A Tinker's Guide to Roulette Math and Smart Choices
Practical, math‑based explanations of roulette odds, house edge, hit frequency, variance, and sensible session choices for curious adult players.
Roulette looks simple: a ball, a wheel, and a stack of possible bets. Under the spectacle, though, are precise probabilities and a single, unbending truth: the house edge. This guide explains how that edge shows up in different bets, what hit frequency and variance mean for your session, and how to make practical choices that keep play interesting without pretending any strategy will beat the math.
The wheel and the basic probabilities
Modern roulette wheels commonly come in two main flavors: single‑zero (European-style, 37 pockets) and double‑zero (American-style, 38 pockets). On a single‑zero wheel any straight‑up number has a probability of 1/37 (≈2.70%); on a double‑zero wheel it’s 1/38 (≈2.63%).
Payouts are fixed by the table: a straight‑up pays 35 to 1, a split 17 to 1, a street 11 to 1, and even‑money bets (red/black, odd/even) pay 1 to 1. The important thing to remember: those payout ratios are set so the house keeps a small expected percentage of every bet.
House edge vs expected value vs hit frequency vs variance
-
House edge is the casino’s long‑run advantage expressed as a percentage of each bet. For single‑zero roulette it’s 1/37 ≈ 2.70%; for double‑zero it’s 2/38 ≈ 5.26%. This is the expected loss per unit staked, on average, over many spins.
-
Expected value (EV) is what you expect, on average, from a particular bet. If you bet $1 on a single number in single‑zero roulette: EV = 35*(1/37) + (-1)*(36/37) = -1/37 ≈ -$0.027. You lose about 2.7 cents per dollar in expectation — that’s the house edge at work.
-
Hit frequency is how often a bet wins. A straight‑up wins roughly 1/37 of the time; an even‑money bet wins about 18/37 ≈ 48.65% of the time on single‑zero. Hit frequency does not change the house edge — it only changes how often you see wins.
-
Variance is how wild the swings are. High‑payout, low‑probability bets (single numbers) have high variance: you’ll rarely win, but when you do, the payout is big. Even‑money bets have low variance: wins and losses are frequent and similar in size. Two bets can have the same EV but very different variance profiles; choose based on the kind of volatility you can tolerate.
Why the house edge is the same across bets
A useful fact: despite different payouts and win frequencies, the house edge is the same on most standard bets (in single‑zero roulette it’s 2.70%). That’s why a $1 straight‑up and a $1 red bet both lose, on average, the same proportion over time — even though their experiences at the table feel very different. Betting systems that change bet sizes or move between bets don’t change the underlying EV; they only change the path you take through variance.
Common strategies — what they actually accomplish
-
Martingale (double after loss): increases the chance of short winning runs but risks a catastrophic loss when you hit a run of losses or the table limit. It does not change EV; it simply concentrates risk.
-
Progressions (D’Alembert, Fibonacci, etc.): aim to smooth swings or recover losses slowly. They also do not change the house edge — they only change the distribution of wins and losses.
-
Flat betting: betting the same amount each spin minimizes surprises in bet sizing and makes long‑run EV easy to track.
In plain terms: no size or progression flips the math. Systems can be entertaining and help structure play, but they trade one form of risk for another rather than removing it.
Practical choices for a sensible session
-
Pick your volatility. If you like steady action, prefer even‑money bets; if you want a shot at a headline win, play singles or splits but accept bigger swings.
-
Size bets to your bankroll. A common sensible approach is to stake a small percentage of your session bankroll on each spin so a few bad runs don’t end your night.
-
Know what you’re paying. You can quickly compute EV per dollar staked: it’s essentially the negative house edge (e.g., about −$0.027 per $1 on single‑zero). That tells you how much, on average, you’ll lose per dollar over many spins.
Tink’s Toolchest can help you reference bet probabilities, compare expected values, or review a session’s outcomes afterward.
A final practical example
Imagine you play single‑zero, place $1 on a single number each spin for 100 spins. Expected loss ≈ 100 * $1 * 2.70% = $2.70. Variance means actual results will vary — you might lose less, more, or even hit a big win — but across many such 100‑spin sessions the average loss will cluster around that expected value.
Responsible play reality check
Roulette is a game of controlled risk and predictable math: enjoy the spectacle and your chosen style of play, but only wager money you can afford to lose and set limits before you sit down.