All betting systems are worthless.

Yes. Thats why people are always trying to sell someone a betting system rather than using it themselves.

The only betting system that seems to work reliably over time is: own the casino.

The only betting system that seems to work in the short run is: get lucky!!

Quote: FleaStiff

The only betting system that seems to work in the short run is: get lucky!!

Partially correct. Get lucky and get the hell outta there before the bad luck returns and you end up losing more money than you came in with.

Quote: RiverRock

get the hell outta there before ... losing more money than you came in with.

Yes, that is often my problem. The theorists all tell us that walking into the casino and puting your entire bankroll on the line is best, but its fun only if it wins.

I'd prefer to spend a few hours shooting craps than just one roll. Yet that house-edge grinds relentlessly the longer I stay there! It takes courage to walk away from the table and far more courage to walk away a winner than a loser.

The just-one-more-roll-and-my-luck-will-return thinking is the casino's major advantage.

I also face the effects of various distractions and lose track of my bets from time to time. Not good!

Since you quoted what theorists say, allow me to chime in here. The best way to approach any gambling situation is to make wagers of optimal size, using Kelly betting, based on your bankroll when you have an edge, and to not make any wagers when the house has an edge.

Quote: teliot

make wagers when you have an edge and not make any wagers when the house has an edge.

Marvelous. I'd love to be an Advantage Player and have an edge against the house. So would we all, I guess. Its just I can't count cards at Blackjack, I can't play poker and I sure don't want to be seen playing Bingo!
So I think that if I'm going to gamble it has to be against a house edge. I just look for the lower house edges: baccarat, craps, occasional blackjack.

For the gambler, optimal play is whatever gives you the most pleasure at the least financial risk.

To be more accurate, let P be pleasure and R be risk. Then you want P to be large and R to be small. Hence you want the fraction (P/R) to be as great as possible. If a large increase in risk doesn't correlate to an even larger increase in pleasure, then don't take the risk. If a decrease in pleasure correlates to a significant decrease in risk, the you'll should forgo the pleasure. Of course, "0" risk (no gambling) maximizes P/R (it is infinte).

This is actually meaningful, because for kelly betting you consider the standard deviation S and the edge E, and you want E to get as large as possible and S to get as small as possible, so you want to maximize (E/S) -- in a very techincal kind of way.

So, maximizing P/R is the gamblers road to the greatest fun with the least risk.

Enjoy your journey!

Quote: teliot

Since you quoted what theorists say, allow me to chime in here. The best way to approach any gambling situation is to make wagers of optimal size, using Kelly betting, based on your bankroll when you have an edge, and to not make any wagers when the house has an edge.

What is Kelly betting, or where can I learn more?

In most games, my betting method is to increase my bet when I win, and return to a base bet when I lose. It seems to really limit my losses while allowing me to capture and make some great gains if I ever get on a roll.

For example, in blackjack I start at $25, and if I win I progress as follows: 35, 50, 75, 110, 165, 250, 375, 550, 800, 1200, etc.

However, if/when I lose 3 times in a row (sometimes 4), I leave. Because obviously luck isn't on your side if you lose 3 in a row.

In other games, I might just increase one unit each time. I don't really have a system for craps.

Quote: Phosphorous

What is Kelly betting, or where can I learn more?
I don't really have a system for craps.

Kelly criterion: https://wizardofodds.com/kelly
I don't really have a system either, I tend to increase bets whimsically which is probably wrong or atleast not optimal.

Quote: teliot

For the gambler, optimal play is whatever gives you the most pleasure at the least financial risk.

To be more accurate, let P be pleasure and R be risk. Then you want P to be large and R to be small. Hence you want the fraction (P/R) to be as great as possible. If a large increase in risk doesn't correlate to an even larger increase in pleasure, then don't take the risk. If a decrease in pleasure correlates to a significant decrease in risk, the you'll should forgo the pleasure. Of course, "0" risk (no gambling) maximizes P/R (it is infinte).

This is actually meaningful, because for kelly betting you consider the standard deviation S and the edge E, and you want E to get as large as possible and S to get as small as possible, so you want to maximize (E/S) -- in a very techincal kind of way.

So, maximizing P/R is the gamblers road to the greatest fun with the least risk.

Enjoy your journey!

That is an awesome formula and the perfect path for every gambler to follow, no matter what their motivation. Wow, I've really got to think about this for a while and apply it to my own situation, which is:

Multi-strike 25c Jacks or Better is the best game available at my home casino and I adore the game. P is significantly increased along with a .25% increase in return. But, the E/S is not kind and the bankroll requirements are really higher than I can afford. Hmmm, so I have to decide exactly what my P/R is. Once I do that, I'll feel a lot more comfortable with the risk I decide to take (or to discontinue).

