pure (round(rand(6))+1).

    42. Re:Why? by gnick · · Score: 1

      Except that with the pips on different faces being different colors, I don't think that missing pips or phantom pips are really a high risk. It almost seems more likely that you'd miss an entire die than miscount the colored pips.

      This isn't facial recognition - It's really basic optical recognition.

      --
      He's getting rather old, but he's a good mouse.
    43. Re:Why? by Eternauta3k · · Score: 1

      If you think software RNGs aren't random enough to play board games with, you'd better delete your Webmail accounts and close your online bank account, because Web security absolutely relies on them.

      I thought they used /dev/random and hired people to wiggle the mouse and type random stuff for entropy :(

      --
      Yeah. Would you choose a neurosurgeon who pokes around people's brains in his spare time? I wouldn't.
    44. Re:Why? by Trinn · · Score: 1

      I hope you enjoy your cold mechanistic universe.

      I, however, choose, however foolishly it may be, to believe in free will, the acausal side of the universe affecting the causal part we normally see by "choosing" how the "random" quantum events collapse (and in other methods, most likely. basically, I believe that the reality rules can be broken)

    45. Re:Why? by Anonymous Coward · · Score: 0

      The guy was being hounded by a bunch of witless nigglers (note the 'L' in that word)

      Right, because when they're on the stand it's always as the defendant, never the witness. They're all witless. Oh wait, you meant the 'l' in the other word. Nevermind.

    46. Re:Why? by rsmith · · Score: 2, Informative

      If this does not qualify as random, what does?

      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 test suite.

      --
      Never ascribe to malice that which is adequately explained by incompetence.
    47. Re:Why? by Tweenk · · Score: 1

      Nuclear decay, as well as some other quantum phenomena, is random in every sense of this word.

      In case you haven't noticed, determinism is long dead. Even though the universe might be deterministic with respect to itself, it is not deterministic with respect to our observations, which is what actually matters.

      --
      Those who would give up liberty to obtain working drivers, deserve neither liberty nor working drivers.
    48. Re:Why? by docbrody · · Score: 1

      What we see here is an example of people being stupid.

      Seems more like an example of people having a little fun as evidenced in these quotes from TFA:

      However, any automation project worth doing is worth over doing, and I way overshot the mark

      As I promised earlier, if you donate to the site and are unhappy about the rolls, let me know and I will pull a die out of the machine, melt it flat and mail it to you, as an object lesson to the other dice.

    49. Re:Why? by profplump · · Score: 1

      He also said the numbers from random.org were no random enough, and they claim to be providing truly random numbers as well.

    50. Re:Why? by gnick · · Score: 1

      Wow. Your post seriously just lead me to tie the simultaneous life/death states of SchrÃdinger's cat to a warm fuzzy world as opposed to a cold mechanical one where, given enough input and a simulator better than we've yet invented, we'd know whether or not the cat was dead.

      Simultaneously living/dead cat awaiting certain doom = Warm fuzzy
      Live cat/Dead cat = Cold mechanical

      Weird.

      --
      He's getting rather old, but he's a good mouse.
    51. Re:Why? by profplump · · Score: 1

      For dice rolls I don't know that it's terribly important for the data to be random. People *think* it needs to be random, and it makes them feel better about losing money, but so long as the values are not predictable by the players and have an even distribution among possible values -- both of which are easy to do with any of several decent PRNG algorithms -- there's probably no practical benefit to increased randomness, no matter how people feel about it.

    52. Re:Why? by Jurily · · Score: 1

      http://en.wikipedia.org/wiki/Hardware_random_number_generator

      FTFW:

      Random number generators can also be built from macroscopic phenomena, such as playing cards, dice, roulette wheels and lottery machines. The presence of unpredictability in these phenomena can be justified by the theory of unstable dynamical systems and chaos theory. These theories suggest that even though macroscopic phenomena are deterministic in theory under Newtonian mechanics, real-world systems evolve in ways that cannot be predicted in practice because one would need to know the micro-details of initial conditions and subsequent manipulation or change.

      I think this device qualifies.

    53. Re:Why? by asdf7890 · · Score: 2, Insightful

      Why would you need this? And how is this better than a RNG?

      Erm, it is an RNG. A proper one at that, not a PRNG.

      OK so there might be a little bias somewhere in the system (a slight manufacturing defect in some of the dice making the chance of getting a six 1 in 5.99999999 instead of 1 in 6, or perhaps some oddity in the optical processing code that makes it fail to recognise the colour representing four more often that it fails to recognise threes) but only a perfect RNG would not have a little bias like this and there is no such thing as a completely perfect RNG. There are statistical analysis and filtering techniques designed to detect and filter/reduce such bias in systems.

      And on the subject of "why would you need something like this?": sometimes wanting something is enough. Sometimes the fun of creating something and the joy of a successful project completed are the whole point.

    54. Re:Why? by The+Archon+V2.0 · · Score: 1

      "Dices"? Really?

      Natural progression from people using "dice" as the singular, I suppose.

      Worse than that, where I used to live, the local vernacular maimed the word "lynx". "Lynx" was used as the plural form, and that's okay, but "link" ("lynk"?) was used as the singular. Drove me NUTS, and no amount of explaining that not every word that ends in an S sound is a plural helped. Still does when speaking to relatives and the topic of what wild animals were recently seen comes up. It's a topic I try to avoid.

    55. Re:Why? by Gerzel · · Score: 1

      Only if you believe that chaos is an illusion based on imperfect knowledge of an ordered universe.

      If you believe that the universe is fundamentally chaotic however then that doesn't hold up.

    56. Re:Why? by retchdog · · Score: 1

      There's no way you can predict the outcome of well-shaken dice. Even if the dice were imperfect, the random numbers could easily be whitened by any number of methods. Atomic decay isn't uniformly random either, of course - the waiting time for decays is exponentially distributed and not stationary (it is decaying after all), but this is easily corrected for.

      Using atomic decay is all about being faster, and less maintenance.

      Stephenson is a pseudo-scientist hack.

      --
      "They were pure niggers." – Noam Chomsky
    57. Re:Why? by asdf7890 · · Score: 1

      But even if your initial source is truly random, you detector may introduce some form of bias through sampling error or the process of turning an analogue readings (i.e. the time between two emissions from your radioactive source) into discrete values, so your generator it not likely to be perfectly random even if its source is.

    58. Re:Why? by ivan256 · · Score: 1

      Randomness wasn't the true goal here. It's quite possible that this machine is less-random than his pseudo-random-number generator.

      The real goal here was catering to the superstitions of the quirky types that play dice games by mail.

      I suspect that this will still be insufficient. Real gaming nerds like to select their weapon (die) for a particular roll.

    59. Re:Why? by SCPRedMage · · Score: 1

      Sounds to me like someone needed to be slapped in the face with volume "L" of any encyclopedia. Or, you know, just shown the "lynx" entry.

      Some people just don't get/accept things like that when explained verbally. Show it to them in writing, like an encyclopedia, and it's more likely to sink in.

      --
      My sig can beat up your sig.
    60. Re:Why? by evilbessie · · Score: 1

      What people believe to be random and what is actually random are usually far apart. Most people would expect to see an approximate even distribution for true randomness, but this is not what happens. You get clumps, actually the size and distribution of these clumps, I believe, can be used to tell exactly how random said data is. My point is most people wouldn't know random if it hit them in the face, so 'believably random' is a non-sequeter.

    61. Re:Why? by wjousts · · Score: 1

      Still not random.

      Yes it is. The idea of a clockwork universe has been dead for decades. I know it's kind of hard to wrap your head around living, as we do, in the macroscopic world (even Einstein had a problem with it, hence the "God doesn't play dice" quote), but you simply cannot know the values of all the variables in play when, for example, an atomic nucleus decays even in theory. Look up Heisenberg's Uncertainty Principle. It's not a statement about how hard it is to measure something accurately, or the limitations of the instruments used, it's a fundamental, inescapable property of the universe.

