Forum adverts like this one are shown to any user who is not logged in. Join us by filling out a tiny 3 field form and you will get your own, free, dakka user account which gives a good range of benefits to you: No adverts like this in the forums anymore. Times and dates in your local timezone. Full tracking of what you have read so you can skip to your first unread post, easily see what has changed since you last logged in, and easily see what is new at a glance. Email notifications for threads you want to watch closely. Being a part of the oldest wargaming community on the net. If you are already a member then feel free to login now.

Hm. At the high level it looks fine enough (I haven't used your actual application). Unity does not have anything approaching True Random, though. At least, to my understanding of it. Have you run just some dice simulations to see if you can dredge up some bias?

M.

I don't think anything has any true random outside of connecting the seed to a quark's quantum state. I'm not sure how unity determines the initial seed value - possibly on the day/time/millisecond. I can set it manually via code, but it appears to be suitably random on a brief examination.

I've run a few 100k and 1mil rolls - and the numbers appear pretty close to what you'd expect, with minor deviation - as expected.

ie: 1mil rolls
roll1 - 499,873 are 4+
roll2 - 500,589 are 4+
roll3 - 500,227 are 4+

so pretty close and well within reasonable margin of error.

Yeah, skews a bit high. But, probably good enough for Mathshammer.

M.

To me the solution would never be to keep the same game design and assume use of digital dice rollers. I think physical dice are a part of the charm and core of the game. I would just like to see them be design with speed in play in mind.
Is it really too non granular? To me I would not mind if the only variants of hit rolls was 3+,4+,5+. Maybe the 2+ and 6+ could be reached with modifiers, or no used at all. Do we really need a unit to have accuracy between avrerage and good? between average and bad?

Or could such a simple solution be enough?

You know what another term for rerolling is don't you?
Double handling.
Of course it's slow.

 Infantryman wrote:
Yeah, skews a bit high. But, probably good enough for Mathshammer.

M.

Actually those results are fairly typical, and make absolute sense given the standard deviation. In this case, it's a million coinflips, which has a standard deviation of 500. So all of those results are within 2 deviations from the mean, where 95% of all trials will fall.

In other words, you could run this experiment 100 times, and you would expect 95 of those experiments to produce a result between 499000 and 501000.

Arguing that there may be bias because "true random" doesn't exist is actually more of a philosophical one than a mathematical one. Computers generating random numbers based on the millisecond is sufficiently random and will follow the appropriate distribution.

In other news, I prefer to roll dice myself. I like dice.