Roulette Betting Systems Can't Beat the Long Run, but...

Could they reduce the house edge, through bet progressions, while staying within normal table limits?

"You don't need to win every spin, just the ones you bet on."

I have a progression, betting through 15 spins, that seems to reduce the overall house advantage to 1.73%.

2 different tests:

1st Million Spin Test
16368 progression losses
947360 decisions

2nd Million Spin Test
16377 progression losses
947373 decisions



Any thoughts? Anyone have test results showing comparable or better results?

No, no system can reduce the house edge. The fact that you got a 1.73% negative return is just a function of luck/variance/whatever you want to call it.

How did you come up with 1.73% anyway?

Quote: teddys

The fact that you got a 1.73% negative return is just a function of luck/variance/whatever you want to call it.

I disagree. 2 sets of spins, with 950k decisions each, and with results within .01% of each other. If this is luck, or variance, how many spins before the numbers can be considered legitimate?

Quote: teddys

How did you come up with 1.73% anyway?

Your question makes me think I used incorrect terminology, and we are discussing differing percentages. But, to answer your question, I simply took [# of losing progressions] / [total decisions].

Quote: sharkbyte

I disagree. 2 sets of spins, with 950k decisions each, and with results within .01% of each other. If this is luck, or variance, how many spins before the numbers can be considered legitimate?


Your question makes me think I used incorrect terminology, and we are discussing differing percentages. But, to answer your question, I simply took [# of losing progressions] / [total decisions].

Number of losing progressions doesn't matter, only number of dollars won and the percentage of total wagers that represents. What did your bankroll do over your two million trials? Up or down?

Quote: MathExtremist

Number of losing progressions doesn't matter, only number of dollars won and the percentage of total wagers that represents. What did your bankroll do over your two million trials? Up or down?

The overall result was a loss. Wasn't trying to imply that it won. However, I realized what was being looked for and the overall rate works out to -5.17% and -5.18% respectively. Not being a numbers expert, I would guess these fall within 1 standard deviation of the expected -5.26%.

Quote: sharkbyte

Could they reduce the house edge, through bet progressions, while staying within normal table limits?

No. On the contrary, progressive betting systems increase the effective house advantage. To see why, calculate your EV per potential winnings rather than per wager (since if you play to win, the wager is not relevant, the winnings are). An 8-bet-limited martingale sequence on a $10 wheel will win $10 170/171 times, but lose $2,550 1/171 of the time. Fair winnings for that risk would be $15 ($2550/171), but you only get $10, making that a 33% vigorish, as opposed to only 5.28% for flat betting.

It is possible to minimize the house advantage per winnings, but that involves maximum bets and parlay, i.e. increasing bets each time you win. That works simply by minimizing house grind. There are legitimate mathematics involved in improving one's game outcomes, but they can never reduce the loss below house edge, just prevent it from going too far above it; unfortunately, these legitimate maths are overlooked due to the association with betting systems.