    62. Re:Why? by Khashishi · · Score: 1

      Because it's awesome.

      And it will attract players because it is awesome.

    63. Re:Why? by Khashishi · · Score: 1

      Quantum processes seem to be truly random. Electronic noise can be considered random. It's not difficult to build a noise module in electronics. Hell, if you have a sound card, you have one already.

    64. Re:Why? by Khashishi · · Score: 2, 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.

    65. Re:Why? by rbarreira · · Score: 1

      As long as the dice are well built, yes. Some other guy posted a link to a video of a professional dice maker saying even a 1/1000th of an inch difference in size between the faces can cause significant bias.

      --

      The AACS key is NOT 0xF606EEFD628B1CA427BEA93A9CA9773F
    66. Re:Why? by Bob-taro · · Score: 1

      Still not random.

      Yes it is.

      I don't think you can say that it has been proven either way scientifically. Classical mechanics assumes everything is deterministic. Quantum doesn't make that assumption, but it doesn't preclude determinism either.

      --
      Prov 9:8 Do not rebuke mockers or they will hate you; rebuke the wise and they will love you.
    67. Re:Why? by Urkki · · Score: 1

      No, they are not truly random. Nor is his dice machine, as the dice are possibly imperfect and subject to gravity or the way it reloads them into the hopper. Influences could be anywhere.
       

      I think you're mixing bias (or whatever is the correct mathematical term) and randomness. Some of the number may be more likely to come up, ie. the distribution is not even (and it's even changing all the time as the dices get slowly worn out in the machine). But the result is still truly random, ie. result of each individual dice throw is totally unpredictable.

      The influences are everywhere, which makes it truly random. Well unless universe itself is deterministic and nothing is random, but I think current view of quantum mechanics is that some things are truly random..

      It's not a perfect RNG of course (uneven distribution is hard to compensate for, when it's not known accurately and is changing as the machine runs), but that's not same as non-random.

    68. Re:Why? by Anonymous Coward · · Score: 0

      Why?! Are you insane!? Just the amazing fun of making it is good enough reason on its own :)

      (And if that's not good enough reason, it's a brilliant promotion for his web page!)

    69. Re:Why? by MyLongNickName · · Score: 1

      Your days must me significantly longer than mine. On mine there are 86,400 seconds per day.

      --
      See my journal for slashdot ID's by year. Mine created in 2005. http://slashdot.org/journal/289875/slashdot-ids-by-year
    70. Re:Why? by Jurily · · Score: 1

      Some other guy posted a link to a video of a professional dice maker saying even a 1/1000th of an inch difference in size between the faces can cause significant bias.

      With one dice, yes. But I don't think it matters that much when you shake up a bowl with 1500 of these.

    71. Re:Why? by Anonymous Coward · · Score: 0

      The randomness of quantum events is a hypothesis.

      An alternative hypothesis is that everything is deterministic. You probably didn't cover hidden variable theory much because it was supposed to have been "disproven" but actually the counterproof was demonstrated to be wrong.

      Assuming a deterministic universe is not a QM fail. The real fail was the argument that if we can't know the value of something then it can't have a value.

    72. Re:Why? by burris · · Score: 1

      random enough for those mathematicians who work for Casinos across the world who use real die for a variety of Casino games and are willing to shell out hundreds of millions in winnings should the roller win.

      Casino ensure their profit by paying less than the true odds. Outside of weighted dice introduced for fraud, the quality of the dice isn't that critical to the success of the casino.

      I can assure you that in any commercial casino you will be asked to stop playing long before you reach 100 million in winnings.

    73. Re:Why? by rbarreira · · Score: 1

      TFA mentions that the way the dice are carried up in the buckets does damage them a bit. It's not a stretch that some of that damage could be self-reinforcing, as in a feedback loop. Potentially ending up at a point where every single dice is biased.

      One solution would be to periodically replace all the die, of course.

      --

      The AACS key is NOT 0xF606EEFD628B1CA427BEA93A9CA9773F
    74. Re:Why? by NightRain · · Score: 1

      Or to accept that fact that anyone using real dice anywhere for anything, rolls damaged dice with possibly self re-enforcing damage and lives with it as being an acceptable level of bias in the dice. Heck, I've roleplayed with the same dice for over 20 years, and those dice have seen better days, yet for all of that they're acceptable to me and the people I game with.

      Remember, this thing is designed to roll dice for a website who's entire aim is to facilitate the playing of physical board games via email. If it's acceptable to bring and roll your own damaged dice in person, I can't see how damaged dice being rolled by a neutral 3rd party is going to be an issue...

    75. Re:Why? by Jurily · · Score: 1

      Potentially ending up at a point where every single dice is biased.

      And what could be more random than a bunch of dice with different unknown bias used together?

    76. Re:Why? by squidfood · · Score: 1

      And what could be more random than a bunch of dice with different unknown bias used together?

      Unless there's a tendency to say, chip around a pip, so (for example) 6s accumulate more chips than 1s...

    77. Re:Why? by Patch86 · · Score: 1

      I think you've hit on the point exactly there.

      This isn't really about making the most random of all possible RNGs as much as providing a service to customers. Tabletop gaming (be it the D&D or casino variety) are traditional games, played for anywhere between decades and centuries, depending on the game. They've always traditionally used dice. These dice are only as random as dice are, and that's as random as they need to be. D&D games have always been played with "biased" dice, and that is either completely unimportant, or is actually part of the appeal in a strange sort of way.

      There's something tactile and enjoyable about rolling dice, which just isn't replicated by a computer presenting you with a number that it assures you is random. The dice-rolling robot is a novel way for people who can't use physical dice (play-by-mail players) to get a little bit closer to the "proper" way of playing their game.

      As long as people will pay for it, it's a great idea. Arguing whether it's scientifically "more random" is missing the point.

    78. Re:Why? by meyekul · · Score: 1

      Did you think they picked lottery numbers from ping pong balls just because its more fun?

    79. Re:Why? by maxume · · Score: 1

      He'd probably be happy to send you the data.

      --
      Nerd rage is the funniest rage.
    80. Re:Why? by Tired+and+Emotional · · Score: 1

      The information density of postings to Slashdot is completely random on P in [-1,1], but the distribution is not uniform - E(P)0. (-1, of course, represents a post that is 100% wrong)

      --
      Squirrel!
    81. Re:Why? by Anonymous Coward · · Score: 0

      Nope. With a complete mapping of the brains of every person on Earth (including current status) and a sufficient simulator, the posts could be predicted.

      But they are, indeed, random enough to escape predictably.

    82. Re:Why? by Anonymous Coward · · Score: 0

      > the dice are possibly imperfect

      Imperfect dice? Somebody should have ordered from Lou...

    83. Re:Why? by Anonymous Coward · · Score: 0

      There's no way you can predict the outcome of well-shaken dice.

      What if I say that rolling a 6 is 0.1% more likely than rolling a 1? What if that non-uniformity is enough to very slightly bias the figures in a one-time pad? It's a tiny change, but because one-time pads need to be truly random in order to be secure, that tiny change is enough to break your code. Your enemy gathers statistics about the messages you send using your one-time pads, and sees that the number distribution is skewed slightly. They make a guess about which numbers are more likely, and that gives them a handle: they can make a start on decoding your message. This is what cold war cryptographers actually did; Stephenson didn't make it up.

      Even if the dice were imperfect, the random numbers could easily be whitened by any number of methods.

      You are also mistaken about this. Imagine that you have a box filled with pseudo random number generators. You can connect them up any way you want, and use any whitening method you choose provided that it's deterministic. Is your output truly random? No; it's still pseudo random, because whitening does not help if there is bias. All you can do is make the bias smaller, and that's never enough.

      Point is, random numbers are really hard to generate, and these mechanical systems are not the right way to do it.

