Slot Mathematics Explained: RNG, Hit Rates & Expected Value

Every Slot Is a Mathematical Model

Every slot machine is a mathematical model. The colourful graphics, sound effects, and animations are a presentation layer wrapped around a probability engine. This guide explains what that engine does — and why no presentation, theme, or marketing claim can change the underlying maths.

Random Number Generation

Every outcome is determined by a Pseudo-Random Number Generator (PRNG). The PRNG produces a sequence of numbers that are statistically indistinguishable from true randomness. When you press spin, the PRNG generates a number. That number maps to a specific combination of symbols on the reels. The mapping is defined by the game's math model — a lookup table that assigns probability weights to every possible outcome.

The PRNG runs continuously, generating thousands of numbers per second regardless of whether anyone is playing. The exact moment you press spin determines which number in the sequence is used. This is why each spin is independent — the previous outcome has no influence on the next, because the PRNG has already moved through millions of numbers between your spins.

Probability and Weighting

A physical reel with 20 symbols has a 1/20 chance of landing on any given symbol. Virtual reels are different. Each 'stop' on the virtual reel can be weighted — a cherry might occupy 5 virtual stops while a wild occupies 1. A 5-reel slot with 100 virtual stops per reel has 100^5 = 10 billion possible combinations. The game designer assigns payouts to winning combinations and adjusts the virtual reel weights until the overall return matches the target RTP.

Hit Frequency

This is how often any spin produces a win (including sub-stake wins). A slot with 30% hit frequency produces a winning combination on roughly 3 out of 10 spins. High hit frequency does not mean good RTP — a game can produce frequent tiny wins (e.g., 0.10x stake) and still have a high house edge because most 'wins' return less than the bet. See RTP vs Volatility for how hit frequency relates to bankroll experience.

Expected Value (EV)

The expected value of a single spin is: EV = Stake × (RTP - 1). At £1 stake on a 96% game: EV = £1 × (0.96 - 1) = -£0.04. You expect to lose 4p per spin on average. Over 1,000 spins: expected loss = £40. This is the mathematical truth that no strategy can overcome — the house edge ensures negative expected value on every spin. Use the RTP Calculator and Session EV Calculator to model your own scenarios.

Variance and Standard Deviation

RTP is a long-run average. In the short term, your actual results will deviate from the expected value. This deviation is variance. A low-volatility game has small variance — your session results cluster closely around the expected loss. A high-volatility game has large variance — individual sessions can produce results far above or below expectation. The standard deviation of a slot's returns determines how likely you are to experience a result significantly different from the RTP over a given number of spins.

The Law of Large Numbers

Over infinite play, your observed return converges on the deployed RTP. Over 100 spins, your observed return might range from 50% to 150%. Over 10,000 spins, it narrows to perhaps 90-102%. Over 1,000,000 spins, it converges to within 0.5% of the true RTP. This is why 'testing' a slot's RTP through personal play is futile — you cannot play enough spins in a lifetime to reliably measure it.

The House Edge Is Absolute

No betting system, strategy, or pattern can overcome the mathematical house edge. Martingale (doubling bets after losses) doesn't work because it requires infinite bankroll and there are table/stake limits. Timing spins doesn't work because the PRNG runs continuously. 'Hot' and 'cold' detection doesn't work because each spin is independent. The only player choice that affects long-run cost is selecting which game and which casino — i.e., choosing the deployed RTP. See Slot myths debunked for the full list of strategies that don't work and why.