Back in 1998, I knew a guy named Frank who was absolutely convinced that D'Alembert was the system to beat roulette. He carried a notebook everywhere, tracking his bets, adjusting unit sizes based on wins and losses. He would raise his bet after a loss and lower it after a win, thinking that variance would eventually balance out in his favor.
Then there was Marco, a completely different personality, who swore by Fibonacci. He would sequence bets using 1, 1, 2, 3, 5, 8, 13, 21. Every loss pushed him further up the sequence. Every win? He'd reset to the beginning.
Both men came to the same Atlantic City blackjack hall on Friday nights. Both left poorer.
The Difference Nobody Noticed
The difference between D'Alembert and Fibonacci is actually enormous when you look at how quickly they escalate. D'Alembert adds one unit each time you lose. After ten losses, you've gone from ten dollars to twenty dollars.
Fibonacci goes 1, 1, 2, 3, 5, 8, 13, 21, 34, 55. After ten losses? You're betting fifty-five dollars on what is fundamentally a coin flip with a 2.7 percent house edge against you.
Frank's Spiral
Frank played for three years. His notebooks filled an entire filing cabinet. He was meticulous. He never drank at the table. He never rushed his decisions. He treated the casino like a laboratory and his bets like experiments.
What he missed was that no amount of experimental discipline changes the mathematics. The casino's edge is built into every spin. His D'Alembert system was not fighting the house edge. His betting progression was fighting variance, which is a battle nobody wins.
By year three, Frank had accumulated losses of about 85,000 dollars. He quit. He said the system was sound but his bankroll was insufficient.
Marco's Crash
Marco lasted eighteen months before a brutal downswing broke him. He hit a sequence of thirteen consecutive losses. His bet had escalated to 89,000 dollars on a spin. He had started the day with 50,000.
The spin landed on black. He had bet red. He could not cover the loss. He owed the casino and walked away from the table in shame.
That was the end of Marco's gambling. He never returned.
What They Should Have Known
Both systems are variations on the same failed concept: that your previous results dictate your future betting size, and that changed betting sizes can overcome a fixed house edge.
This is not how probability works. Roulette outcomes are independent. Frank's system could not overcome independence. Marco's system could not either.
The Actual Difference in Play
If you wanted to compare them objectively, you would simulate 10,000 spins of each system and track the total loss.
D'Alembert: you'd lose approximately 2.7 percent of all wagered money over 10,000 spins. Your total wagering would be higher than with flat betting because of the system. Your total loss would be higher.
Fibonacci: same 2.7 percent of wagered money. Your total wagering would escalate much faster due to the sequence. Your losses would be even larger.
The Fibonacci system is actually worse because variance hits harder and faster. When you hit that downswing, the Fibonacci system forces you to bet larger, which converts a moderate loss into a catastrophic loss.
Why They Persisted
Both systems persist because they are seductive. They suggest that you can control the house through discipline and mathematics. This is not true.
But the appeal remains. If I just follow this pattern. If I just stay disciplined. If I just have enough capital. Then I can beat it.
The house has heard this for 300 years. The house knows how this story ends.
