A Statistical Benchmarking of Keystroke Entropy Pools Versus State-of-the-Art Software Random Number Generators

Randomness lies at the foundation of computational modeling, simulation, and cryptography. While keystroke dynamics have been explored as a supplementary source of entropy for cryptographic systems, rigorous, modern benchmarking of their standalone statistical quality is rarely reported. This project provides a empirical evaluation of a keystroke-derived entropy pool in contrast with two state-of-arts pseudorandom generators: Python’s built-in random library and NumPy’s PCGC64. Using a suite of established tests (Shannon entropy test, Wald-Wolfowitz runs test, autocorrelation state), a comparitive analysis has been performed to prove the viability of this data suite. Additionally, the pool has been tested with respect to a Quantum Entanglement simulation loosely based on Monte-Carlo Acceptance Rejection method to demonstrate an application.