Roulette Online PRO
Walkthroughs · Walkthrough · 8 min

Walkthrough: A Martingale Session

Ten dollars on red, doubled after every loss, at a $500 table. Spin by spin, including the spin that ends it.

Short answer

A simulated Martingale session from a $1,000 bankroll. Forty-one small wins, then six losses in a row and a refused seventh bet. The session ends down $220 after being up $410, which is what the system does.

base
$10 on red
table limit
$500
losses to break
6
p(6 losses in a row)
1.8% per attempt

The plan

Bet $10 on redbetRedEven-money bet on the 18 red numbers. French: rouge.Read full →. After a loss, double. After a win, back to $10. Table limittableTable limitMinimum and maximum bet allowed. The maximum is what stops doubling progressions.Read full → $500, so the stakes are 10, 20, 40, 80, 160, 320 and then a $640 bet that the table will refuse. Six consecutive losses break the system; each attempt has (19/37)⁶ ≈ 1.8% of doing so, and a 100-spin session contains about 50 attempts. Expected: roughly a 60% chance of breaking within 100 spins, and a loss of $630 when we do.

Roulette Table Limits

Spins 1–30

Ordinary MartingalestrategyMartingaleProgression that doubles the bet after each loss. See the Strategy section for what it does and does not do.Read full →: strings of one to three losses, each ended by a win that nets +$10. By spin 30 we have had 17 wins, each closing a cycle, and are up $170. The largest stake so far was $80 after three losses. It feels like a machine for printing $10 bills.

cyclesequencestakesnet
1W10+$10
2L W10, 20+$10
3L L W10, 20, 40+$10
4W10+$10
5L L L W10, 20, 40, 80+$10
6–17mixed, max 3 losses≤ 80+$120
after 30 spins17 cycles+$170

Spins 31–60

More of the same, with one four-loss cycle (stakes to $160, then a win) that briefly put $310 on the table and felt like courage. After 60 spins: 41 cycles completed, up $410, and the bankrollmoneyBankrollThe total amount set aside for play. Bankroll management is deciding how much of it to risk per bet and per session.Read full →’s low point was $1,040. The histogram of every Martingale session looks like this for most of its length.

Spins 61–66

L L L L L L. Stakes 10, 20, 40, 80, 160, 320: $630 lost. The next bet would be $640; the limit is $500. The system is broken. We can bet $500 (and if it wins, still be $130 down on the cycle) or reset to $10 and accept the loss. Six blacksbetBlackEven-money bet on the 18 black numbers. French: noir.Read full → or zeroswheelZeroThe single green pocket that no even-money, dozen or column bet covers; the source of the house edge. Also: single zero, 0.Read full → in a row has probabilitymathProbabilityChance of an outcome, from 0 to 1 or 0% to 100%. One number: 1/37 = 2.70%.Read full → 1.8% per cycle. We had run about 45 cycles. The chance of reaching this point somewhere in 45 cycles was about 55%. It was not bad luck; it was the median.

This session and nine like itresult vs start
+$2,450$0−$1,000+$1,730−$5000125250375500spinsresult
1,000 sessions × 500 spins · violet p10/p25/p75/p90, light blue median, 83 faint samples · move over the chart for quantiles at a spin
  • p10 −$1,000
  • p25 −$1,000
  • median −$1,000
  • p75 +$790
  • p90 +$1,690
1,000 sessions × 500 spins · y-axis clipped to the 1st–99th percentile (extremes −$1,000 to +$2,630) · 500 spins · $10 base · even money
Most lines creep upward in $10 steps. The ones that fall, fall $630 at once. Over many sessions the mean is well below a flat bettor’s.

The accounts

Up $410 after 60 spins, down $220 after 66. Total wagered: about $2,100 over 66 spins, against $660 for a flat $10 bettor. Expected lossmathExpected lossExpected value times total amount wagered; the average amount lost over a session or a period.Read full → on $2,100 of turnovermoneyTurnoverThe total amount wagered over a session or period, counting re-bet winnings. Also: handle, action.Read full → is $57; our actual −$220 is about a third of a standard deviationmathStandard deviationSquare root of variance; the typical distance of a result from the expected value.Read full → below that, which is unremarkable for a system whose distribution has a long left tail. Had we stopped at spin 60 with a stop-winstrategyStop-winA pre-set profit target at which you end the session.Read full → of $400, we would have left up $410, and the next session would have carried the same 1.8% per cycle.

