Seeded PRNG
A pseudo-random number generator initialized with a fixed seed, producing identical sequences on every run.
A pseudo-random number generator initialized with a fixed seed, producing identical sequences on every run.
A seeded PRNG (pseudo-random number generator) produces a deterministic sequence of numbers from a fixed starting value called the seed. Given the same seed, the same sequence is always produced. Given a different seed, a statistically independent sequence is produced. The outputs appear random to statistical tests but are fully reproducible. Common algorithms include Mulberry32, xoshiro256**, and LCG variants. The key property is reproducibility: seeded PRNGs make randomized systems testable because any run can be exactly reproduced by re-supplying the original seed.
How It Works
Initialize generator with a seed integer → call next() to get the next pseudo-random value → the same seed always produces the same sequence of values in the same order.
Example
Replacing all Math.random() calls in a campaign pipeline with a seeded Mulberry32 PRNG made 73 encoding variants fully reproducible. Any run with seed 42 produces the same 73 outputs on any machine. Snapshot testing became possible because the outputs stopped drifting.