Betting Systems, and Why Every One of Them Fails

Updated

Every betting system makes the same promise in different clothes: change the size of your bets and the odds change with them. They do not. The next hand does not know what you bet on the last one. Here is the Martingale with the actual numbers, because it is the one that sounds most convincing.

The short version
Every figure here is computed with our own solver and checked against an independent reference. The calculator applies the same numbers to your table.

The Martingale, with numbers

Double after every loss, and the first win recovers everything plus one unit. It works exactly as advertised — right up until it does not.

Ignoring pushes, you lose about 53.8% of decided hands. So a run long enough to break a progression has this chance of appearing:

4 doublings 8.4% · 1 in 12 6 doublings 2.4% · 1 in 41 8 doublings 0.7% · 1 in 142 10 doublings 0.2% · 1 in 488
The failure gets rarer as you add steps, and more expensive at exactly the same rate. That trade never comes out ahead.

Why the trade never comes out ahead

Adding a step halves the failure rate and doubles the loss when it fails. Those cancel exactly. A ten-step progression fails about 1 in 488 sequences, and when it does you are down 1,023 units — wiping out a thousand successful sequences at one unit each.

And the table caps it anyway. A $10 table with a $500 maximum allows six doublings. That is not an oversight in the system; it is the reason the limit is printed on the sign.

The other systems

Paroli doubles after wins instead of losses. D’Alembert adds and subtracts one unit. Labouchère crosses numbers off a list. Oscar’s Grind grinds toward one unit of profit.

They differ in feel and in nothing else. Each produces a different distribution of session results — more small wins and rarer large losses, or the reverse — while leaving the average exactly where it was. You cannot add up a series of negative-expectation bets and get a positive one, whatever order you put them in.

What changing your bet size actually can do

One thing, and it is not a system. If you have information that the next hand is genuinely better than average — which is what counting cards gives you — then betting more on those hands does raise your return, because you are not varying the bet by results but by the composition of the shoe.

That is the whole difference. Counting changes your bet based on what is left in the shoe. Systems change it based on what already happened to you, which is information about nothing.

Keep reading

Card countingthe one that is not a system Losing streakshow normal they are Staying in controlwhy chasing cannot work

By Martin Kosek · every number on this site is computed by us, not copied.

← Back to guides