The question behind the question
When someone asks who has studied roulette, they are usually not looking for a bibliography. They want to know whether anyone has found something the rest of us missed: a bias, a pattern, a method that turns a spinning wheel into a source of income. The honest answer is that roulette is one of the most studied gambling games in history, and the conclusion has been remarkably stable across centuries of serious effort.
Mathematicians, physicists, statisticians, computer scientists and professional gamblers have all taken runs at the wheel. Some studied it to understand probability. Some studied it to beat it. A few succeeded for short periods under specific physical conditions. None of that changes what the wheel does to an ordinary bet placed under modern conditions.
What the early mathematicians established
The first serious mathematical treatments of roulette appeared in the 18th and 19th centuries, when probability theory itself was being formalised. The wheel was a natural test case: a finite set of equally likely outcomes, a clear payout structure, and a house advantage that could be calculated exactly rather than estimated.
What those early studies showed is that roulette is a game of independent trials. Each spin is unaffected by the spins before it. The wheel has no memory, and neither does the ball. A number that has not appeared for fifty spins is no more likely to appear on the fifty-first than any other number. This is not a matter of opinion or casino policy; it follows from the physical structure of the game and the mathematics of independent events.
How the edge is built into every bet
The house edge in roulette comes from a simple mismatch: the payout for a winning bet is slightly smaller than the true odds of that bet winning. The wheel itself is the reference point.
A European wheel has 37 pockets and a single zero. A straight-up bet on one number pays 35 to 1. If the payouts matched the true odds, a winning straight-up bet would pay 36 to 1, because there are 36 losing numbers and one winning number. The missing unit is the house edge. An American wheel has 38 pockets with a double zero added, and the same 35-to-1 payout applies, which widens the gap further.
The edge is not hidden in the spin. It is printed in the payout table. Every bet on the layout carries its own edge, and the edge is applied to every unit wagered, win or lose.
RTP and house edge under one rule set
Return to Player (RTP) and house edge are two ways of describing the same relationship. RTP is a theoretical percentage of total wagered money a game returns to players over a very large number of spins or hands. It is a long-run statistical average, not a per-session guarantee. House edge is the mathematical advantage the operator holds on a given game, expressed as a percentage of each wager. House edge and RTP are complements: they sum to 100% for the same game.
This complement only holds for the same bet under the same rule set. A European wheel and an American wheel are different rule sets. A straight-up bet and an even-money bet on the same wheel carry the same edge in roulette, which is unusual among casino games, but the edge itself is a property of the wheel layout and payout table, not of the player's mood or system.
What the physicists looked for
Physicists have studied roulette from a different angle: the mechanics of the spin. The ball, the wheel, the deflectors, the rotor speed and the initial conditions all follow deterministic physical laws. In principle, if you knew the exact position and velocity of the ball and wheel at release, you could predict where the ball would land.
In practice, this is extraordinarily difficult. The system is chaotic in the technical sense: tiny differences in initial conditions produce large differences in outcome. A fraction of a millimetre in release position or a fraction of a degree in rotor speed changes the result completely. Studies that have attempted prediction have generally concluded that the required measurement precision exceeds what a person can achieve at the table, and modern wheels are designed to reduce predictable bias further.
There is a separate line of study around biased wheels: physical imperfections that make some numbers appear more often than chance would predict. This was a real phenomenon in the past, and some players profited from it by recording thousands of spins and betting on the biased numbers. Modern wheels are manufactured to tighter tolerances, rotated and serviced regularly, and monitored electronically. A bias large enough to overcome the house edge is rare, short-lived, and not something a player can rely on.
Why short sessions look nothing like the long run
The house edge is a long-run average. A single session is a short-run sample, and short-run samples are noisy. This is the single most important thing to understand about roulette, and it is the reason the game feels beatable even when it is not.
