The maximum win is not a limit, it is a design

A game does not pay less than its maximum win because a rule says so - it never contained a prize larger than that. This page explains what a maximum-win multiple is, how it enters the published return, what it costs to reach, and why two lobbies can list the same title with different figures.

Open the partner account Start with the price you were shown

In the rules: a maximum payout, a cap per day, a market liability limitPaid as priced: everything below the ceilingPaid at the cap: everything the ceiling was drawn through

Direct answerA maximum win is the largest multiple of the stake a game's mathematics can return, expressed as a figure like 5,000×. It is part of the game's design rather than a clause: prizes above it do not exist in the build, and the return figure the game publishes is calculated with it already applied. It is reached rarely, by a specific combination of outcomes, and the expected cost of reaching it is normally greater than the ceiling itself.

Shape: a multiple of stakeHome: the game's own mathsPublished return: calculated with the capReachable by: one exact combination

What the figure is made of

part 1

the top prize line One outcome in the maths. The largest prize on the game's paytable, usually a full screen of the highest symbol in a bonus feature. Its size is the number divided by the stake, and it has a probability attached to it like every other outcome.

part 2

the multiplier ladder How small prizes become the big one. A win is compounded by multipliers during a feature. The maximum is not one spin's prize; it is the compounded total of an entire bonus sequence, which is why the probability is measured in the hundreds of thousands of spins.

part 3

the truncation What the cap removes. A design without the ceiling would place a little more probability on the very top of the range. Adding the cap removes that weight, so the game's published return is a percentage point or two below the return of the same machine before the ceiling was drawn.

Worked example - the same maths with and without the ceiling (illustrative) Top prize: 25,000× the stake, at a probability of 1 in 250,000 spins. Contribution to the game's return: 25,000 ÷ 250,000 = 0.100, or 10.0 percentage points. Same prize line capped at 5,000×: 5,000 ÷ 250,000 = 0.020, or 2.0 percentage points. Difference: 8.0 percentage points of return that the capped build does not contain. The published return of a game already includes its own ceiling - nothing is being withheld from it. The point of the arithmetic is the opposite one: when two games advertise different maximum wins, they are describing two different return distributions, and a reader comparing headline percentages is comparing two ceilings as much as two sets of maths.

What it costs to reach

The number most worth working out on this page is not the ceiling, it is the price of the chance at it.

In spins. A 1-in-250,000 top prize, at a brisk 600 spins an hour, is about 417 hours of continuous play - a little over seventeen days without stopping - for one expected hit.
In money. Over 250,000 spins of 1.00 each, a game with a 4% edge costs an expected 10,000.00. The ceiling in the same example pays 5,000.00.
The comparison that follows. The expected cost of reaching the maximum exceeds the maximum. That is not a quirk of the example: it is the ordinary relationship between a ceiling and the probability attached to it, and it is why a maximum win is a marketing figure and not an expectation.

Why the same title pays differently twice

Because a title is a piece of software an operator configures rather than a fixed object. Different maximum-win figures on the same name are different builds or different settings of the same build, each with its own published return. The game supply desk covers the chain that produces those configurations - which studio, which platform, which approval. The consequence here is narrow and practical: the maximum win a reader saw in a lobby is a property of the version that lobby serves, and it can differ from the figure in the game's own help panel, which is the version-specific one.

Four things the maximum win is not

Not a limit on a session. It caps one game round's return, not a reader's day, week or account. A day-level ceiling is the clause-side cap covered by the previous page.
Not a target. Nothing in the maths makes a big win more likely after a long run of small ones. The series' game-maths desk covers why, and it is the page to read instead of any of this if the question is about return percentages themselves.
Not a promise about the top of the range. A 5,000× ceiling means 5,000× is the most the build can pay, not that it is available. Most sessions return a fraction of the stake and a minority return a multiple of it.
Not negotiable. Unlike a clause, a design has no version a reader can appeal to. The number in the help panel is what the reader is playing.

What this desk will not do with a maximum win. It will not rank games by their ceiling, and it will not treat a large maximum as a reason to play one. A larger ceiling usually means a lower probability attached to the top of the range and a distribution with more of its weight at the bottom - which is a description of a game, not a recommendation about it. The only claim here is structural: the number exists, it is designed in, it is published with the game rather than in the operator's rules, and its expected cost is ordinarily larger than itself.