Random number generator

A random number generator produces values intended to be unpredictable or statistically random. It can use physical processes or deterministic algorithms to support games, simulations, cryptography, and sampling.

Pseudorandom number generator: A deterministic algorithm that generates a sequence of numbers designed to imitate random values. Most software games and simulations use this algorithmic form.
Monte Carlo method: A computational method that uses repeated random sampling to estimate quantities or solve problems. Its estimates depend on generating samples with suitable statistical properties.
Cryptographically secure pseudorandom number generator: A pseudorandom generator designed to make its output infeasible to predict without secret information. Ordinary simulation generators may be predictable and unsuitable for security-sensitive outcomes.
Reproducibility: The ability to obtain consistent results by repeating a procedure under specified conditions. Recorded seeds let simulations reproduce results while changing other inputs.
Hardware random number generator: A device that derives random values from measurements of physical processes. It can supply entropy that deterministic software cannot create on its own.
Procedural generation: The algorithmic creation of content, such as landscapes or levels, from rules and input values. Games use generated values to vary content without storing every possible version.
Linear congruential generator: A pseudorandom number generator that updates its state using a linear recurrence modulo an integer. Its simplicity makes it a useful example of a generator with detectable statistical weaknesses.
Random number generator attack: An attack that exploits predictable or weak random values to defeat a system's protections. Predictable outputs can expose encrypted data or enable cheating in games.
Entropy source: A process or measurement that supplies unpredictability to a random-number system. Physical generators depend on a source whose outcomes cannot be reliably predicted.
Random sampling: A method of selecting observations so each eligible member has a specified chance of selection. A generator can supply the draws used to form a sample.

Linked from 9 pages