    84. Re:Why? by Manuel+M · · Score: 1

      That's true RNGs are not truly random. But, then again, neither is anything else. [...]

      You fail at Quantum Mechanics.

      Hardware random number generators

      No, sir, you fail. You are forgetting the very essence of applied Science.

      The randomness that appear in Quantum Physics is only part of an axiomatic model that tries to describe reality as far as we can measure it. The model is not the reality. What current Quantum Mechanics describes as random could very well be deterministic in a later, more accurate model.

      Those Hardware RNGs definitely seem random in all possible ways, but what actually happens inside them can be ultimately deterministic. Who knows. Maybe God actually does play dice (or his own version of Warcraft, which is my personal guess).

    85. Re:Why? by DerekLyons · · Score: 1

      People who know better will trust a good RNG just as well (as long as it's open source)

      And how will they know it's a 'good' RNG? Just being able to see the source, or even being able to program, doesn't mean you have the mathematical props to determine how 'random' the generator is.

    86. Re:Why? by retchdog · · Score: 1

      First, there are PRNGs which can boost a seed of perfect randomness into arbitrarily much cryptographically guaranteed pseudo-randomness. (It'll be as hard to break the pseudo-randomness, as it is to break the crypto.) I'm no expert on this, but read this if you're interested: http://www.wisdom.weizmann.ac.il/~oded/foc-book.html

      However, a shaken die isn't even a PRNG in any reasonable sense of the word. A die well-shaken in a symmetric cup, by a chaotic system, is real randomness in the sense that the rolls will be a stationary independent sequence. It'd take some work to convince me that I'm wrong on this, but I'll listen.

      Here is a shitty way to use a biased die in such a system for randomness (it just uses the die as a coin, and uses von Neumann's method; if I were smarter I'd do better). Label the faces 1,2,3 as "A". Label the faces 4,5,6 as "B". Roll the die repeatedly and each time "AB" comes up, mark that as "0". Each time "BA" comes up, mark that as "1". (Making sure not to reuse letters. I.e., "ABA" does not count as "01" but "ABBA" does.) Note that the sequence of 0s and 1s you mark down is uniform independent random.

      And yes, I'd much rather have an atomic decay RNG. However, "perfect enough for cryptography" randomness isn't magic. People screwed up in the past, just because they weren't paranoid enough. Le chiffre de Vigenère was good enough for centuries.

      --
      "They were pure niggers." – Noam Chomsky
    87. Re:Why? by retchdog · · Score: 1

      Let me amend that. The cup doesn't even need to be symmetric, as long as the orientation stays the same across rolls (i.e. you don't spin it). You'll still get a stationary ergodic sequence of rolls.

      --
      "They were pure niggers." – Noam Chomsky
    88. Re:Why? by gnick · · Score: 1

      I never said anything about knowing that it was a good RNG. I said people would trust it as a good RNG. Open source (for most people) is about knowing that it can be reviewed by peers. For a small minority, it's about going through the code and deciding, "well it looks good to me - and surely if there are problems, an expert in the field will catch them." And, for a very small minority, it's about reviewing expert-level details and critiquing them.

      I never said that open-source was better (although it quite often is). I said it was more trusted.

      --
      He's getting rather old, but he's a good mouse.
    89. Re:Why? by FiloEleven · · Score: 1

      Nope. With a complete mapping of the brains of every person on Earth (including current status) and a sufficient simulator, the posts could be predicted.

      I believe the Greeks called such a system Deus Ex Machina =p

    90. Re:Why? by alexbgreat · · Score: 1

      Hows that for counterfeiting? We now have fake random numbers that are indistinguishable from REAL random numbers

    91. Re:Why? by Anonymous Coward · · Score: 0

      Check your math. There are 86,400 seconds in a day.

    92. Re:Why? by FiloEleven · · Score: 1

      Didn't QM experiments show that nonlocal effects exist? And don't nonlocal effects violate causality, throwing universal determinism out the window? Or is this a falsifiability thing?

    93. Re:Why? by Anonymous Coward · · Score: 0

      I've oddly heard this too. Some casino umm, afficionados... yeah.. I know will ONLY play dice games at a table, they swear the machine equivalent is not the same and they get better rolls at the table. And one guy claims the opposite, that he does better on the machine (he varies his bets a lot and things he gets better luck on the machines.)

              This is great, it makes perfect sense for some online dice gaming to use dice, that's all I can say 8-). (I do see one other poster pointing out unpitted dice would be better though.)

    94. Re:Why? by FiloEleven · · Score: 1

      The model is not the reality. What current Quantum Mechanics describes as random could very well be deterministic in a later, more accurate model.

      ...or not. Unless and until it's shown to be deterministic, there's no point in treating it as such. It's fun to think about future models, but the reality is that we can currently see no determinism there, so for us it is functionally, meaning truly, random.

    95. Re:Why? by m50d · · Score: 1
      Oh, and regardless of Shannon entropy, it is a bit more obvious that this approach will more satisfy a feeling of randomness. Unless the die are weighted, their effect will be random. Or as random as one will possibly get, or as random enough for those mathematicians who work for Casinos across the world who use real die for a variety of Casino games and are willing to shell out hundreds of millions in winnings should the roller win.

      Dude, that's what the Shannon entropy would measure. If it isn't _very_ close to log_2(6), there's going to be exploitable patterns (and calculating the entropy is a much better way to determine this than "look and see if I can spot any"). As someone else mentioned, he's not using casino-grade dice, which are manufactured to higher quality standards than those one normally uses.

      It's already agreed that if we knew the entire state of the universe, that we could see past present and future (just as we can determine numbers either side of a target number within a known sequence or set).

      No. Not so.

      So as each dice has vastly different atomic landscapes on it's edges, and as the surfaces vary greatly with wind currents randomly blowing through, variations in humidity, vibrations, electro magnetics and all sorts of subtle forces and their variations and effects on one another and the die... those dice rolls are as random as they can get.

      It's possible. But it's also entirely possible that there's some slight systematic bias towards certain values (e.g. the shapes of all the dice may be very slightly skewed in the same way, from the manufacturing process). Which is why rather than theorising we should do as the GP suggested: run the machine, calculate the entropy and see how it turns out.

      --
      I am trolling
    96. Re:Why? by m50d · · Score: 1
      If this does not qualify as random, what does?

      I think atomic decay counts for more - according to our best understanding of physics it's truly random, rather than merely chaotic. It's possible to have (intentionally or otherwise) a biased dice or roller, wheras you can't make a biased atom.

      --
      I am trolling
    97. Re:Why? by pastafazou · · Score: 1

      I play the games on that site regularly, and the RNG he was using had a tendency to throw out some astronomically rare results a little too often. Having played the board game Axis and Allies for a dozen years, and then playing the computer version at gamesbyemail, I can tell you that an RNG isn't really random like the dice are. And apparently the RNG server he used to use was supposedly a good one.

    98. Re:Why? by LordLucless · · Score: 1

      On individual rolls, probably not. But that's not a particularly good measure of randomness. Over many iterations, you can detect patterns in well-shaken die. Ask any gamer with a "lucky" d20.

      --
      Just because you're paranoid doesn't mean there isn't an invisible demon about to eat your face
  7. Not running Linux? by Anonymous Coward · · 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."

    1. Re:Not running Linux? by ratbag · · Score: 2, Funny

      A Linux user might also completely fail to read the article and discover that some of his users weren't happy with the results of various electronic and noise-based generators that he's used in the past.

      Rob (a not-PC user, in a snarky mood for some reason)

    2. Re:Not running Linux? by jsnipy · · Score: 1

      You should be more concerned that the laptop is held up with a bunjee cord

      --
      -- if you mod me down, I will become more powerful than you can possibly imagine
    3. Re:Not running Linux? by Anonymous Coward · · Score: 0

      I've read the article now, and all I can say is that those people are idiots. But hey, there's a market for homeopathic "remedies" and Monster Cables, so why not "random numbers that are specially random because they come from dice"?

