Martingale and the Ruin Problem: Why the Table Maximum Matters More Than the Streak

PROPOSAL DRAFT — needs Andrew review. Not publication-ready. Facts in source notes must be verified before publish.

Martingale has been the go-to system for 250 years because its logic sounds unbreakable: double after every loss, and when you finally win, you recover everything plus one unit. The logic is not unbreakable. Two constraints are.

What Martingale requires

To guarantee recovery after N consecutive losses, you need a bankroll of at least 2^N units and table limits that allow a bet of 2^N units. A five-loss streak requires 32 units. A ten-loss streak requires 1,024. A $10 bettor needs $10,240 just to survive ten straight losses.

Where the sequence ends

Consecutive losing streaks are uncommon but not rare. In a 6-deck game with roughly 46% win probability per hand, a six-loss streak occurs in about 1.6% of six-hand windows. Across a 300-hand session, you will encounter one with meaningful probability. The doubling sequence hits the table maximum long before a bankroll of realistic size can sustain it.

The table maximum as the real constraint

A $10 minimum game with a $500 maximum table cap stops your doubling at the sixth bet: $10, $20, $40, $80, $160, $320 — the next bet would be $640, which the casino will not accept. At that point, you have lost $630 and cannot bet enough to recover it in one hand. The system is broken by the casino's own rules, not by bad luck.

What players experience

Short sessions with Martingale produce a lot of small wins and one catastrophic loss. The small wins reinforce the system. The catastrophic loss is dismissed as exceptional. Over time, the expected value per hand is unchanged: you are betting more per session and accumulating expected losses proportionally.

Reader prompt: Load the Bot Arena and run a Martingale bot for 500 hands. Note the bankroll path, not just the end balance. Where did the largest drawdowns occur?