Zocci has been giving that psuedo-science marketing spiel for years and it's changed nothing.
It's only recently that people have started taking his sales pitch far more seriously than it deserves.
RPGshop.com sells "precision" dice; but are they Gamescience brand?
Zocci has been giving that psuedo-science marketing spiel for years and it's changed nothing.
It's only recently that people have started taking his sales pitch far more seriously than it deserves.
Henrix
Explorer
No, no, they are truly random, but the distribution isn't even. Don't confuse things - a dice with half of it's number the same is still random, just distributed unevenly. And the same goes for a dice with uneven sides.
No. If you're bothered about randomness a program is totally off.
A program does not give you a random number, it's calculated using an algorithm, often using, say, the eigth decimal in the number previously generated.
The distribution is good (if the program's not crap), though, and it seems random, which generally is good enough. But not random.
resistor
First Post
Speaking as a computer scientist and cryptographer, that's bogus. Pseudo-random number generators are MUCH more sophisticated than that, generally built out of mathematically-studied recurrence relations that have demonstrably extremely low correlations between subsequent terms, or out of cryptographic primitives which themselves are intensely studied to minimize correlation effects.
This impression is mostly based on a period from the 70s and early 80s before effective PRNGs were well understood, during which a large number of really crappy LCG generators were used. Nowadays we know better, and either use better generators, or are more careful in how we use LCGs.
You're right, they're not random, but they're likely as just as random, or more so, than anything but casino dice. For example, cryptographic primitives generally cannot allow even minute variations for absolute uniformity. A cipher with a key-length of 128-bits generally can't allow internal correlations that vary from uniform by more than 2^{-64} (really freakin' small).
Pretty much any pseudorandom number generator on a modern system is statistically random to the point that you're likely to notice imperfections in your dice before you notice imperfections in the generator.
If you're really paranoid, cryptographically strong PRNGs also require that, even if you know how the PRNG works and the current value, it's infeasible to predict the next value. For the Blum-Blum-Shub generator, it's demonstrable that predicting the next value is as hard as general integer factorization. And if you know how to do that fast, well, we've got bigger problems than non-random dice rollers. ;-)
Last edited:
Slife
First Post
Not if it connects to http://www.random.org/, or if you use your hardware in non-standard ways to get seeds.
Last edited:
frankthedm
First Post
All that really matters is if zocci was telling the truth about the use of
rock tumblers
to round off the edges of the dice and apply paint to dice. I stopped liking the more rounded edged dice [like the old armory 2 set packs], and now I know why the rounded edges started to bother me.
aurance
Explorer
I agree. Players are not likely to notice any difference in distribution between any gaming dice for table-top RPG purposes. You can set up a chi-square test for your dice if you're really worried, but most likely your Chessex / D&D Basic Set / Cheap Off-brand / whatever dice are just as good as the Gamescience ones.
I do dislike the rock tumbler rounded edges, but not because I believe they cause any significant bias (they don't), but because they have a tendency to roll off the table.
doctorhook
Legend
Obviously you don't think Gamescience dice are worth it. Care to elaborate about why? It could help this discussion a lot, I believe.
Henrix
Explorer
You're wilfully and with ill intent misinterpretating my words. I wasn't talking about sophistication, but the principle.
I didn't think it fit to discuss details - as you say, programs don't generate random numbers, but calculate a seemingly random number.
Many die rolling programs out there are not particularily sophisticated, though.
There are ways to get real random numbers, as Slife says. But it's akin to rolling a dice and us it as a seed (but generally faster methods are preferred). They are not generatad by the computer.
My point is that all this buzz about imperfect dice is really uninteresting - they are more random and have a better distribution than anyone will notice without doing rather extensive statistical analysis of their dice.
If a dice is so imperfect that it'll really matter it'd have to be so skewed as to look bad.
It's far easier to get weird results from different dice throwing tricks. (And that is easier if there are fewer sides on the dice, or if it's a spindown dice.)
Get a dice shaker if you want to minimize skewed results. It probably won't matter much, but probably more than getting Lou's 'perfect' dice.
(And get them if you want - dice are fun! And a little superstition while playing RPGs is merely fitting, if you recognize i for what it is. I prefer to have dice that fits my character, and I won't let anyone else touch them.
)gamestationnet
First Post
Chessex has been selling GameScience for several years, to keep the brand from dying. We (GameStation) will be buying up their existing inventory at the time we launch the product into distribution. They are the same dice...its only the past couple of months, though, that Chx has started putting the GameScience name on the packaging.
You are right in that they only carry the gem and glow colors, and only the most "popular" of those. By the end of April, all colors will be available for your local stores to purchase from distributors or us directly.
lantern314
First Post
I'm of two minds about the Gamescience dice. I love the pointy edges, but dislike the fact that they leave little dents in the lamination on my maps.