      I hear if you dip your dice in snake oil, you get better results.

    4. Re:Not running Linux? by Anonymous Coward · · Score: 0

      Yeah, and you've never seen someone blowing on dice, or wearing a lucky tie or any shit like that either, have you?

      There are a lot of idiots in the world.

    5. Re:Not running Linux? by Cassini2 · · Score: 1

      No, it's not running Linux.

      It is a really simple machine with an uptime of only 1.5 hours per day. It is really tough to get Windows to do real-time control tasks for long periods of time. Even for activities where reaction times are in the few second range, Windows updates and Anti-virus updates are killers. With Linux, you can make the hard drive read-only, lock everything down, install a real-time kernel, and you can get much more consistent quick reaction times.

      Some days, I feel like organizing a contest: who can keep your Windows machine alive the longest? It just there isn't much point. Windows can barely keep up with a long-term 24/7 data logging activity.

      Does Vista and/or Windows 7 have the bug where if you write files to a directory constantly (say once every few seconds for a month), and then do a directory operation, "dir", the machine locks up for a few minutes?

    6. Re:Not running Linux? by dlakelan · · Score: 1

      Yes, this.

      The real reason to do this can NOT be to get better quality random numbers, since you'd be better off just hooking up a webcam with a piece of tape over the lens and hashing the resulting diode noise.

      The best reason to do this is because you want to play mechanical engineer in your spare time.

      --
      ((lambda (x) (x x)) (lambda (x) (x x))) http://www.endpointcomputing.com a scientific approach to custom computing.
    7. Re:Not running Linux? by profplump · · Score: 1

      If I were this guy I'd fire up my local hardware RNG, then post some blurry pictures of a dice-rolling machine.

    8. Re:Not running Linux? by profplump · · Score: 1

      And because people believe that dice have magical powers.

    9. Re:Not running Linux? by Culture20 · · Score: 1

      If I had to choose between Linux of Windows, it would be Linux, hands down, but if this machine's only network connection exists to pass the random number data, then there's no reason why it needs constant Windows Updates and Anti-virus scans (shouldn't need AV at all*). Pretty much everything you stated except real-time kernel you can do with stock Windows. It doesn't even have to be on the internet itself.

      * http://xkcd.com/463/

    10. Re:Not running Linux? by RiotingPacifist · · Score: 1

      Rob (a not-PC user, in a snarky mood for some reason)

      Posting from a mobile device?

      --
      IranAir Flight 655 never forget!
  8. Jumping to conclusions by Dan+East · · Score: 4, Funny

    "No, it's not running Linux"

    I hate it when people to conclusions. Obviously, it is running linux, just with an XP-themed window manager.

    --
    Better known as 318230.
    1. Re:Jumping to conclusions by Anonymous Coward · · Score: 0

      "The images captured by the USB camera are processed by software I wrote in .NET. "

      Ummm, that makes it pretty likely to be running Windows...

    2. Re:Jumping to conclusions by Anonymous Coward · · Score: 0

      Or Linux with mono.

    3. Re:Jumping to conclusions by Anonymous Coward · · Score: 0

      I hate it when people to conclusions too!!

    4. Re:Jumping to conclusions by AlecC · · Score: 1

      He did sat the software was using #NET, not Mono. Which is pretty MS-ish.

      --
      Consciousness is an illusion caused by an excess of self consciousness.
    5. Re:Jumping to conclusions by Anonymous Coward · · Score: 1, Funny

      Impossible! He said he was using a USB camera!

    6. Re:Jumping to conclusions by PitaBred · · Score: 1

      What? You mean Mono doesn't make .NET 100% cross-platform? Say it ain't so! That's what Microsoft promised!

    7. Re:Jumping to conclusions by Cr4wford · · Score: 1

      I hate it when people to conclusions.

      Dude, did you accidentally the whole bottle?

      --
      Freelance Web Designer - Portfolio
    8. Re:Jumping to conclusions by Anonymous Coward · · Score: 0

      I hate it when people to conclusions.

      I hate it when people accidentally the whole thing.

  9. One-time pads by ickleberry · · Score: 1

    It's the only way to be sure that your numbers are actually random.

    You could use something that uses clock drift to generate random numbers like VIA Padlock but unless you have taken your processor apart, inspected with an electron microscope and put back together you will never know if its the real deal. Government back-doors, etc.

    I wonder how many minutes of kiddy porn one can fit into the ~2 megabits per day generated by this.

    1. Re:One-time pads by $RANDOMLUSER · · Score: 1

      I miss the SGI lava lamp for just that reason.

      --
      No folly is more costly than the folly of intolerant idealism. - Winston Churchill
  10. Yahtzee! by hansamurai · · Score: 1

    From the first picture, all the rolls are 4's, 5's, and 6's! Where's the Yahtzee game that runs off this?

  11. What does that by Anonymous Coward · · Score: 0

    last half dice roll look like?

  12. Great by ForAllTheFish · · Score: 0

    Now we can reliably generate a random number between 1 and 6 millions of times a day.

  13. Okay 1..2..3..4..5..6 by LinuxOverWindows · · Score: 1

    What are they trying to Observe here, Do they think there going to find a magical pattern in rolling dice that can predict future evens :-O Thats it.

    I just don't see the application this Robot solves and how anyone or why anyone would fund this but then again maybe it will produce something we can use.

    1. Re:Okay 1..2..3..4..5..6 by goodmanj · · Score: 1

      RTFA. The machine was built for a good reason.

    2. Re:Okay 1..2..3..4..5..6 by LinuxOverWindows · · Score: 1

      I did and it never says why it's rolling the dice, It says how it's rolling the dice and what happens to the dice after there rolled but in the end rolling rice will only let you randomize a number usually between 1 .. 6.

    3. Re:Okay 1..2..3..4..5..6 by LinuxOverWindows · · Score: 1

      I know he's trying to get Random Dice rolls but there are easier and more practical methods to produce random numbers. If he really needs randomized numbers then grab an advanced Math lib or even better use a atomic or molecular random event. Either way this isn't very practice.

    4. Re:Okay 1..2..3..4..5..6 by JSBiff · · Score: 1

      /dev/random

      Seriously, there are many applications (particularly in cryptography) where having a fast supply of true random data is important. I don't really know if this bot succeeds in that role, but I suppose it could. Before someone objects that 1 to 6 isn't sufficiently random, consider combining multiple dice roles into a single value. . . that is, treat each die as a digit in a base-6 number, to create a value as large as you need - you can create numbers with as many bits as you need this way:

      35105304250512114023

      Seems like a pretty strong source of entropy, although, you would likely want to do something like for whatever the value n on the die is, subtract one, so you have a range of 0 to 5, instead of 1 to 6 (0 is a very important number to have). You will also need to apply some logic to compensate for the bit-width difference between base-2 and base-6 - that is, if you wanted, say, an 8 bit binary value, you need to be able to represent up to at least 0-255(base-10). The minumum number of base-6 digits required to represent values in the range 0-255(base-10), is 4 base-6 digits, but that actually gives a range of 0-1295 (base 10), which requires at least 11 binary bits. So, to get to a random binary value, you need to somehow force the range from 0-1295, down to 0-255. One approach would be to use the modules operation, e.g. n = n mod 255, but that will bias the results a tiny bit (I say that, because, since the range isn't a perfect multiple of 255 [1295 ~= 5.07843 * 255, where the notation ~= is used for 'is approximately equal to'], you will have a slightly greater chance of getting values in the low-end of the range), although I'm not sure if that slight bias would be enough to be a problem in most applications - still, if you're designing a machine to provide cryptographically strong random values for use in binary computers, then you should be using powers-of-two-sided dice (e.g. 4-, 8-, or 16- sided dice).

