The vocabulary of the Random Data Generator API

Pseudo-Random Number Generator - an algorithm that produces a sequence of numbers that approximates random behavior.

PRNGs use mathematical formulas to generate sequences that appear random but are actually deterministic. Given the same starting value (seed), a PRNG always produces the same sequence. Common algorithms include Mersenne Twister and Linear Congruential Generator. PRNGs are fast and suitable for games, simulations, and testing.

ExampleMath.random() in JavaScript uses a PRNG internally

An initial value used to start a pseudo-random number generator's sequence.

The seed determines the entire sequence a PRNG will produce. Same seed = same sequence, enabling reproducibility. For testing, you might use a fixed seed to reproduce bugs. For production, seeds should come from high-entropy sources. Never use predictable seeds (like timestamps) for security applications.

Examplerandom.seed(42) in Python produces reproducible "random" numbers

A measure of randomness or unpredictability in a system, used to seed random number generators.

High entropy means more randomness. Operating systems gather entropy from unpredictable sources: mouse movements, keyboard timing, disk I/O, network traffic. This entropy seeds the system's random number generator. Insufficient entropy can make random numbers predictable—a security risk.

Example/dev/random on Linux blocks until sufficient entropy is available

Cryptographically Secure Pseudo-Random Number Generator - a PRNG designed to be unpredictable for security applications.

CSPRNGs produce output that cannot be predicted even if an attacker knows the algorithm and some previous outputs. They're essential for generating encryption keys, session tokens, and passwords. Examples include Fortuna, CTR_DRBG, and /dev/urandom. Always use CSPRNGs for security-sensitive randomness.

Examplecrypto.getRandomValues() in browsers uses a CSPRNG

A probability distribution where every value in a range has an equal chance of being selected.

In uniform distribution, each outcome is equally likely. For random integers 1-10, each number has a 10% chance. Good random generators produce uniform distributions. Biased generators might favor certain values, which is problematic for games (unfair) and simulations (inaccurate). Test your random source for uniformity.

ExampleA fair die roll is uniformly distributed over 1-6

A computational technique using random sampling to obtain numerical results, often for complex problems.

Monte Carlo methods run many simulations with random inputs to estimate outcomes. Used for: risk analysis (financial modeling), physics simulations (particle behavior), optimization (finding best solutions), integration (estimating areas/volumes). The more iterations, the more accurate the estimate. Requires high-quality random numbers.

ExampleEstimating pi by randomly throwing darts at a square with inscribed circle