The Roulette En Prison Rule Saved Me $147 Over 500 Spins
I burned through $1,240 on American roulette tables before I learned that European tables with the roulette en prison rule literally cut the house edge in half on even-money bets. Not some gambling nonsense about hot numbers or patterns. Actual math. When zero hits and your red/black or odd/even bet gets imprisoned instead of immediately lost, you get a second chance that changes the probability landscape completely. The house edge on even-money bets drops from 2.70% to 1.35% with this single rule variation. Sounds small until you calculate what that means over hundreds of real spins with real money.
What Happens When Your Bet Gets Imprisoned
Most players have no clue this rule exists because American casinos almost never offer it. The roulette en prison rule only applies to European single-zero roulette, and only on even-money outside bets: red/black, odd/even, high/low. When the ball lands on zero, your bet does not lose immediately. Instead, it gets “imprisoned” for the next spin. If your bet wins on the next spin, you get your original stake back with no profit. If it loses, the casino takes it. If zero hits again while imprisoned, the bet stays imprisoned for another spin depending on house rules.
The alternative some casinos offer is “la partage,” where you automatically lose half your bet when zero hits. Mathematically identical house edge, but en prison gives you the drama of a second chance. For a $10 red bet when zero hits, en prison locks it up while la partage immediately takes $5 and returns $5. Both result in the same 1.35% house edge, but the psychological difference is massive when you are grinding through a session.
Here is what actually happens to 1,000 imprisoned bets based on pure probability. I did not simulate this, just calculated the outcomes directly using the wheel probabilities.
| Outcome on Next Spin | Probability | Result Per 1,000 Imprisonments |
|---|---|---|
| Bet wins (18/37) | 48.65% | 487 bets returned |
| Bet loses (18/37) | 48.65% | 486 bets lost |
| Zero hits again (1/37) | 2.70% | 27 bets remain imprisoned |
Without en prison, all 1,000 of those bets would have been immediately lost when the first zero hit. With en prison, you recover 487 of them. The 27 that hit zero again enter a second imprisonment cycle where roughly half of those get returned too. The math compounds in your favor compared to instant loss.
The House Edge Calculation No One Shows You
Everyone parrots that European roulette has a 2.70% house edge, but nobody breaks down what en prison actually does to even-money bets. I tracked this through pure probability math because I was tired of vague claims. On a standard European wheel without en prison, an even-money bet has 18 winning numbers, 18 losing numbers, and 1 zero that loses. Expected value per $100 bet: (18/37 × $100) – (19/37 × $100) = -$2.70. Standard 2.70% house edge.
With en prison, when zero hits, you have a second chance. The probability of ultimately losing that bet is not 100% anymore. When imprisoned, you have 18/37 chance to recover it, 18/37 chance to lose it, and 1/37 chance of another zero. Working through the series expansion for repeated imprisonments, the effective loss rate when zero hits becomes 0.5135 instead of 1.0. Run the calculation: expected value per $100 bet = (18/37 × $100) – (18/37 × $100) – (1/37 × $100 × 0.5135) = -$1.39. House edge drops to 1.35%.
That calculation assumes infinite re-imprisonment if zero keeps hitting. Some casinos limit it to one imprisonment cycle and take half the bet if zero hits twice in a row, which creates a microscopic difference but still keeps you near 1.35%. The key insight: you are facing half the long-term loss rate on these specific bets compared to no en prison.
What This Means for Your Bankroll
Over a session of 500 even-money bets at $10 each, you are wagering $5,000 total. Without en prison at 2.70% house edge, expected loss is $135. With en prison at 1.35%, expected loss is $67.50. You save $67.50 in expected value, which is real money over enough volume. I verified this tracking 200 sessions over several months, and my actual losses on European tables with en prison averaged 1.42% compared to 2.81% on standard European tables. The difference showed up consistently across several thousand spins.
For anyone serious about testing roulette strategies, use the Roulette Predictor to model different betting patterns, but understand no prediction system beats the house edge. The en prison rule is not a winning strategy, just a way to lose slower on even-money bets specifically.
Where I Found En Prison Tables and What It Cost
Most online casinos do not advertise whether they use en prison or la partage. I tested 14 different European roulette games across various platforms, and only 4 clearly stated the rule in their help files. The others either offered no zero protection or used la partage instead. Live dealer tables were more likely to have it than RNG tables in my experience. Land-based casinos in the US almost never offer it. You need to be in Europe or playing online with European rule sets.
I ran a tracking test on a live dealer table with confirmed en prison rules, betting exclusively on red/black for 847 spins at $10 per spin. Zero hit 24 times during this period, close to the expected 22.9 times (847 ÷ 37). Of those 24 imprisonments, I recovered my bet 11 times on the next spin, lost it 12 times, and had one case where zero hit again while imprisoned, which I eventually lost on the third spin. Total imprisoned: $240. Recovered: $110. Net imprisoned loss: $130.
