A Fair Coin, a Biased Coin, and the Odds of Doubling Before Ruin
Scenario: you start with 1,000 dollars and bet 100 dollars per flip until you reach 2,000 dollars or 0. What are your chances of doubling?
Fair coin. With a 50 percent chance on each flip, and the bankroll measured in units of 100 dollars, you start at 10 units and need 20. The chance of reaching 20 before 0 is the start divided by the target: 10 divided by 20, which is 50 percent.
Biased against you, 45 percent. Let r be q divided by p, which is 0.55 divided by 0.45, about 1.2222. The chance of reaching 20 before 0 from 10 is (1 minus r to the 10th) divided by (1 minus r to the 20th). r to the 10th is about 7.44, and r to the 20th is about 55.4. So the chance is (1 minus 7.44) divided by (1 minus 55.4), which is minus 6.44 divided by minus 54.4, about 11.8 percent.
A 5-point disadvantage per flip cuts the chance of doubling from 50 to about 12 percent. The ruin probability is the complement: about 88 percent.
Biased in your favor, 55 percent. r is 0.818. r to the 10th is 0.134, r to the 20th is 0.018. The chance is (1 minus 0.134) divided by (1 minus 0.018), which is 0.866 divided by 0.982, about 88 percent.
So a 5-point edge in either direction moves the odds from a coin flip to near certainty, because of repetition.
Lesson: with enough trades, small edges dominate. And costs, which turn a 50 percent game into a 45 percent game, are decisive.
Exercise: compute the doubling odds if you bet 50 dollars per flip instead, with a 45 percent coin.
A Fair Coin, a Biased Coin, and the Odds of Doubling Before Ruin
-
FXS Prop Desk
- Posts: 1689
- Joined: Wed Oct 07, 2026 9:52 am
Re: A Fair Coin, a Biased Coin, and the Odds of Doubling Before Ruin
The biased coin version is where it gets instructive. With a 55 percent edge and ten units to target, the odds of doubling before ruin climb to roughly 88 percent, using the gambler's ruin formula with q/p = 0.818. Shrink the bet to 50 dollars, twenty units each way, and it goes above 98 percent. Same edge, very different survival, purely from bet size. That is the whole case for small fixed risk in one exercise.
FXS Official Team | The Idea Lab & Strategy
-
Amanhasnoname
- Posts: 2060
- Joined: Fri Oct 30, 2026 9:45 am
Re: A Fair Coin, a Biased Coin, and the Odds of Doubling Before Ruin
I worked the exercise through on a spreadsheet because the answer surprised me the first time I saw it. With 50 dollar bets you start at 20 units and need 40, so with the 45 percent coin the chance becomes (1 minus r to the 20th) divided by (1 minus r to the 40th). r to the 20th is about 55, r to the 40th is roughly 3,060, which puts the odds of doubling at under 2 percent. Halving the bet made things far worse, not better.
That flipped how I think about small size. With a real edge, smaller bets let the edge play out and survival gets close to certain. With a negative edge, smaller bets just give the casino more flips to grind you down, and the only half decent chance is one big bet.
For trading the practical takeaway for me was that cutting size never rescues a losing system. I had a scalping method that was slightly negative after spread, and trading it at a quarter of normal risk only made the slow bleed take longer.
That flipped how I think about small size. With a real edge, smaller bets let the edge play out and survival gets close to certain. With a negative edge, smaller bets just give the casino more flips to grind you down, and the only half decent chance is one big bet.
For trading the practical takeaway for me was that cutting size never rescues a losing system. I had a scalping method that was slightly negative after spread, and trading it at a quarter of normal risk only made the slow bleed take longer.