      Still, the basic concept of a dice-rolling robot could be a good source of entropy.

    5. Re:Okay 1..2..3..4..5..6 by ender- · · Score: 1

      I know he's trying to get Random Dice rolls but there are easier and more practical methods to produce random numbers. If he really needs randomized numbers then grab an advanced Math lib or even better use a atomic or molecular random event. Either way this isn't very practice.

      Posts like this make me wish there were a "Didn't RTFA" moderation...

    6. Re:Okay 1..2..3..4..5..6 by Volante3192 · · Score: 1

      Second paragraph.

    7. Re:Okay 1..2..3..4..5..6 by LinuxOverWindows · · Score: 1

      But I did read that article, did you. It clearly says he's trying to get better random dice rolls, which is exactly what I said. He should grab a random number lib, analyzing million of dice rolls wont help produce better random dice production. What they have to do is use a natural random production method. This idea just doesn't work, seriously I don't understand how this is going to help produce random Die rolls.

      If people aren't happy with the random dice rolls in games then how about getting off the computer and opening up the board game, although of course a person can influence the dice rolls with how they throw the dice so are you telling me that humans don't have random enough rolls. This problem is not serious if people are so concerned that there games don't have random enough dice rolls then don't play the game. There meant for fun, it's not like there going to use these dice rolls to vote a government in.

      People need to grow up and stop being so bothered by something so small, this is a very very little problem, I'm sure there get by.

      So I did actually read the article, So it rolls dice very nice, It take pictures of the rolls and sends the pictures to a server. The dimensions of the is machine are The Dice-O-Matic is 7 feet tall, 18 inches wide and 18 inches deep. It has an aluminum frame covered with Plexiglas panels. Any other question about me not reading the article. I've read it and I don't see the point of this machine.

    8. Re:Okay 1..2..3..4..5..6 by 5of0 · · Score: 1

      If his only goal was to get true random numbers, then yes, your solution would be fine.

      As it is, however, his goal was to convince his users that the dice rolls they're getting are truly random, and not just some magical formula in the computer (which is what they will see any algorithm as, no matter how much you reassure them otherwise). For that purpose, building a giant machine that rolls a bunch of dice and takes pictures of them is the only way to do that.
      It's not the ultimate solution for generating random numbers, no. But for generating believable dice rolls (to the average human, not a math PhD), not much can top it. Not only that, but it's pretty entertaining, and was surely a lot of fun to get up and running.

      --
      You all have Oo.o and Firefox, so get World Wind.
    9. Re:Okay 1..2..3..4..5..6 by binford2k · · Score: 1

      You fail at life. Or at least at reading comprehension.

      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.

      A few years ago I tinkered with a dice rolling machine made from Legos. Though great fun, it was noisy and cantankerous and unreliable, and it never recovered from the move two years ago. But it had made players happy, at least for a while. So I decided to make a 'professional' grade rolling machine

    10. Re:Okay 1..2..3..4..5..6 by LinuxOverWindows · · Score: 1

      Thanks you, you see my point :-)

      Your awesome

    11. Re:Okay 1..2..3..4..5..6 by LinuxOverWindows · · Score: 0
      Wow, Just wow lets do the analysis here, I need to prove that over analyzing things are bad.

      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.

      A few years ago I tinkered with a dice rolling machine made from Legos. Though great fun, it was noisy and cantankerous and unreliable, and it never recovered from the move two years ago. But it had made players happy, at least for a while. So I decided to make a 'professional' grade rolling machine

      Okay so this paragraph in summary says dice rolls aren't random enough! Thats it, nothing else, the summary says dice rolls aren't random.

      Good good glad we got that done next paragraph

      A few years ago I tinkered with a dice rolling machine made from Legos. Though great fun, it was noisy and cantankerous and unreliable, and it never recovered from the move two years ago. But it had made players happy, at least for a while. So I decided to make a 'professional' grade rolling machine

      Okay this paragraph tells us that there is a machine to roll dice, hmm okay awesome lets sum up the two paragraphs....

      Dice aren't random enough and machines roll dice

      Next paragraph to summerize

      The result is what you see here: 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.

      Okay this machine takes picture and sends to a server! done, this paragraph is done.

      Done I've summerized the entire article, how hard is it to read NOT between the lines, don't expand, don't read into, don't over analysis anything. Look it over get the fact and then your done, Only "ender" and me can seem to do that, and BTW I got the quotes from the article

      http://gamesbyemail.com/News/DiceOMatic

      So what don't I understand, This machine rolls dice because dice rolls in computers aren't random enough, it takes pictures and sends them to a server to be looked at. THATS IT THAT IS ALL THIS MACHINE DOES. Stop telling me I don't know how to read english, If you didn't get that, then you don't how to read english because you over read that article, that's all this is about nothing else.

      I've just proven my point and shown that I know what I'm talking about!

      Thanks
      LinuxOverWindows

    12. Re:Okay 1..2..3..4..5..6 by Khashishi · · Score: 1

      but would they be as cool?

    13. Re:Okay 1..2..3..4..5..6 by 2short · · Score: 1

      Your proof of your reading comprehension doesn't even get off the ground:

      "Okay so this paragraph in summary says dice rolls aren't random enough! "

      No, it does not say that. Not even once. You fail. It says "[I] have always received numerous complaints that the dice are not random enough." That is not the same thing as saying the rolls are not, in fact, random enough. The difference is significant and relevant. Scores of people have read the article, and understood why he built the machine and why he thinks it is cool. You have read the article, and apparently do not understand these things. I'm not sure why you think more detailed analysis will convince people this is due to your superior comprehension.

      "I've just proven my point and shown that I know what I'm talking about! "

      Any time you have to end a post with a sentence like that, you may be sure it isn't true.

  14. What a waste, by Anonymous Coward · · Score: 1, Funny

    it can only roll D6s.

    1. Re:What a waste, by goodmanj · · Score: 5, Funny


      it can only roll D6s.

      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!
      .
      .
      .
      .
      (If you're furious with nerd rage right now: I'm kidding. If you're not furious: don't try this at home.)

    2. Re:What a waste, by oldspewey · · Score: 4, Funny

      You know, I never thought there would be a story on /. that could earn the "1500000d6" tag, and then this came along.

      --
      If libertarians are so opposed to effective government, why don't they all move to Somalia?
    3. Re:What a waste, by Anonymous Coward · · Score: 0

      wow, even if your D6 rolls are random, now you *Poof* destroyed your randomness...

    4. Re:What a waste, by Datamonstar · · Score: 1
      --
      The eternal struggle of good vs. evil begins within one's self.
    5. Re:What a waste, by shliddle · · Score: 2, Informative

      Please don't use this method. You are not creating an evenly distributed d8. From all 36 possible outcomes of 2d6, your "mod" d8 comes out like this: 1 - 5 times; 2 thru 6 - 4 times each; 7 - 5 times; 8 - 6 times. You COULD use some kind of conversion table, but even then you would need to throw away 4 of the 36 combinations to make it work.

    6. Re:What a waste, by shliddle · · Score: 1

      This is silly. If you have truly random d6s, applying math against their outcome doesn't make them any less random unless you divide the roll by itself.

    7. Re:What a waste, by shliddle · · Score: 2, Funny

      I got so furious with nerd rage, I didn't even read the last line of your post! :-)

    8. Re:What a waste, by Fwipp · · Score: 1

      Please don't use this method. You are not creating an evenly distributed d8. From all 36 possible outcomes of 2d6, your "mod" d8 comes out like this: 1 - 5 times; 2 thru 6 - 4 times each; 7 - 5 times; 8 - 6 times. You COULD use some kind of conversion table, but even then you would need to throw away 4 of the 36 combinations to make it work.

      (If you're furious with nerd rage right now: I'm kidding. If you're not furious: don't try this at home.)

      Whoooosh.