Thanks so much. What a helpful post!

In Macau when I was playing Pai Gow, a guy told me that if you just keep making min bets, the house edge will eventually grind you down, but if you make just a few bets that are much larger than your average bet, you give yourself a chance to win.

I think this concept is similar to the idea of making just one large bet instead of a lot of small ones, but you get the added enjoyment of actually playing for a while. Of course, from an EV perspective, it is still more negative than the alternatives of 1) not playing when the casino has the edge, 2) making one large bet only, or 3) making only min bets, but at least for me it offers the greatest pleasure. Individual results may vary.

Note: While I like this for games like Pai Gow, I don't think it is necessary in a game like craps where the variance is so high that you can legitimately play for a lifetime and actually win (I've run several simulations of 10,000 rolls of craps where I ended up well ahead--though many more where I lost. The point is high variance of outcomes over a large number of trials).

I read somewhere of a hotel that had a sign posted on the back of the door.
"Please walk to our casino and enjoy our games, However if you have a betting system then just sit there. We'll send the limousine over to get you".

==========================================
Belief in betting systems is fundamental to the human psyche. That's why it's so hard for mathematicians to shake people's belief in them. If you think about it, why do people insist on betting black after a long run of reds on a roulette wheel? At the same time people run to get a piece of the action if someone has a long run at a craps table. It's because there is a basic need in people to think they have some influence over outcomes. So a long run at craps is perceived as being caused by the dice thrower, while a long run at roulette is an inanimate object that is defying logical probabilities. It's not logical to bet on the dice thrower, and bet against the wheel, but it satisfies a basic psychological drive.

==========================================
The same way with Martingale. People know that losing six times in a row (assuming no house edge) has odds of 1/64 out of the first 6 plays. That is fairly simple mathematics. But the more you play, the more likely that you will get a run of 6 losses. It only takes 89 plays until you have a 50% chance of losing 6 times in a row. But unless you are extremely lucky you won't have even doubled your minimum bankroll of 64 units in the first 89 plays. Casino games are even worse, because you have a house edge as well.
= Generally these calculations involves Markov Transition matrices or the equivalent mathematics. It's graduate level mathematics, and not just algebraic formulas.

==========================================
Consider this website that claims This system [Martingale] of betting in roulette can stand to make you a decent amount of money, though will only work in the short term. It's kind of a self fulfilling statement. If you have a bankroll that is big enough to cover 6 losses in a row; i.e. 63 units of say $5 (63*$5=$315) and your goal is to win $30 to have a steak and beer then you will probably make the money using Martingale. However Martingale is not important, because you will probably win that much money with flat betting. The fact that you use Martingale or flat betting or any other system is not important. What is important is that your bankroll is high enough relative to your goal that you risk of ruin is extremely low. However, your odds of winning ten dinners in a row on successive nights is by no means assured.

removed
silly

Quote: mustangsally

Quote: pacomartin

Generally these calculations involves Markov Transition matrices or the equivalent mathematics. It's graduate level mathematics, and not just algebraic formulas.
==========================================
Consider this website that claims This system [Martingale] of betting in roulette can stand to make you a decent amount of money, though will only work in the short term. It's kind of a self fulfilling statement. If you have a bankroll that is big enough to cover 6 losses in a row; i.e. 63 units of say $5 (63*$5=$315)

and your goal is to win $30 to have a steak and beer then you will probably make the money using Martingale. However Martingale is not important, because you will probably win that much money with flat betting.

I used simple algebraic formulas to solve for a Martingale, Bold Play and simple Gambler's Ruin.(double up after a loss, Bet all bankroll or just enough to hit win goal or bust, Flat betting)

What bet (from Craps and Roulette) and method (from the 3 below) gives the best chance at winning $30 by gambling for that Beer and Steak dinner?
The best method may be a choice on how long it *can take* or just how hungry are you.
Some methods take way longer than others ;)

Results: [formula =(1-((1-p)^6))^6 or (1-((1-p)^s))^u where S=number of steps in the Marty and u=the unit wins needed to hit win goal]
Martingale 6 Step
pass line (p=244/495) 90.224613%
11.97 expected number of trials (pass line)
18/38 roulette
87.905305% 6Marty
12.4 expected number of trials
18/37 roulette
89.490398%
12.1 expected number of trials

Bold Play: (spreadsheet calculations to 4 decimal accuracy)
89.38% 18/38
1.59 trials
90.35% 18/37
1.59 trials
90.814% pass
1.59 trials

Flat Beting:
18/38
53.1114563%
500.7 trials
18/37
71.6153%
502.7 trials
pass line
81.8070%
463.4 trials

Gambler's Ruin formula links can be found by starting here
Gambler's Ruin

Quote: pacomartin

The fact that you use Martingale or flat betting or any other system is not important. What is important is that your bankroll is high enough relative to your goal that you risk of ruin is extremely low.
However, your odds of winning ten dinners in a row on successive nights is by no means assured.

