Random Number Generator & Range Picker

In standard web applications, random numbers are generated using deterministic software routines like JavaScript's native Math.random(). These standard engines are classified as Pseudo-Random Number Generators (PRNGs). While fast, PRNGs rely on mathematical formulas initialized by an internal state seed (typically the system microsecond clock). Given knowledge of the seed or a sufficiently long output sequence, future values can be predicted with 100% mathematical certainty.

A common flaw in custom range generators is naive modulo arithmetic: mapping a raw 32-bit unsigned integer $R \in [0, 2^32-1]$ to a target range $[0, M-1]$ using $R \pmod M$. When the total state space $2^32$ is not evenly divisible by $M$, lower numbers in the range receive a subtly higher statistical probability of selection—a flaw known as modulo bias.

To eliminate modulo bias entirely, our generator implements rejection sampling. The engine establishes an exact threshold cutoff $T$:

Any randomly drawn byte sequence evaluating to a value equal to or greater than $T$ is discarded immediately, and a fresh draw is executed. This guarantees absolute mathematical uniformity across every selected bound.