Suppose, as an example, a player places 100 even-money bets of one unit each on a European wheel. The expected loss from the house edge is a little over two and a half units over that session. But the actual result will rarely land exactly on the expectation. The player might finish ahead by several units, or behind by several units. The variance swamps the edge over short horizons. Over 100,000 bets, the edge dominates and the result converges toward the expected loss. The player who wins a session has not beaten the wheel; they have experienced a favourable draw from a distribution whose centre is negative.
This is why progression systems feel effective. A player who doubles after losses will often recover a small loss and finish a session ahead. The system works often enough to feel reliable. What it does not do is change the expected value of the bets placed.
What decisions per hour do to total exposure
Roulette moves quickly. A player at a busy table might place forty or fifty bets an hour; an online player can place far more. The house edge applies to every unit wagered, so the total expected loss scales with total action, not with time spent at the table in any simple way.
If a game returns a given RTP, the expected loss on any amount of action is the complement of that RTP applied to the total wagered. A player who wagers 1,000 units over an evening has more expected loss than a player who wagers 200 units, even if both play the same wheel and the same bets. The edge is a tax on turnover. Slowing down, betting fewer units per spin, or taking breaks does not change the edge, but it reduces the total amount the edge is applied to.
This is the one variable a player genuinely controls: how much total money passes across the layout. Bet selection in roulette does not change the edge the way it does in craps, where different bet types on the same table carry materially different house edges. In roulette, the main choice is which wheel to play, because the extra pocket on the American layout increases the house edge on every bet placed there.
Why progression systems redistribute outcomes
A progression system is a rule for changing bet size based on previous results. The most famous is the martingale: double after every loss, return to the base bet after a win. The appeal is that a win eventually recovers all previous losses plus one unit of profit. The problem is that the bets required to keep the progression alive grow quickly, and every table has a maximum bet.
What a progression does is change the shape of the outcome distribution. It turns many small wins into occasional large losses. The expected loss remains the product of the house edge and the total amount wagered, because the edge applies to every unit regardless of when it was bet or how the bet size was chosen. No betting pattern changes the built-in edge of a negative-expectation game. A progression is a way of choosing when to lose, not whether to lose.
What the studies agree on
Across centuries of mathematical, physical and practical study, the conclusions about roulette have stayed consistent. The wheel produces independent trials. The payouts are set below the true odds. The edge is applied to every bet. Variance dominates the short run and the edge dominates the long run. No betting pattern removes the edge, and no player can reliably predict where the ball will land under modern conditions.
The people who have studied roulette most carefully are not the ones selling systems. They are the ones who describe the game accurately: as a negative-expectation game with a known, calculable edge, where short-term results are noisy and long-term results are predictable. That is not a secret. It is the most thoroughly documented fact in gambling mathematics.
FAQ
Has anyone ever reliably beaten roulette?
There are documented cases of players profiting from biased wheels in the past, and of teams using prediction devices with some short-term success. Under modern wheel maintenance and monitoring, neither approach is a reliable source of profit. The house edge remains in place on every bet.
Does the house edge apply to every roulette bet equally?
On a given wheel, the edge is the same across the standard bets because the payout shortfall is built into the payout table uniformly. The main difference in edge comes from the wheel itself: the American layout with its extra pocket carries a higher edge than the European layout.
Why do people still win at roulette if the edge is against them?
Because the edge is small relative to the variance of a single session. A player can finish ahead over dozens or hundreds of spins through ordinary luck. The edge expresses itself over very large numbers of bets, where the accumulated expected loss dominates the random fluctuation.
If you or someone you know is struggling with gambling, help is available. Contact the National Problem Gambling Helpline at the helpline printed in the footer below for confidential support, or visit the National Council on Problem Gambling website for resources.
Play responsibly
Gambling should be treated as paid entertainment, never as a way to earn income or recover losses. 1-800-MY-RESET (1-800-697-3738)
18+ or 21+ depending on where you are; follow the minimum age that applies to you.
This article is general information about how these mechanics work. It is not legal advice and not a recommendation to gamble or to use any particular operator. Availability and legality differ by jurisdiction — check the rules that apply where you are.