    9. Re:What a waste, by Anonymous Coward · · Score: 0

      Heh, what if I am merely amused by your swindling ways?

    10. Re:What a waste, by TheRagingTowel · · Score: 1

      Then you can use this little thingy that I've built:
      Led Matrix Dice - Shake to Roll
      It does D99 if you'd really like :)

      --
      4Z5TX
    11. Re:What a waste, by 5of0 · · Score: 1

      Actually, it does. I was just reading about this from another link in a comment hereabouts.

      --
      You all have Oo.o and Firefox, so get World Wind.
    12. Re:What a waste, by crrkrieger · · 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

    13. Re:What a waste, by Anonymous Coward · · Score: 0

      You knew one of us nerds had to do the math on this:

      This method is most biased toward the 8, which takes up 1/6 of the possible dice rolls. Slightly less favored but still biased are the 1 and the 7, each having a 5/36 chance of being rolled. The rest of the rolls (2 through 6) come up 1/9 of the time.

    14. Re:What a waste, by goodmanj · · Score: 4, 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.

    15. Re:What a waste, by Culture20 · · Score: 1
      I know this is a joke (you said so yourself), and the punch line is the uneven distribution, but can't you do:

      digit0 = (1d6)%2 ;
      digit1 = (1d6)%2 ;
      digit2 = (1d6)%2 ;
      d8result = binarytodecimal(texttobinary(concatinate(digit2,digit1,digit0))) + 1 ;

      To get an even distribution?

    16. Re:What a waste, by goodmanj · · Score: 1

      Yes, but you're rolling far more dice than you need to, since you're making a die roll the same as a coin flip.

      And your texttobinary thing made me throw up in my mouth a little bit... :)

    17. Re:What a waste, by goodmanj · · Score: 1

      The "right" way to do it is something like

      floor(8*(36*roll3+6*roll2+roll1 - 43)/215)) + 1

    18. Re:What a waste, by shliddle · · Score: 1

      I read that link and it's true, you could easily reduce precision, and yes, you COULD reduce the randomness, but they are mutually exclusive. If I wanted to turn a d6 with a particular level of randomness into a coin flip (d2) by doing modular arithmetic, you don't automatically increase the predictability of that d2 merely because of the math you're performing.

    19. Re:What a waste, by Anonymous Coward · · Score: 0

      chance of rolling a 2 is 1/36, which mods to 2, +1 is 3.
      chance of rolling a 3 is 2/36, which mods to 3, +1 is 4.

      showing less work:
      5 = 3/36
      6 = 4/36
      7 = 5/36
      8 = 6/36
      rolling an 8 will mod to 0, so now:
      1 = 5/36
      2 = 4/36
      3 = 3/36 (and this will be added to the previous chance
      4 = 2/36
      5 = 1/36 (a roll of 12, modded to 4, +1)

      The numbers 3, 4 and 5 can be rolled two ways.

      Final probability
      1 = 5/36
      2 = 4/36
      3 = 4/36
      4 = 4/36
      5 = 4/36
      6 = 4/36
      7 = 5/36
      8 = 6/36

      So yes, it's pretty even, but not perfectly even. 8 will come up the most often.

    20. Re:What a waste, by Anonymous Coward · · Score: 0

      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.

      Cue the Dilbert comic: "Seven Seven Seven Seven Seven Seven..." "Are you sure it's random?"

      Yes, i thought the distribution will be more a lot lot more biased.
      With 6/36 chance for 7 and 5/36 chance for the other numbers, there is one bias on 8. Ouch on the hitpoints.

    21. Re:What a waste, by Anonymous Coward · · Score: 0

      Cue the Dilbert comic: "Seven Seven Seven Seven Seven Seven..." "Are you sure it's random?"

      1 : 5/36 (0.83/6 or 1.25/9)
      2 : 4/36 (1/9)
      3 : 5/36 (0.83/6 or 1.25/9)
      4 : 4/36 (1/9)
      5 : 4/36 (1/9)
      6 : 4/36 (1/9)
      7 : 6/36 (1/6)
      8 : 6/36 (1/6)

      It's as horrible as i thought

  15. I believe by FlyingSquidStudios · · Score: 1

    I believe you'll find that six photos is really all you require in this situation...

  16. What has the netbook got to do with anything? by NoNeeeed · · Score: 1

    In other new: Computer "used for computer type work" shock!

    I mean this is a really neat little hardware project (reminded me a uni hardware project I did), but the bit about the Dell is just fluff. What's so amazing about using a netbook? it's just a small laptop.

    If it was being run off a trinary mechanical computer powered by a hamster then *that* would be quite interesting.

    Actually, I think I have a new project....

  17. Accuracy by wangerx · · Score: 2, Interesting

    Very cool device! It does lack in accuracy. Pitted dice are off balance and the 1 will land on the bottom more often than not. That is why Vegas does not use that type of die. There is error in the machine; look closely at the video where the dice get stuck at the top.

    1. Re:Accuracy by Anonymous Coward · · Score: 0

      Very cool device! It does lack in accuracy. Pitted dice are off balance and the 1 will land on the bottom more often than not. That is why Vegas does not use that type of die.

      The only way to measure the randomness of this machine, regardless of the particular dice used, is to measure how accurately it achieves a uniform distribution of results. Run it for a little more than a week to get 12 million rolls, see how each number count deviates from the expected 2 million.

      For one time pad generation, call all odd rolls ONE and all even rolls ZERO. Assuming a verified 50-50 Odd-Even split, this would work even if the machine proves to be off-balance, right?

    2. Re:Accuracy by Anonymous Coward · · Score: 0

      "For one time pad generation, call all odd rolls ONE and all even rolls ZERO. Assuming a verified 50-50 Odd-Even split, this would work even if the machine proves to be off-balance, right?"

      Almost, you still need one more step: You would need to XOR two consecutive rolls after you performed your ONE/ZERO determination. Use the output of the XOR as your random bit. That would remove any bias in the distribution.

      Roll 1 Roll 2 Output
          0 0 0
          0 1 1
          1 0 1
          1 1 0

  18. From TFA by neoflame · · Score: 2, Informative

    "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." Math.random is an LCG and so therefore of dubious quality. Random.org, though, is a true RNG (not a PRNG). If random.org is not random enough, either they're doing something quite horribly wrong or (far more likely) players don't actually understand what random means.

    1. Re:From TFA by Volante3192 · · Score: 1

      The customer is always right. The customer is also always an idiot.

    2. Re:From TFA by The+Archon+V2.0 · · Score: 1

      If random.org is not random enough, either they're doing something quite horribly wrong or (far more likely) players don't actually understand what random means.

      I'd bet money on "players don't actually understand what random means". This device is interesting, though. Broken down:

      PROBLEM: His users complained because they don't trust a computer to say "I rolled a die, and it came up 6."

      SOLUTION: Build a device that rolls real dice, and hook it to a computer. Then the computer says "I rolled a die, and it came up 6."

      Let's face it - if the video camera was a dummy and the PC was just still getting results one of the old ways, would anyone notice? Either way the ultimate result is a computer saying "Yeah, it's a five. Now it's a four. Now a five. etc."

  19. how much $? by ilblissli · · Score: 0

    its nice and all to know that it can roll 1.3 million rolls a day, but how is it useful? is it actually being put to use by any gambling sites? if not, why not? if so i think a more interesting stat is how much money does this little machine rake in each day?

    1. Re:how much $? by Anonymous Coward · · Score: 0

      RTFA

  20. wrong lava lamp by OglinTatas · · Score: 1

    heh, I actually linked to the wrong lava lamp random number generator, but after clicking around the site, I decided that this one was better, though no actual lava lamps were harmed in the creation of this rng.
    so again: http://www.lavarnd.org/

  21. My favorite homemade entropy is by CustomDesigned · · Score: 1

    a digital camera with the lens cap on. Especially, if you can get the raw pixels, it contains a large component of true quantum randomness. Just run the bits into PRNGD (which runs the bits through a secure hash and adjusts the input/output bit rates according to the estimated randomness of the sources) with a conservative estimate of the percentage of quantum randomness.

