simulationprngrandom

Is it possible to generate random numbers through physical process simulation?


Is it possible to generate random numbers through physical process simulation?

If I simulate the physical roll of a dice (i.e. you picking it up, shaking it in your hand, releasing it onto the table and recording which side ends up "up"...) will that produce a "random" number or would I just have a complex simulation which really accomplishes nothing.


Solution

  • You would have a complex simulation which really accomplishes nothing.

    (other than to return the same value every time, or add unnecessary complexity to an existing PRNG)

    Unless you seed the simulated environment with "random" variables (i.e. depend on a different means of generating random numbers), you will always get the same value back, because every step in the process will happen the same way every time.

    Say you did use a base pseudo random number generator (PRNG):

    Since your simulation would only be as random as the other PRNG you were using, you may as well just use the other PRNG directly, without all the extra work! (In fact, your simulation may not do anything except make the output less random than the values from your base PRNG)