The site's dice are accurately pseudorandom.

The thing about randomness is that in small numbers lots of these 'clumps' or how you want to call them happen. In small numbers (like 20 vs 20) everything is just is completely inpredictable. The only way you can check if this is close enough to true randomness, is by running lots and lots of tests, and measure the results to what mathematics tell you the result should be. That is exactly what Vexer did in this topic - the dice are so close to true randomness that you don't see the difference. So there is not such a thing as extra amount of 'clumpiness' here.

I have not seen nor claimed to have seen an extra amount of 'clumpiness,' because I've also seen and experienced the 'clumpiness' balance out(although it does feel that way at times) but with the amount of people suggesting there is, I was wondering if it was a possibility. I've played the board game for years and can tell you that this happens with real dice as well. My only complaint was the last one where it felt like I couldn't roll positive no matter what, but I've also played where I felt invincible. This happens with real rolls as well. I was just wondering if the possibility that this particular generator created slightly more 'clumpiness' than seems normal because of the limited amount of numbers input.

It is possible that it generates a little bit more or a little bit less clumpiness, as true randomness is not possible, but it is only such a little bit that calling it nothing is close enough :)

So is it fair to say that the dice are not truly random but as close as you can get if you want to keep the game play going at a decent pace? And if the gambling websites use a TRNG how do they keep up the pace? Just curious if anyone knows the answer.

They do not use true randomness either, they approximate it as well. Computers simply cannot generate true randomness. Though, if you are talking about gambling websites, it might just as well be horribly biased :P

I have 2 points to chime in about: 1. Pseudo-random only means that the numbers are not [u]mathematically[/u] random in the strictest sense. That is, it is theoretically possible for a computer (or hyper-intelligent multi-dimensional being) to reverse-engineer the pattern and based on that formula, predict future rolls. Since the seed often comes from the clock, then you would also have to know [i]exactly[/i] when the rolls are occurring in order to do this. And just because this is theoretically possible, doesn't mean it's practically meaningful. Using a pseudo-random generator, the next roll still has a 16.66 % chance of being 1, 16.66% chance of being 2... The only difference is that IF one were to really try hard, it might be possible to reverse-engineer the formula. If someone could do this for the purposes of winning at Risk, I say let them. 2. Las Vegas exists solely because people feel that after a string of losses, they are somehow more likely to start winning. Unfortunately, The roulette table is still 50% Red, 50% Black each and every time*. Actually, 47.4%, 47.4% and 5.2% double-zero. Because the house always wins.

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

Thank you for all the info. I am not trying to argue, prove or disprove. I am only trying to learn and appreciate the patience involved in the explanations provided.

You are welcome, glad to help :)

Re: random number generators http://www.smbc-comics.com/index.php?db=comics&id=2984#comic

Honestly I feel that the dice are either hot or cold. Not provable, just a feeling. I will say this: Random number generator functions work pretty well, but they always are based on the time clock of computer running the script. You may consider getting a true random number generator that is based off a simple PCI card that creates true random vibrations. All you would have to do is just run that bad boy 1 million times and create a table to randomly pull from during the dice roll.

Lol, actually saving a milion numbers would be less random than this one. Really, the difference is so small it is impossible to know whether they are more hot, or less hot than real dice.

Matty, you don't get it. A lot of enterprise companies use a true random number generator and store the results, but have a weekly or monthly process in place to update the results with new ones to keep it fresh.

That would be a lot of work for no difference in results.

Actually we used to use dice that came from a true random number generator. We have 92,000 such rolls stored in the database and we used to cycle through them. We probably needed a lot more but we stopped using them anyway to reduce load on the sql server. No one noticed a difference when I made the switch and the amount of complaining was the same.