  22. No random, no paradox by iYk6 · · Score: 1, Funny

    Nothing is random, but you can still imagine the results of a random event. There is no paradox here.

    1. Re:No random, no paradox by aurb · · Score: 1

      But you can still imagine the paradox...

    2. Re:No random, no paradox by JustOK · · Score: 3, Funny

      I have random thoughts like that all the time.

      --
      rewriting history since 2109
    3. Re:No random, no paradox by Culture20 · · Score: 1

      Paradoxically, those thoughts deliberately Soviet Russia about you all the time.
      Wait, I screwed that up somehow...

    4. Re:No random, no paradox by masterzora · · Score: 1

      It seems like a perfect example of a paradox to me. It seems contradictory but, as you pointed out, it's not.

      --
      Remember, open source is free as in speech, not free as in bear.
  23. 1.3 million rolls a day! ... by tbi · · Score: 4, Funny

    ...Imagine a beowulf cluster of those...

    1. Re:1.3 million rolls a day! ... by Verdatum · · Score: 1

      Damnit, you beat me to it.

    2. Re:1.3 million rolls a day! ... by Anonymous Coward · · Score: 0, Redundant

      Finally a sensible way to play a 4000 pt Imperial Guard list!

    3. Re:1.3 million rolls a day! ... by Evan+Meakyl · · Score: 1

      When will the petarolls barrier be broken? I can't wait!

    4. Re:1.3 million rolls a day! ... by skeeto · · Score: 1
      1.3 million rolls only comes out to about 410 kB of random data per day, so they may eventually need more machines. Imagination becomes reality.

      l(6)/l(2) * 1300000 / 1024 / 8
      410.21133434295691356335

  24. Video by 0x2A · · Score: 1

    That looks pretty fast... http://www.youtube.com/watch?v=7n8LNxGbZbs

  25. I did read the article by Anonymous Coward · · Score: 0

    I know its a terrible faux pas but i did read the article. So I'll sum it up.
    The owner runs a site were people play board games via email. He tried using computer randomizing programs but the customers complained. Some people put more effort into the statistical analysis of the dice roles than they put into there graduate thesis.
    So he made a device to role a crap load of dice so customers couldn't complain. but they probably still will.

    So yes he tried the usual, easy roots first and the machine does have a purpose.

  26. Oh, man, I need one of these! by Tetsujin · · Score: 1

    If I can get my hands on one of these things I can automate the process of rolling all the ones out of my collection of D20's... While an ordinary elimination process might use an average of 400 dice to yield a single die with a 1:8000 chance of rolling another "one" on the next roll, this kind of automation could process thousands of dice in a reasonable amount of time - yielding either a higher volume of dice with a small chance of rolling another one, or producing dice with an even smaller probability of rolling a one...

    --
    Bow-ties are cool.
  27. I'm going to be rich by Nighttime · · Score: 1

    Finally, I can automate my production of D6s/D20s with all the 1's rolled out of them.

    --
    I've got a fever and the only prescription is more COBOL.
  28. Correcting for loaded dice. by Animats · · Score: 1

    If you build something like this, there's another step needed to get reliably random numbers. Take two successive outputs. If A>B,, output a 1. If A Even radioactive random number generators have to use that step. That was discovered in the 1950s.

    1. Re:Correcting for loaded dice. by Animats · · Score: 1

      (Sorry, HTML escape problem.)

      If A < B, output a 0. If A==B, ignore and try again.

    2. Re:Correcting for loaded dice. by Anonymous Coward · · Score: 0

      Isn't this about getting the result on the face of the dice (between numDice and 6*numDice) and not about generating random numbers in [0,1]?

  29. Practical randomness by wayward_bruce · · Score: 1

    If you want good randomness: buy a cheap microcontroller, one thermistor, one photoresistor, hook it all up to the charger of your old cell phone (or power from USB but then there's not much noise) and sample away. Or alternatively just grab that Brownian motion detector.

    1. Re:Practical randomness by drinkypoo · · Score: 2, Funny

      Or alternatively just grab that Brownian motion detector.

      If only it would just hold still for a sec...

      --
      "You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
  30. 1.3 million rolls... by seandiggity · · Score: 1

    ...ought to be enough for anybody.

    --
    Geeks like to think that they can ignore politics, you can leave politics alone, but politics won't leave you alone.-rms
    1. Re:1.3 million rolls... by __aasqbs9791 · · Score: 1

      I take it you've never played Rolemaster then? ;^)

  31. Saving throw? by Zenmonkeycat · · Score: 1

    Now I can finally realize my dream of a real-time GURPS campaign, where you roll on every single action.
    "Okay, I step forward."
    "Success. You move one hex."
    "I step forward again."
    "Success. You move another hex."
    "I take another step."
    "Oh, critical failure! You actually trip and fall backwards one hex!"
    "I'm going to kill the GM."
    "Failure. You do no damage."
    "No, that wasn't an action, I really am going to come across the table and kill you."

    --

    *****
    Dear Mary,
    I yearn for you tragically,
    A.T. Tappman, Chaplain, U.S. Army.

  32. D20 by Anonymous Coward · · Score: 0

    I'll make the obvious geek-troll comment: call me when it will automate Dungeons and Dragons for me! Until it can recognize a fistfull of d20's, d12's, d10's, d6's and d4's, I'm not giving him any of my cheetos!

  33. how long do the dice last? by Dan+Ost · · Score: 2, Insightful

    The article mentioned that the dice get beat up pretty bad at the bottom of the machine. I have three questions:

    1. how long do the dice last before needing to be pulled out of the machine and replaced?

    2. how are damaged dice identified to be removed?

    3. does the software recognize when damaged dice are causing errors (for example, when the paint from a pip has been completely chipped off)?

    --

    *sigh* back to work...
  34. Awesome by zztong · · Score: 1

    Awesome.

    Some dice vendor should get one made to show at game conventions.

  35. All your dice suck - Testable! by braindrainbahrain · · Score: 2, Informative

    Now finally, we have the technology to experimentally verify the claims made by a certain dice manufacturer!

  36. Why? Because revenge is possible by Actually,+I+do+RTFA · · Score: 2, Interesting

    if you donate to the site and are unhappy about the rolls, let me know and I will pull a die out of the machine, melt it flat and mail it to you, as an object lesson to the other dice.

    'nuff said

    --
    Your ad here. Ask me how!
  37. Version 2 by serutan · · Score: 1

    On-demand dice roller with multiple numbers of d4, d6, d8, d10, d12 and d20. Type in 4d8+1 and see your dice rolled live on streaming video.

  38. the point is that they're dice by Anonymous Coward · · Score: 0

    there's all this talk of this contraption vs rngs, but when you play an actual boardgame, you use actual dice, and these are actual dice. so in that sense, it is more realistic than rngs. maybe rngs would produce more random numbers, but it's irrelevant. they want what dice would do.

  39. How to improve effeciency by shliddle · · Score: 1

    It seems there is a fair amount of waste by the buckets not being full. When the machine started up, the buckets were all packed with dice. Only after they started cycling did you see the gaps. This tells me you could put many more dice in the hopper to ensure a full bucket each trip. If you were okay with the current rate of rolls, you could slow the track down, thereby extending the life of your dice. So long as each bucket was full, you would get the same number of total rolls per day.

  40. How do you know QM is random? by Colin+Smith · · Score: 1

    Perhaps it's completely deterministic, but we just don't know how to look at it.
     

    --
    Deleted
    1. Re:How do you know QM is random? by mrsurb · · Score: 2, Interesting
      It's been a decade since I looked at QM...

      Bell's theorem loosely states: No physical theory of local hidden variables can ever reproduce all of the predictions of quantum mechanics.

