Skip to content

Types of Randomness

Randomness feels like a single, simple idea — unpredictable, chaotic, beyond control. But dig a little deeper and you'll find that randomness isn't one thing. It's a spectrum with two very different ends: true randomness and pseudo-randomness. Understanding the difference changes how you think about security, games, simulations, and even the dice you roll on game night.

True Randomness

True randomness comes from the physical world. It's produced by processes that are fundamentally unpredictable — so chaotic and complex that no amount of computing power can foresee the outcome.

Common physical sources include:

  • Radioactive decay — the moment an unstable atom decays is genuinely unpredictable, even in principle
  • Atmospheric noise — static picked up from the weather, lightning, and cosmic radiation
  • Thermal noise — the tiny jitter of electrons inside electronic components like resistors
  • Quantum effects — measurement outcomes at the quantum level defy deterministic prediction

A hardware device that harvests this kind of entropy is called a true random number generator (TRNG). Because the source is physical, the output isn't just practically unpredictable — it's fundamentally unpredictable. No pattern exists to find.

True randomness doesn't fake unpredictability. It borrows it directly from the universe.

In short: true randomness is unpredictable because the underlying process itself is unpredictable. You can't reverse-engineer it, no matter how much computing power you have.

Pseudo-Randomness

Pseudo-random numbers look random — and often fool even trained eyes — but they're actually produced by a deterministic algorithm. Feed the same algorithm the same starting value (called a seed), and you'll get the exact same sequence every time.

The most famous family of these algorithms is the linear congruential generator (LCG), which produces numbers using a simple formula:

  • Start with a seed value
  • Multiply it by a constant, add another constant, then take the result modulo a large number
  • Repeat — and the output dances around in a way that looks convincingly random

Because pseudo-random number generators (PRNGs) are just math, they have some useful properties:

  • They're fast — no physical hardware required
  • They're repeatable — great for testing and debugging
  • They're portable — they run anywhere

But they also have a fatal flaw: they're only as random as their seed. If the seed is predictable, the entire sequence is predictable. And even with a good seed, every PRNG is periodic — eventually, its sequence repeats.

In short: pseudo-randomness is a crafted illusion. It's a sequence that looks random but is secretly following a recipe.

Why the Difference Matters

You might think, "If both look random, who cares?" The answer: it depends entirely on what you're doing.

Security & cryptography — Encryption keys, authentication tokens, and password salts must be unpredictable to an adversary. A pseudo-random sequence can be reconstructed if the attacker guesses the seed or observes enough output. That's why cryptographic systems demand true randomness — or at least randomness hardened by unpredictable physical entropy.

Games — A game that uses a weak PRNG with a visible seed can be exploited. If the shuffle order or loot drops are predictable, players will notice. Most games use PRNGs for performance, but they seed them from truly random sources so the outcome feels fresh every time.

Simulations & science — Monte Carlo simulations, weather models, and financial models rely on random sampling. Here, the priority is often statistical quality — the numbers must follow the right distribution and not repeat too soon — which good PRNGs can deliver perfectly well. Repeatability is actually a bonus: you can rerun the same experiment with the same seed.

Everyday tools — Random acts of kindness like "shuffle my playlist" or "surprise me with a restaurant" sit on the casual end. Pseudo-randomness is more than good enough. The stakes are low, and nobody is trying to predict your commute playlist.

The rule of thumb: use true randomness when the stakes are high and secrecy matters. Use pseudo-randomness when speed, repeatability, or simplicity matters more.

Everyday Examples

You've experienced both kinds of randomness without thinking about it. Here's where each shows up in daily life:

  • Dice — a well-rolled die is a beautiful little source of physical chaos: the bouncing, air resistance, and friction of the table make the outcome genuinely hard to predict
  • Coin flips — the tumble of a coin in the air is sensitive to tiny variations in force and air currents, pushing it toward true randomness
  • Card shuffling — a careful human shuffle mixes cards through physical entropy; a digital "shuffle" in an app is almost certainly a PRNG at work
  • Your phone's lock screen — that "swipe to unlock" pattern generation mixes touch timing and movement into a seed
  • YouTube or Spotify shuffle — pseudo-random sequences, often seeded with the current time, because nobody wants a shuffle that repeats the same three songs forever

The beautiful thing is that both kinds of randomness are useful. True randomness is the real deal — honest unpredictability borrowed from physics. Pseudo-randomness is the clever illusion that powers most of the digital world at lightning speed. Knowing which one you're dealing with is the first step to using randomness well.

Dig Deeper

  • How to Generate Randomness — practical ways to produce both kinds, from hardware to software
  • Uses — where randomness quietly powers the things you rely on every day
  • FAQ — answers to the most common questions about chance, probability, and chaos