AI-assisted content
Drafted with AI assistance and reviewed by Lei Abelai. Odds shown are for standard single-zero and double-zero wheels; multiplier variants change payouts, so check the game rules panel. Full AI disclosure.
Roulette has a purity the other games lack. There are no decisions after the bet, no rule variations to memorise, no strategy chart. There is a wheel, a set of payouts, and a mismatch between them that produces the house edge. Everything else is presentation.
That mismatch is easy to see once it's pointed out, and once you've seen it you can never quite unsee it.
1. Three wheels, three prices
| Wheel | Pockets | House edge | Cost per ₱1,000 | Verdict |
|---|---|---|---|---|
| French (with la partage) | 37 (single zero) | 1.35% on even-money bets | ₱14 | Best if offered |
| European | 37 (single zero) | 2.70% | ₱27 | Standard choice |
| American | 38 (0 and 00) | 5.26% | ₱53 | Avoid |
That's the whole decision, and it's the most consequential one you'll make at the table. Playing American roulette instead of European costs you almost double for an identical experience. The extra green pocket is the entire difference.
La partage is worth understanding because it's genuinely good value: on a French wheel, if you've placed an even-money bet (red/black, odd/even, high/low) and the ball lands on zero, you get half your stake back rather than losing it all. That single rule halves the edge on those bets from 2.70% to 1.35%.
2. Why every bet carries the same edge
Here's the mechanism, using a European wheel. There are 37 pockets. A straight-up bet on a single number wins once in 37 spins on average — so "fair" odds would be 36:1.
The casino pays 35:1.
That one-unit shortfall, spread across 37 outcomes, is 1/37 = 2.70%. Now run the same calculation for red/black: 18 winning pockets out of 37, paying 1:1 when fair odds would be 19:18. The shortfall works out to exactly 2.70% again.
Every bet on a European layout is underpaid by precisely the same proportion. The zero is the edge, and it taxes all bets equally.
This has a liberating implication: there is no "smart" roulette bet. Betting a single number is not braver or worse value than betting red. It's the same price with a different distribution.
3. The complete odds table
All figures for a single-zero European wheel.
| Bet | Covers | Payout | Win chance | House edge |
|---|---|---|---|---|
| Straight up | 1 number | 35:1 | 2.70% | 2.70% |
| Split | 2 numbers | 17:1 | 5.41% | 2.70% |
| Street | 3 numbers | 11:1 | 8.11% | 2.70% |
| Corner | 4 numbers | 8:1 | 10.81% | 2.70% |
| Six line | 6 numbers | 5:1 | 16.22% | 2.70% |
| Column / Dozen | 12 numbers | 2:1 | 32.43% | 2.70% |
| Red / Black | 18 numbers | 1:1 | 48.65% | 2.70% |
| Odd / Even | 18 numbers | 1:1 | 48.65% | 2.70% |
| High / Low | 18 numbers | 1:1 | 48.65% | 2.70% |
Look at the last column. Nine different bets, wildly different payouts and probabilities, one identical price. That table is the single most useful thing to know about roulette.
Red/black is not a coin flip
It's 48.65%, not 50%. The green zero is neither red nor black, and that missing 1.35% on each side is where the entire house edge lives. Any system built on the assumption of an even-money proposition is built on a false premise.
4. Choosing bets for the ride, not the return
Since cost is fixed, bet selection is purely a question of what kind of session you want.
| If you want… | Bet | What ₱1,000 over 100 spins feels like |
|---|---|---|
| Long playing time, small swings | Even money (red/black, odd/even) | Roughly half the spins win. Balance drifts slowly downward. |
| Moderate swings, occasional lift | Dozens or columns | Wins about a third of the time at 2:1. Noticeable runs both ways. |
| Rare, large moments | Straight up or splits | Long dry stretches; a single hit returns 35× the chip. |
If your goal is to make a modest budget last an evening, even-money bets on a French wheel with la partage is the objectively cheapest way to sit at a roulette table. If your goal is the thrill of a straight-up hit, that costs the same per peso wagered — you'll just get there faster or not at all.
5. Lightning and multiplier roulette
Multiplier variants add randomly selected "lucky numbers" each round, paying up to 500× instead of 35:1. They're visually spectacular and enormously popular. The pricing deserves scrutiny.
The multipliers are funded by cutting the standard straight-up payout from 35:1 to 29:1. So on any spin where your number wins without being selected as a lucky number — the large majority of wins — you're paid substantially less than on a normal wheel.
Net effect: house edge on these variants typically lands somewhere around 3% to 5%+ depending on the specific game, compared to 2.70% on a standard European wheel. Even-money bets in these games often don't qualify for multipliers at all while still being subject to the game's structure.
They aren't a scam — the rules are published. But they are a more expensive product sold on the strength of a bigger headline number, and it's worth knowing which one you're buying.
6. Systems, wheel bias and dealer signature
Betting systems
Martingale, Fibonacci, d'Alembert, Labouchère — all fail for the same structural reason. A sequence of bets each carrying a 2.70% edge has a 2.70% edge in aggregate, regardless of the order or size in which they're placed. Progressions redistribute variance; they don't touch expectation. Table maximums exist specifically to cap the doubling that Martingale requires. Full explanation in why betting systems fail.
Wheel bias
Historically real. In the 19th and 20th centuries, physical wheels with manufacturing defects could favour certain pockets, and a few teams exploited it profitably. Today: modern wheels are precision-engineered and monitored statistically by the casino itself, which will spot a bias long before you do. In RNG roulette there is no physical wheel at all. Treat wheel bias as history, not method.
Dealer signature
The theory that a croupier develops a consistent release producing predictable landing zones. Live studios rotate dealers, vary release points, and spin against wheel rotation specifically to disrupt this. Any detectable pattern would need to survive thousands of observations to be distinguishable from noise, and you'd be funding that research at 2.70% per spin.
Tracking the board
Live tables display recent results, and players read them for "hot" and "cold" numbers. The wheel has no memory. A number appearing five times in an hour has exactly the same 1-in-37 chance on the next spin as one that hasn't appeared all day.
7. A sensible way to play
- Play French with la partage if available, otherwise European. Never American.
- Pick bets for the experience you want — cost is identical, so choose the variance you enjoy.
- Flat stakes. No progressions, no doubling, no "recovery" sequences.
- Check table minimums before joining. A minimum above 2% of your session budget is too high.
- Treat the results board as decoration.
- Set stop-loss and stop-win before the first spin and honour them.
The honest summary
Roulette charges 2.70% on a European wheel, uniformly, forever. You cannot arrange chips into an advantage, and no one ever has without a physically defective wheel. What you can control is which wheel you sit at, how much you stake, and how long you play — and those three levers are the whole game.
Why no system beats the wheel
Martingale, Fibonacci, Labouchère — the full explanation of why progressions cannot work, in one place.
21+ only · Gambling involves risk of loss · If it stops being fun, stop.