Advanced: The Arithmetic of a Decaying Balance and a Fixed Target

This page contains no technique. It contains the arithmetic that explains why the conditions behave as they do, and one conclusion that follows from it: in this market the best offers are usually the smallest.

The two quantities

A requirement is a fixed target: winnings times the multiple. It does not move.

The balance generating that turnover decays geometrically. Each round returns some fraction of its stake on average — a game at 96% keeps 4p per pound — so after n rounds the expected balance is the original times 0.96 to the power n, while the turnover accumulated is the sum of the balances staked along the way.

That mismatch, fixed target against geometric decay, is the whole of it.

Working it through

Start with £12 and stake the whole balance each round at 96%.

Round one: £12 staked, £11.52 remains, £12 of turnover recorded. Round five: about £9.83 remains, £53 of turnover. Round ten: about £8.05 remains, £98 of turnover. Round twenty: about £5.40 remains, £163 of turnover. Round forty: about £2.42 remains, £238 of turnover.

So a ten-times requirement on that £12 win — £120 — is met around round twelve with most of the money intact. A twenty-times requirement, £240, needs about forty rounds and finishes with roughly a fifth left. A forty-times requirement, £480, is not reached on the average path at all: the balance is functionally gone before the target.

Why the averages flatter the outcome

Everything above uses the expected return, and real play does not follow it. Slot outcomes are high-variance: the balance jumps and drops rather than declining smoothly, and it reaches zero far more often than the average path suggests.

That asymmetry matters because zero absorbs. A balance at zero does not recover however good the next spin would have been, so every early bad run truncates a path that the average calculation continues. The consequence is that the average overstates how far a real balance gets — the arithmetic above is the optimistic case, not the expectation.

Where RTP enters, and how little it changes

A higher-returning game decays more slowly, so it generates more turnover before running out. Moving from 92% to 96% roughly halves the loss per round and meaningfully extends the run.

It does not change whether the average path reaches a forty-times target. It changes how far it gets, and on requirements of ten or twenty it genuinely improves the odds of finishing. This is the only sense in which game selection affects a bonus, and it is a smaller effect than the difference between two multiples.

Volatility, on the other hand, changes the shape

Two games with identical RTP and different volatility produce very different distributions over fifty spins. The high-variance title reaches a large balance more often and busts more often. The low-variance title grinds.

Which is preferable is decided entirely by the terms. With no requirement, anything won is money, so the distribution with the fatter tail is better — the upside is realisable. With a steep requirement, the grinder survives longer and gets further through the turnover. The same game is the right and wrong choice depending on one number elsewhere on the page.

The expected-value statement, made honestly

On a wager-free offer with no cap, the expected value of the spins is positive, because nothing was paid in and whatever returns is money. It is a small positive number — a few pounds of play times a return below one hundred per cent — and it is the entire mathematical case for preferring wager-free offers.

Attach a requirement and that changes sign in practice, because the probability of conversion collapses as the multiple rises. Attach a cap and the upper tail is truncated, which removes precisely the outcomes that carried the value.

Why systems do not help

A staking progression changes the shape of the distribution and leaves the mean where it was. Any scheme that reaches the target more often busts more often by exactly the amount that keeps the expectation constant. There is no ordering of bets on a negative-expectation game that produces a positive one, and the arithmetic above is indifferent to the order in which the rounds occur.