a made up, not worthless betting system

I have a theoretical question.

Assume I came up with a betting system that "beat the house."

What would it have to do mathematically for it to be considered a winner - or one that "worked?"

For example, If you could double your bank roll 50.1% of the time, and loose your bank roll 49.9%, is that winning system?

What are the parameters to define a winning system?

The only winning system is to own the casino.

Quote: slackyhacky

What would it have to do mathematically for it to be considered a winner - or one that "worked?"

For example, If you could double your bank roll 50.1% of the time, and loose your bank roll 49.9%, is that winning system?

If you do have such a system, then congratulaions you are done.

Quote:

What are the parameters to define a winning system?

As above. A winning system is a system which on the average wins you more than you lose. Simple as that.

But if you focus on system design, you are well misguided. It is not the system which makes it a winning one. It is the game itself which must allow for an advantage. Once you find such a game, basically *any* system will be a winning one.

Hence, focus on game selection, not on system design.

EV > 0

Quote: AZDuffman

The only winning system is to own the casino.

Thanks. Not even close to the point of the question though.

Quote: wudged

EV > 0

Again, not close to the question. This is talking about a game.

Quote: MangoJ

As above. A winning system is a system which on the average wins you more than you lose. Simple as that.

Thanks! I think that makes sense.

Another question then - if you have a make believe system that actually "beats" the house, how many rolls would it take to satisfy the answer?

Quote: slackyhacky

Another question then - if you have a make believe system that actually "beats" the house, how many rolls would it take to satisfy the answer?

Not sure what a "make believe system" is.

If you have a system and you want to test if it is profitable (by playing that system, not by calculating) , you need to measure the standard deviation for each session (it shrinks with the number of sessions you play).

Play as long as your profit/loss is within 3 standard deviations above and below zero. If you a below you can be "pretty" sure you have a losing system. If you are above you are "pretty" sure you have a winning system. With "pretty sure" I mean the usual 98% certainty.

Quote: MangoJ

Not sure what a "make believe system" is.

If you have a system and you want to test if it is profitable (by playing that system, not by calculating) , you need to measure the standard deviation for each session (it shrinks with the number of sessions you play).

Play as long as your profit/loss is within 3 standard deviations above and below zero. If you a below you can be "pretty" sure you have a losing system. If you are above you are "pretty" sure you have a winning system. With "pretty sure" I mean the usual 98% certainty.

I like it. 3 std.

Quote: slackyhacky

3 std.

Who has 3 STD's?

Quote: Beethoven9th

Who has 3 STD's?

Ha ha! Love that one!

Quote: slackyhacky

I have a theoretical question.

Assume I came up with a betting system that "beat the house."

What would it have to do mathematically for it to be considered a winner - or one that "worked?"

For example, If you could double your bank roll 50.1% of the time, and loose your bank roll 49.9%, is that winning system?

What are the parameters to define a winning system?

If you had a mathematical chance of doubling your bankroll 501/1000 times and losing the remaining 499 times, that would be a winning system.

The trick is that you need billions of samples, not 1000. So if you double 501 out of 1000 trials, that's not enough to know anything about the system.

Most folks who think they have winning systems don't have the programming expertise to test their system through a simulator and get to enough samples to know whether the system will work or not.

But the short answer to your question is that if you can get your system to withstand a billion or more events, that is at least a good start.

Using that system with real money is a lot more work, though. And a system that works in a computer doesn't always work in a casino. There are real world costs that also serve to act as an additional impediment to success.

GOOD LUCK!

Systems of holding the rights cards on video poker machines that are set to give back more than 100% do work, for example. Those are working systems. You can simulate and prove that these systems work. But they are built on advantage play settings of a machine being enabled.

Quote: Ahigh

But they are built on advantage play settings of a machine being enabled.

Can you explain what you mean by that please? I assume your saying that even mathematically you may have an edge with perfect play, however the casinos gaff the machines, so in reality A perfect VP player playing a 100%+ machine will lose in the long run.

Quote: Ahigh

If you had a mathematical chance of doubling your bankroll 501/1000 times and losing the remaining 499 times, that would be a winning system.

Doesn't this depend on what kind of edge you have? Lets say someone had a 25% edge would this still hold true?

Quote: AxelWolf

Doesn't this depend on what kind of edge you have? Lets say someone had a 25% edge would this still hold true?

He gave the specific example of doubling your bankroll, not 'doubling or better' your bankroll. I'll give you an example of when you might have greater than 25% edge but will rarely double your bankroll... Lets say you are playing a VP game that would normally pay 95%, but they decide to pay a million dollar bonus for any royal. For your $1000 bankroll, say, you will probably lose it all pretty quickly on those situations where you don't hit a royal, but if you do hit one you multiply your bankroll by a thousand. So you will not double your bankroll that often, even though each hand you play would be positive EV.