I can leave that up to the reader to do the math.
Hint X^10 where X = probability of winning one session (system progression)


There may be methods that can produce even higher win probabilities using a 35 to 1 payoff for example.
I just did not have the time, yet, to work on it. Even money payoffs are easy.

Love your posts mustangsally.

I have a question for you.

Is the probability the same to hit a number on roulette if I randomly pick a number each roll, or just lay one number?

Or asked another way, say I am trying ovoid losing my money on a lay bet on the 4 or 10 (lay bet = hoping a 7 hits before a 4 or 10) - If I play a lay 10 every time, versus randomly selecting the 4 or 10, am I just as likely to loose?

Yes, Yes.
From Wikipedia
"In probability theory and statistics, a sequence or other collection of random variables is independent and identically distributed (i.i.d.) if each random variable has the same probability distribution as the others and all are mutually independent."

This is basic prob and stats stuff as you may already know.
Both Roulette and Craps falls into that category.
It makes no difference if you choose what bet, a coin chooses what bet or a blind squirrel chooses what bet. (ME's idea)
The probability or each trial or spin or roll remains the same each time the bet is made.

But if you start to talk about streaks, that is a bit different, each trail is still independent but now the result of, say 2 in a row, is not mutually independent.
2 in a row can only happen given the first trial was a success (In the case of winning 2 in a row)

Thanks for the question.
Sally

Quote: pacomartin

Belief in betting systems is fundamental to the human psyche.

Yes, we all want to believe that we have control over our life or at least sufficient knowledge about what will happen. We go to a bar knowing some hot chick will throw herself at us because we are witty and charming hot young studs who have a magic influence over females that is almost equal to our magic influence over a pair of dice. When the results fail to conform to our belief system, we hope for better results next time but don't change the belief system. We still have knowledge of the world and control over our lives and we still have the magic touch. This time we are going to find a hot chick, win at dice and beat the living daylights out of that three-cart monte creep who fooled us the last time.

Quote: pacomartin

If you have a bankroll that is big enough to cover 6 losses in a row and your goal is to win $30 to have a steak and beer then you will probably make the money using Martingale. However Martingale is not important, because you will probably win that much money with flat betting. The fact that you use Martingale or flat betting or any other system is not important. What is important is that your bankroll is high enough relative to your goal that your risk of ruin is extremely low. However, your odds of winning ten dinners in a row on successive nights is by no means assured.

That seems to be what the better gamblers do: have a few cocktails to unwind, try a slot machine or two, have a dinner, use a match play coupon, go to the bar in time for the jazz group, have some more drinks, meet up with friends, chat with table mates, take time over an exotic cocktail, etc. They accept the losses and the winning nights and just try to be lucky but have a sufficient bankroll to last a few early adverse events. By taking advantage of all the "extras" the casino offers they are making themselves willing to accept a positive but fairly low level return.

Quote: mustangsally

Yes, Yes.
From Wikipedia
"In probability theory and statistics, a sequence or other collection of random variables is independent and identically distributed (i.i.d.) if each random variable has the same probability distribution as the others and all are mutually independent."

This is basic prob and stats stuff as you may already know.
Both Roulette and Craps falls into that category.
It makes no difference if you choose what bet, a coin chooses what bet or a blind squirrel chooses what bet. (ME's idea)
The probability or each trial or spin or roll remains the same each time the bet is made.

But if you start to talk about streaks, that is a bit different, each trail is still independent but now the result of, say 2 in a row, is not mutually independent.
2 in a row can only happen given the first trial was a success (In the case of winning 2 in a row)

Thanks for the question.
Sally

If this is true, that it doesn't matter if you randomly play a number - then how come we teach our kids that if they are ever lost in the woods, to stay put in the same place? Doesn't that same rule apply in roulette?

Quote: slackyhacky

If this is true, that it doesn't matter if you randomly play a number - then how come we teach our kids that if they are ever lost in the woods, to stay put in the same place? Doesn't that same rule apply in roulette?

No. In roulette the future outcomes don't depend on the past. If a kid is lost in the woods and you're looking for him, you don't look in the same place twice: your future actions do depend on the past. Searching through the woods is like dealing down a deck of cards. If you don't find the ace of spades in the first half of the deck, it's gotta be in the second half. In roulette, if you go 19 spins without seeing a double zero, you can't say for sure that it will appear in the second 19 spins. Unlike card games, which use the "selection without replacement" model, roulette and dice use "selection with replacement".