Roulette Column and Dozen Bets: Same Edge, Different Risk

Published October 14, 2026 · 9 min read · By the OddsLighthouse Editorial Team

Most new roulette players learn the wheel through its simplest bets — red or black, odd or even, high or low — and stop there. One level up sit column and dozen bets, a pair of 2:1 payout wagers that cover exactly the same twelve numbers as each other, laid out two different ways on the table felt. They carry the identical house edge as every other inside and outside bet on a standard wheel, but the way that edge actually plays out across a session is genuinely different between the two, which is the part most quick explainers skip.

What column and dozen bets actually cover

A dozen bet covers twelve consecutive numbers as they're numbered on the felt: 1 through 12 (the "first dozen"), 13 through 24 (the "second dozen"), or 25 through 36 (the "third dozen"). A column bet also covers twelve numbers, but grouped by the table's physical columns rather than by numerical sequence — the numbers in a single vertical column on a standard layout are spaced three apart (1, 4, 7, 10, 13, 16, 19, 22, 25, 28, 31, 34 forms one column, for example). Both bet types exclude zero (and double zero on an American wheel), both pay 2:1 on a win, and both are placed on the outer edge of the number grid rather than directly on a number.

At a glance, that makes them look interchangeable — same count of numbers, same payout, same place on the table. The grouping logic underneath is where they actually diverge.

Why the house edge is identical to every other even-money and 2:1 bet

On a European single-zero wheel, both column and dozen bets carry the same 2.70% house edge as red/black, odd/even, and high/low. On an American double-zero wheel, both carry the same 5.26% edge as those same even-money bets. This isn't a coincidence specific to columns and dozens — on a standard wheel with no special rules like la partage or en prison, almost every standard bet on the layout carries the same house edge, because the edge comes entirely from the zero pockets not paying out on any outside bet, regardless of how the remaining 36 numbers get grouped or how the payout is structured to compensate.

The payout ratio is specifically calibrated to each bet's coverage so the long-run edge comes out the same across the board. A dozen bet pays less per number covered than a single-number bet, but it also wins far more often — the math is built to converge on an identical edge no matter which square on the felt you choose.

The real difference: how the same 12 numbers get grouped

Since both bets carry the same edge and the same payout, the only place they can meaningfully differ is in which specific numbers land together, and how that grouping overlaps with other bets a player might also have on the table. Dozens group numbers sequentially, so a dozen bet lines up cleanly with how most players mentally picture "low," "middle," and "high" numbers on the wheel. Columns group numbers by their position in the physical grid, which has no inherent numerical pattern a player would notice just by looking at recent spin results — two numbers in the same column can be as far apart as 1 and 34.

This distinction mostly matters for one practical reason: covering multiple bets at once. Betting two different dozens together covers 24 of the non-zero numbers, same as betting two different columns together. But because columns and dozens slice the wheel along two different axes, combining one column bet with one dozen bet covers a different, overlapping-but-not-identical set of numbers than combining two dozens or two columns would — useful to understand if you're deliberately trying to cover a specific portion of the layout rather than just picking bets at random.

A worked comparison across 100 spins

Say a player bets $10 on the first dozen (1–12) for 100 spins on a European wheel, and a second player bets $10 on the left column for 100 spins on the same wheel. Both bets cover 12 numbers out of 37, so both have roughly a 32.4% chance of winning any individual spin and an expected loss of about 2.70% of total action over the long run — in this case, roughly $27 in expected losses on $1,000 total wagered, for both players equally.

Where the two sessions could look different in practice isn't the math — it's which 12 numbers happen to hit. If the wheel runs unusually hot on numbers in the 13–24 range during that stretch, the dozen bettor on 1–12 is shut out of every one of those spins, while the column bettor's result depends entirely on how the hot numbers happen to be distributed across the three columns, which bears no relationship to the numerical range at all. Over a large enough sample, this evens out to the identical expected result for both bets — but it's a real source of session-to-session variance that has nothing to do with either bet being better or worse.

European vs. American wheels change the math the same way for both

As covered in our roulette basics guide, the core difference between European and American wheels is the extra 00 pocket on the American layout, which doesn't belong to any column, dozen, or other outside bet. That pocket is what pushes the house edge from 2.70% to 5.26% — and it affects column and dozen bets identically to every other outside bet, since none of them include either zero pocket in their coverage. There's no version of a column or dozen bet on either wheel type that carries a different edge than the other outside bets on that same wheel.

Why some players combine column and dozen bets

A common approach is betting one dozen and a column that doesn't fully overlap with it, aiming to cover more of the layout than either bet alone while still collecting a 2:1 payout on whichever bet hits. This doesn't change the underlying math — the combined bet still carries the same 2.70% (or 5.26%) edge on each dollar wagered, since you're simply placing two separately-priced bets with identical edges side by side. What it does change is the shape of a session: more frequent small wins, interrupted by stretches where neither bet lands, rather than the wider swings that come from concentrating action on a single straight-up number or a single dozen alone. That's a preference about variance, not a way to reduce the house's actual cut.

This is the same principle covered in our broader piece on casino bankroll management — combining bets changes how smoothly your bankroll moves through a session, not how much of it the casino expects to keep over time.

Common misconceptions about column and dozen betting

The most common misconception is that because columns group numbers in a less obviously patterned way than dozens, they're somehow "more random" or carry better odds. They don't — the grouping method has zero effect on the math once zero pockets are accounted for. A second misconception is that tracking which dozen or column has hit most recently reveals anything useful about the next spin. Each spin is an independent event; a dozen that's hit five times in a row has exactly the same probability of hitting on the sixth spin as it did on the first, regardless of how striking the pattern looks on a results board. A third is assuming a 2:1 payout bet is automatically "safer" than a single-number bet simply because it wins more often — safer in the sense of smaller swings, yes, but not safer in terms of the percentage of money the house expects to keep, which is identical either way.