Martingale calculator

This is a live simulation of the martingale system: double after every loss, reset after a win. It is preloaded with a $1,000 bankroll, a $1 base bet and a 49% win chance — change anything and every number recomputes.

Scenario

Current game: an abstract betting round — each round wins with probability 49.00%, paying 1 to 1. This is not any real casino game; to try one, pick a preset above, such as baccarat.

Drag this first. It changes the conclusion more than anything else on the page.

Advanced — win chance, payout, push, table limits, seed

Rounds are capped at 100,000.

72.5% of sessions finish ahead — typically by $245.

27.2% run out of bankroll or can no longer place the next required bet, typically down $871.

Across all 50,000 simulated sessions the expectation is −$48.

Estimated from 50,000 simulated sessions and still converging. Figures carry a margin of error that narrows as the run completes.

startmeanmedian
Ending bankroll. Grey shows flat betting on the same scenario. Estimated by simulation.

Expected versus typical

Expected ending bankroll
$952

The average across every simulated session (−$48).

Typical ending bankroll
$1,239

What the middle session experienced ($239).

These differ because the distribution is lopsided. The average is pulled by outcomes that occur rarely; the median describes the session in the middle.

Preparing sample paths…

Outcomes

Finish ahead
72.5%
Finish behind
27.5%
Bankroll exhausted
0.1%

Could not place even the minimum bet.

Strategy could not continue
27.1%

The next required bet exceeded the bankroll or the table maximum, with money still left.

Full detail — percentiles, risk metrics, and the exact figures

Distribution of ending bankroll

P1
$8
P5
$44
P25
$502
P50
$1,239
P75
$1,249
P95
$1,261
P99
$1,270

Risk

Average largest bet
$307
Median maximum drawdown
$255
Average longest losing streak
8
Average total wagered
$2,244

Exact figures

Computed in closed form. No simulation, no margin of error.

Expected loss per $100 wagered
$2
Fair payout for this win chance
1.0408 to 1
Break-even win chance at this payout
50.00%
Chance the next 10 resolved rounds all lose
0.12%
Chance of at least one 10-loss streak within 500 rounds
25.08%

A different question from the one above, and a much larger number.

Typical longest losing streak
8

95th percentile: 12.

What actually happens

Over 50,000 simulated sessions of 500 rounds, 72.5% finish ahead — but the average result is still −$48, because 27.2% of sessions end unable to place the next doubled bet, typically down $871. Frequent small wins and rare large losses are the same coin; martingale only chooses which side shows.

What martingale cannot do

Bet sizing cannot move the expectation. At these odds the expected loss is $2 per $100 wagered — computed exactly, no simulation — and it is the same whether you bet flat, double after losses, or follow any other progression. What doubling changes is the shape of the distribution: it trades many small wins for occasional total losses. Drag the rounds slider upward and watch the failure probability climb.

Related experiments