Stop Loss / Stop Win

What the session shows

Martingale converts a 48.65% bet into a sequence that wins 98% of its cycles and loses 63 unitsmoneyUnitThe base stake a system or a player works in. Progressions are described in units so they apply at any table.Read full → on the other 2%. It does not change the expected loss per dollar; it multiplies the dollars. The flat bettor beside us wagered a third as much and lost, on average, a third as much. The player who runs this system for a year will win most evenings and lose the year.

Flat Betting

How often a losing streak comes

An even-money bet on a single-zero wheel loses 19 spins in 37. Each row is the chance that a given run of spins all lose, and what a Martingale player, who doubles after every loss, has lost by then.

Base bet
losses in a rowchanceabout 1 inlost so farnext bet
151.4%2$1$2
226.4%4$3$4
313.5%7$7$8
47.0%14$15$16
53.6%28$31$32
61.8%55$63$64
70.94%106$127$128
80.48%207$255$256
90.25%403$511$512
100.13%784$1,023$1,024
110.065%1527$2,047$2,048
120.034%2974$4,095$4,096
130.017%5792$8,191$8,192
140.0089%11279$16,383$16,384
150.0046%21965$32,767$32,768

The chance is 19/37 multiplied by itself once per loss, counted from one given spin. A long session has many spins where such a run can start, so over an evening it is far more likely than one row suggests.

Questions

What if the limit were $5,000?

Nine losses break it instead of six: 0.25% per cycle, and $5,110 lost when it happens. Rarer cliff, higher cliff, same −2.70%.

Should I have bet $500 at spin 67?

It would have won 48.65% of the time and left the cycle at −$130 anyway. There is no good move after a broken Martingale; there is only the stake you choose next.

Did the stop-win matter?

Yes, for that session. Not for the mean over all sessions.

Fun with Math

Try these yourself first. Each solution is folded; open it when you are ready.

  1. At a $1 to $500 table, a MartingalestrategyMartingaleProgression that doubles the bet after each loss. See the Strategy section for what it does and does not do.Read full → player starts at $1 on redbetRedEven-money bet on the 18 red numbers. French: rouge.Read full → on a single-zerowheelZeroThe single green pocket that no even-money, dozen or column bet covers; the source of the house edge. Also: single zero, 0.Read full → wheel. How many losses in a row stop the doubling, how much has been lost by then, and how often does it happen?

    Show the solution

    Answer Nine losses: $511 lost, and the next bet would be $512. It happens once in about 403 sequences.

    The stakes are 1, 2, 4, 8, 16, 32, 64, 128 and 256: together $511. The tenth would be $512, over the limit. A $500 bet that wins still leaves the player $11 down; one that loses leaves them $1,011 down.

    Red loses with chance 19/37, so nine losses in a row have a chance of (19/37)⁹ ≈ 0.25%, once in about 403 sequences. Each completed sequence wins $1, so the player expects about $402 of small wins before the one $511 loss, and the 2.70% of everything wagered is still lost on average.

    On a double-zero wheel the chance is (20/38)⁹, once in about 323. Bollman (2021) prints “about 1 in 833”, which is (18/38)⁹, the chance of nine wins in a row.

Comic relief

Forty-one small wins and one large loss. That is the whole system.

Forty-one small wins and one large loss. That is the whole system.

Sources

  1. Ethier, S. N. The Doctrine of Chances: Probabilistic Aspects of Gambling. Springer, 2010. Chapter 13, roulette.
  2. Thorp, E. O. “Optimal gambling systems for favorable games.” Review of the International Statistical Institute 37(3), 1969.
reviewed 2026-09-28model: European, single zeromethodology ↗sources ↗