      Quantum mechanics is inherently statistical and non-deterministic in nature. If Bell's theorem holds (and experiments have so far gone its way), then the only way to retrieve your deterministic universe from the clutches of quantum mechanics is to allow non-local effects - which brings in problems of instantaneous travel, faster-than-light communication etc...

      Sorry Einstein, it looks like God DOES play dice with the universe.

    2. Re:How do you know QM is random? by ppqq · · Score: 1

      What about non-local hidden variables?

    3. Re:How do you know QM is random? by Culture20 · · Score: 1

      Ah, but to the layman (as myself), "random" does not equal "non-deterministic". There is(may be) a method by which quantum stuff functions, so it is not truly random, but we _can't_ know what that method is, so it's non-random, and non-deterministic. To a physicist, the two terms might equate, but they better start using the language we use before we think they're saying something regoddamdiculous. Tee hee, see what I did there?

    4. Re:How do you know QM is random? by Anonymous Coward · · Score: 0

      Actually Bell's theorem proves nonlocality.

    5. Re:How do you know QM is random? by m50d · · Score: 1

      If you assume an entirely deterministic universe then you can of course make it work. But the particle has to somehow "know" which direction the experimenter is going to choose to measure it in, before he does - which seems rather implausible.

      --
      I am trolling
  41. I fell in love with this site 4 years ago! by Alascom · · Score: 1

    Gamesbyemail is awesome. I've been playing Axis&Allies on the site for almost 4 years now, and I must say its quite brilliant.

    Great gaming engine, and he has covered all the classic games.

  42. No, it's not running Linux by warrenb10 · · Score: 1

    Linux doesn't play dice with the internet.

  43. Not random enough by AlpineR · · Score: 1

    Magic Online generates a lot of complaints about randomness. That's a card game with decks of at least sixty cards built from a pool of thousands of different cards, shuffled, and drawn one by one. The most common complaint is uneven distribution of land cards (about 40% of most decks).

    One theory for why people complain is that everybody's brain is wired to notice unusual draws, and when you play online you have far more draws than in paper. A typical rate of play with paper is six games an hour, for two to four hours a week. A typical rate of play online is twelve games an hour, for two to four hours a night. With that larger sample, you see more weird stuff and accumulate more anecdotes about how broken the shuffler is. (You also have someone to blame other than yourself and a global audience for your complaints.)

    1. Re:Not random enough by neoflame · · Score: 1

      The thread you linked to contains what I think is the most likely explanation: humans are not used to actual randomness. Back when I played MtG, people would regularly "shuffle" by separating out lands and non-lands, interleaving them, and then doing a few rounds of hand-shuffling. This is great if you're trying to avoid mana screw and your opponent doesn't mind this sort of visible cheating (or is doing it themselves), but it's not random, and it's not what one gets from a truly random shuffle.

    2. Re:Not random enough by Kaboom13 · · Score: 1

      As a long time mtg player, I can attest that almost every player complains about land distribution at one point or the other. The difference between a good player and a bad one is in building the deck with the right amounts of land, card draw, mana fixing, etc. to overcome all but the unluckiest of hands (I for example, once pulled 7 land cards in my hand of 7 in a deck that only had 15 out of 60 cards there were land, mulliganed to redraw with a 6 card hand, and drew 5 land. It happens (and yes I shuffled very well, as my playing group will reshuffle your deck for you if you do a halfassed job at it). But I, and everyone who was there, will remember that hand for a LONG time. The 10+ normal hands I got that night are already forgotten. Humans are terrible at determining randomness, and the sample any human is likely to see in his life from any given system is too small to properly calculate randomness with any accuracy. If a hundred thousand people are playing MTG:online, the chance of someone getting consistently bad hands is very high. That person will call the system broken of course, but the reality is if someone wasn't getting shafted, the system wouldn't be random.

    3. Re:Not random enough by Kaboom13 · · Score: 1

      IIRC, the rules of mtg allow you to cut the opponents deck in any manner you choose (within reason of course) so long as you don't flip any cards over. The easy way to teach people to shuffle correctly is to cut their deck into 60 seperate piles, then start randomly putting them back. This method is also fun if they are playing without card sleaves, and their land is from an older edition. The older cards have a different coating and tend to have less vibrant colors, so you can easily pick them out and stack them all in one big clump.

  44. Hmm, Where's Ashy Larry, and Leonard Hamilton... by MetaPhyzx · · Score: 1

    A new contestant for the World Series of Dice!

    --
    Blacker than my baby girl's stare. Black like the veil that the muslimina wear. Black like the planet that they fear...
  45. Who wants to bet by Bruiser80 · · Score: 2, Funny

    that the programming required to make this machine worked required at least one random function :-)

    --
    Arguing with an engineer is like wrestling a pig in the mud. After a while, you realize the engineer enjoys it.
  46. What's so clever? by thtrgremlin · · Score: 1

    Why not use traditional white dice with black dots on a white background. Then it is as simple as
    visgrep <CameraImage.png> <DotImage.pat> -t 10? | wc -l
    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.

    --
    Want Big Business out of government? Take away the incentive and start by getting government out of big business!
    1. Re:What's so clever? by Culture20 · · Score: 4, Informative

      Why not use traditional white dice with black dots on a white background. Then it is as simple as

      visgrep <CameraImage.png> <DotImage.pat> -t 10? | wc -l

      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.

  47. Ig Nobel prize! by Megane · · Score: 2, Insightful

    This guy deserves to be nominated for an Ig Nobel prize in statistics!

    --
    #naabhaprzrag, #sverubfr-000, #agi-fcbafberq, negvpyr[pynff*=' negvpyr-ary-'] { qvfcynl: abar !vzcbegnag; }
    1. Re:Ig Nobel prize! by Anonymous Coward · · Score: 0

      He would need to do some sort of research first. All he's done is made a really cool machine. Were he to perform some of the analyses suggested by folks here, then certainly he should be nominated.

  48. This is perfect for DnD by anexkahn · · Score: 1

    You know, for when I need to make that 1.3 million D6 roll for my nuclear bomb attack :)

    --
    Curious about Storage and Virtualization? Check out
  49. Perfect for Champions by Anonymous Coward · · Score: 0

    Any serious player of ICE's Champions game needs one of these machines.

  50. Cheating possible! by Ken+Broadfoot · · Score: 1

    From the article:

    "The data is uploaded to the server via wi-fi in blocks of 1,000 rolls."

    Then:

    "When the server has enough rolls in the bank to last for a few weeks (I am currently storing a million or so), the conveyor shuts down until needed again. It runs for maybe 1.5 hours total each day."

    Somehow get the Wi-Fi blocks then find out where the rolls are being used: ( Online Casino? )
    Compare rolls to realtime results and CHEAT the casino! Woo woo!

    Yes, I know, It would be hard but not impossible.

    Remove the Wi-Fi link is all I am saying.

    --
    Bitcoin pyramid: Join here: http://www.bitcoinpyramid.com/r/1427 it's FREE!
  51. Too stochastic... by jaitropmange · · Score: 1

    "The generation of random numbers is too important to be left to chance." -- Robert R. Coveyou

    --
    But I AM a troll you insensitive clod!
  52. best line from TFA by crazybilly · · Score: 1

    As I promised earlier, if you donate to the site and are unhappy about the rolls, let me know and I will pull a die out of the machine, melt it flat and mail it to you, as an object lesson to the other dice. Tangible revenge.

    An object lesson to the other dice. Bahahah.

  53. I especially like the high-tech... by macraig · · Score: 1

    ... bungie cord holding the Dell Mini-9 in place.

  54. Bad exit strategy! by macraig · · Score: 1

    In the YouTube video, did anyone else notice the half dozen or so dice trapped below the top of the conveyor belt, because they got the "exit strategy" wrong? Where do we file a bug report?