Notes on chance, odds and the design of games
Regulation accumulated in layers, each responding to a failure the previous layer could not address: first who may operate, then whether the money is safe, then whether the game does what it claims.
The earliest legal responses to commercial gambling were prohibitions, and they were largely unsuccessful in the narrow sense that the activity continued. What prohibition did achieve was to determine who conducted it. An activity that cannot be conducted openly is conducted by parties willing to operate outside the law, and disputes arising from it cannot be settled in court.
That is the recurring difficulty with prohibition as applied to wagering. The transaction is voluntary, small, easily concealed and leaves no aggrieved party willing to complain, so enforcement depends on detecting the activity rather than on receiving reports of it. Where prohibition was sustained it tended to survive as a rule about premises and promotion, since those are visible, rather than about the wager itself.
Licensing replaces the question of whether the activity is permitted with the question of who may conduct it and on what conditions. The change is structural rather than moral: a licensed operator has something to lose, which makes conditions enforceable, and a licence can be made to carry obligations that no criminal prohibition could.
The obligations that arrived first were about money rather than about games, because that is where the visible failures had been. An operator collects stakes long before it must pay prizes, so it holds funds belonging in substance to other people. The lottery failures described in the history category are the classic case. Requirements to hold player funds separately from operating funds, to maintain reserves against outstanding liabilities, and to demonstrate the source of the operator's own capital all follow from that single observation.
Suitability requirements grew from the same root. If a licence is the instrument of control, then the question of who holds it matters, and licensing regimes accumulated processes for examining ownership, controlling interests and changes of control. These are unglamorous provisions that make up the bulk of most licensing frameworks.
Money rules do not answer the question raised at the end of the entry on reel machines: whether the game behaves as its paperwork says. A wheel can be inspected and a die can be weighed, but a machine whose odds live in an internal weighted table cannot be checked by observation, and no amount of financial supervision reveals a paytable that does not match the probabilities behind it.
The response was technical certification. Independent laboratories, separate from both the operator and the manufacturer, examine the submitted game and confirm that the random selection is uniform and independent, that the internal tables produce the return the documentation claims, and that the version approved is the version deployed. Change control matters as much as the original test, since a certified game that can be silently modified in the field has not been certified in any useful sense.
This layer is the point at which regulation began to require mathematics rather than paperwork, and it is why regulatory technical standards read like specifications: distribution requirements for the generator, seeding and cycling rules, tolerances for observed return against theoretical return, and rules about what the display may imply about the outcome.
Testing establishes that a return figure is accurate but does not necessarily tell anyone what it is. Disclosure requirements close that gap by obliging the return, the rules and the paytable to be available to the person deciding whether to play, on the ordinary consumer-protection principle that a price should be visible before a purchase rather than after it.
Disclosure has spread unevenly and remains the least settled of the four layers. It is easy to state for a machine game with a single published figure and much harder for products whose margin depends on a distribution of stakes, such as the balanced book described in the bookmaking entry, where no single number describes what any individual participant faces.
The sequence is not arbitrary. Each layer addresses a failure that the previous one made visible without being able to fix. Prohibition made the activity clandestine, which made the handling of money unsupervisable; licensing made the money supervisable but said nothing about the games; testing made the games verifiable but left the figures unpublished; disclosure publishes them. A regime that skipped a layer generally acquired it later, after encountering the failure that layer exists to prevent.
Reading the history this way also explains why regimes in different places look similar despite having no common source. They are responses to the same small set of structural features: money held before it is owed, odds that cannot be inspected from outside, and a margin that compounds with volume. Those features belong to the arithmetic set out in the first category of this board, and any regime dealing with commercial games of chance eventually has to deal with them.
All times on this board are UTC. Dates shown are the date an entry was last revised. Bet On Good.