true random or pseudo-random number generation for die rolls?

6 posts / 0 new
Last post
I can't decide whether the group I'm DMing for should use a true random number generator or a pseudo-random number generator for die rolls. I am unhappy with the true random one because it generates far too many 20s and 1s, but don't know whether pseudo-random would be better. Any thoughts?
Generally, it shouldn't matter if a RNG (random number generator) is pseudo-random or not.  And most RNG you would like use would be psuedo-random.  Psuedo-random generators aren't truly random because the results are based on a complex algorithms performed on a random-ish seed such as the time of day down to the (micro) second.  If you had enough time, patience, and a large enough set of numbers, you could crack a pRNG.  Though effectively, the results are well close enough to random to be equated as such unless you were working in a field where generating truly random results was of utmost importance, such as computer encryption.

True random number generators are usually seeded based on some natural phenomenon that can't be predicted or reproduced, such as atmospheric noise, or the behavior of subatomic particles.  This is about as random as you can get.  So if you are getting more 1's or 20's with a tRNG, blame the dice gods.  Some services, like Random.org, can supply you with true random number, though their interfaces maybe be a bit cumbersome for an environment like D&D where a large volume of different types of rolls are needed quickly.


So most RNGs you would use would be psuedo-RNGs.  And, effectively, their should results should be random enough.  If you can confirm that the results of your RNG are skewed, then the problem isn't with pRNG in general.  Somebody probably screwed up programming the generator, Perhaps 'ceiling' when they should have 'floored', or vice-versa.  In that case, just search for a different one.

TLDR: Just try a different RNG.
Thinking about creating a race for 4e? Make things a lil' easier on yourself by reading my Race Mechanic Creation Guide first.
I am confused by the idea that a 'true random' generator produces too many 20's and 1's
I am confused by the idea that a 'true random' generator produces too many 20's and 1's



This is probably just a matter of perception.  Because 1s and 20s are at the extreme results, and produce special results on attack rolls (automatic miss, automatic/critical hit), they are more memorable.  It's the same way people have 'lucky dice'.

OP, have you actually tracked your RNG results and checked for statistical anomalies over a significantly large sample of rolls?
I am confused by the idea that a 'true random' generator produces too many 20's and 1's



This is probably just a matter of perception.  Because 1s and 20s are at the extreme results, and produce special results on attack rolls (automatic miss, automatic/critical hit), they are more memorable.  It's the same way people have 'lucky dice'.



Availability Heuristic is the term for this phenomenon. 
Thinking about creating a race for 4e? Make things a lil' easier on yourself by reading my Race Mechanic Creation Guide first.
Let me correct what I said: the rng I was using, random.org, SEEMS TO generate far too many 20s and 1s. No, I haven't done any kind of statistical analysis, so this could well be a matter of perception.

Thanks to everybody for their feedback.