You ever sit down at a roulette table and watch the guy next to you methodically work through a leather notebook filled with numbers? That was Tommy, and I watched him for six hours straight at the Bellagio back in 2018. He was running the D'Alembert system, which is basically this: you start with a base bet, say twenty bucks. Every time you lose, you add twenty more to your next bet. Every time you win, you subtract twenty from your next bet. The idea is that you're chasing equilibrium. Eventually, wins and losses have to balance out, right?
Tommy believed in it like a religion. His notebook had columns dated back months. He showed me a particularly brutal losing streak: fourteen reds in a row against his black bets. But he kept climbing the stakes because the system promised him that a comeback was mathematically inevitable. The casino was going to supply it any second now.
How These Two Systems Actually Work
The D'Alembert assumes that roulette wheels have memory, that they owe you a win if you've lost enough. It's the opposite of aggressive. You're not doubling your bets like in Martingale; you're stepping up gradually. The beauty, or the trap, is that it feels controlled. You think you're managing risk. You're tracking your money methodically. You're not chasing insanity like the guy next to you who's screaming about hot numbers.
Fibonacci is different. It's a sequence: one, one, two, three, five, eight, thirteen, twenty-one, and so on. Each number is the sum of the two before it. You use this as your bet progression. Lose? You move one position forward in the sequence. Win twice in a row? You reset. The Fibonacci proponents argue that this geometric relationship somehow maps onto the nature of probability recovery. It sounds smarter than D'Alembert because it invokes an ancient mathematical pattern.
Neither system changes the house edge. This is the part Tommy never wrote in his notebook. American roulette has a 2.7 percent edge on every single spin, red or black, European roulette 1.35 percent. No sequence of bets changes the probability of the next spin. The wheel doesn't care what happened before. The wheel doesn't owe you anything.
What I Saw at the Table
Tommy ran D'Alembert for about four hours before his base unit got too expensive to sustain. He'd climbed from twenty-dollar bets to three-hundred-dollar bets on a single losing streak, trying to chase equilibrium. Then he hit a black number and won. But it was too late to recover the session. His notebook didn't show the full picture because he'd stopped writing partway through.
I saw a woman nearby trying Fibonacci. She was quieter about it, just moved through her sequence without announcing it. She lasted longer because her progression climbed more slowly at first. But eventually, the same ceiling: you run into the table limit, or you run into your bankroll ceiling, whichever comes first. Casinos set table limits specifically to shut down these systems.
The thing that gets people is the feeling of structure. D'Alembert and Fibonacci both create the illusion that you're not gambling, you're engineering. You've got a method. You're tracking it. You're applying intelligence to something that's actually indifferent. The house edge eats your money at exactly the same rate whether you bet randomly or whether you've got a leather notebook and a century of mathematical theory behind you.
Tommy left that night down nine hundred dollars. The Fibonacci woman broke even and walked away. But the Fibonacci woman got lucky. Both of them would return eventually because the system felt like it should work. It never does. That's the final detail that doesn't make it into the notebook.