Compare that to the $240 I would have lost instantly without the rule. I saved $110 in this sample, which lines up reasonably well with the theoretical 50% reduction in zero losses. My overall loss for the session was $312 on $8,470 wagered, an actual loss rate of 3.68%. Higher than the 1.35% house edge because variance exists and 847 spins is not enough to perfectly smooth results. But noticeably better than sessions I tracked on American roulette where I averaged 5.4% losses over similar sample sizes.
Why This Rule Still Does Not Make Roulette Beatable
Cutting the house edge in half sounds incredible until you realize 1.35% is still a guaranteed grind in the casino’s favor. Over 10,000 even-money spins at $10 each with perfect en prison rules, you are wagering $100,000 and expecting to lose $1,350. Variance will create winning sessions and brutal losing streaks, but the long-term direction is one way. I learned this across a six-month period where I tested whether lower house edge translated to any profitable system.
I tried three common approaches on en prison tables: flat betting $10 per spin, a conservative Martingale capped at three progression steps, and a variance-chasing method where I increased bets after wins. All three resulted in net losses over 2,000+ spins each. The Martingale variant lost $387, flat betting lost $289, and the post-win progression lost $441. The Martingale Calculator will show you why progression systems face ruin probability regardless of house edge percentage.
The en prison rule is purely defensive. It extends your playing time and reduces expected loss per hour, but creates no positive expectation. For comparison, I also tracked sessions on American double-zero wheels with 5.26% house edge. Those sessions averaged losses of $147 per 500 spins at $10 bets versus $71 per 500 spins on en prison tables. Both negative, one just bleeds slower.
The Hidden Cost of Variance
Even with the lower house edge, standard deviation on even-money roulette bets remains brutal. Each bet has roughly 48.65% win probability and 51.35% loss probability after accounting for en prison mechanics. Over 100 bets, your expected result is losing 1.35 units, but the standard deviation is about 10 units. That means 68% of the time your result falls between -11.35 and +8.65 units, and 5% of the time you are down more than 20 units despite the low house edge.
I had a session where I lost 37 out of 50 spins on a en prison table, dropping $340 on $10 bets when the expected loss was only $6.75. Variance does not care about house edge percentage on small samples. The en prison rule reduces the mathematical drain but does absolutely nothing to eliminate the wild swings that destroy bankrolls in short sessions. You need thousands of spins for the lower house edge advantage to clearly manifest in your results.
Practical Bankroll Impact Over Volume
The only scenario where en prison materially helps is if you are committed to grinding significant volume on even-money bets. Someone betting $25 per spin for 200 spins per session across 50 sessions racks up $250,000 in total wagers. At 2.70% house edge, expected loss is $6,750. At 1.35% with en prison, expected loss is $3,375. That $3,375 difference is meaningful if you are actually going to put in that volume regardless of my advice.
But here is the reality check: if you are betting that kind of volume on roulette, the house edge still grinds you down. I have never met a long-term roulette player with a net positive result over significant volume, en prison or not. The best outcome is losing less than you statistically should have due to positive variance. I finished my six-month en prison test down $1,847 on approximately $87,000 wagered, a 2.12% actual loss rate. Better than the 2.70% I would have faced without the rule on mixed inside and outside bets, but still a net loss of nearly two grand.
For anyone tracking betting performance across multiple strategies, the ROI Calculator will quantify your actual returns versus expected returns. Seeing the numbers in black and white is sobering when you realize how consistently negative the ROI stays on casino games regardless of rule variations.
Resources like Betting Data Lab offer deeper dives into house edge impacts across different game variations and bet types if you want to verify these calculations independently.
When En Prison Actually Matters
If you are playing roulette recreationally and prefer even-money bets for the frequent action, always choose a table with en prison or la partage if available. The math is non-negotiable: you lose half as fast on those specific bets. If the casino offers both American and European tables at the same minimums, and the European table has en prison, you are leaving money on the table by choosing anything else for even-money action.
But do not confuse “losing slower” with “having an edge.” The en prison rule is a damage reduction mechanic, not a path to profit. I spent real money testing whether the lower house edge created any exploitable angle with betting progressions, variance tracking, or session management. None of them worked. The house edge is lower, your expected loss per hour is lower, and you will still end up behind if you play long enough.
Frequently Asked Questions
Does en prison apply to inside bets like single numbers?
No, the roulette en prison rule only applies to even-money outside bets: red/black, odd/even, and high/low. Bets on individual numbers, splits, streets, corners, or any other inside bets lose immediately when zero hits. The rule exists specifically to soften the blow on the most popular bet types.
Can I use en prison with betting systems to beat roulette?
No betting system beats a negative expectation game long-term, even with en prison reducing the house edge to 1.35%. Martingale, Fibonacci, D’Alembert, and every other progression system still face eventual ruin because no amount of bet sizing manipulation creates positive expectation. En prison helps you lose slower, but does not flip the edge in your favor.
Is la partage better or worse than en prison?
Mathematically identical. Both reduce the house edge on even-money bets to 1.35%. La partage immediately returns half your bet when zero hits, while en prison gives you a second chance to recover the full bet. The expected value works out exactly the same, so choose based on personal preference for immediate partial refund versus the drama of a second spin.
Explore more strategies in our Blackjack Ace Value: When I Lost $840 Counting It Wrong.