Craps Betting Strategy Question

Is there a change in optimal betting strategy under the following circumstances:

1.) We know the normal distribution of 2 fair sided dice. We have the belief that over time, through many simulations we will approach this normal distribution.

2.) In our current limited number of trials, our actual distributions are either slightly skewed to the left or right of the average.

For example: we have a distribution with a low number of 7's and a higher number of 5's and 6's or 8's and 9's.

Is there an optimal betting strategy that can be created that is geared toward assuming more iterations will mean revert back to the norm versus another strategy that continues the actual observed distribution of 5's and 6's or 8's and 9's?

Quote: Asswhoopermcdaddy

Is there a change in optimal betting strategy under the following circumstances:

1.) We know the normal distribution of 2 fair sided dice. We have the belief that over time, through many simulations we will approach this normal distribution.

2.) In our current limited number of trials, our actual distributions are either slightly skewed to the left or right of the average.

For example: we have a distribution with a low number of 7's and a higher number of 5's and 6's or 8's and 9's.

Is there an optimal betting strategy that can be created that is geared toward assuming more iterations will mean revert back to the norm versus another strategy that continues the actual observed distribution of 5's and 6's or 8's and 9's?

Boy, did you tee one up

Quote: Asswhoopermcdaddy

Is there an optimal betting strategy that can be created that is geared toward assuming more iterations will mean revert back to the norm versus another strategy that continues the actual observed distribution of 5's and 6's or 8's and 9's?

This is not how the statistics of independent trials works. Nothing is ever due. The way it "reverts back" is that, over many trials, the percentage of 7s will approach 1/6. For each throw of the dice, the chance of a 7 is 1/6, regardless if 10 of the last 20 rolls have been 7, or if 0 of the last 20 rolls have been 7.

Another way to look at this is this way:

If you flip a fair coin over and over and over, the ratio of heads to tails will approach 1 to 1. However, the actual difference between the number of heads flipped and number of tails flipped will continue to increase, forever.

Compare this to blackjack, which when dealt out of a shoe or handheld deck, is a game of dependent trials. If you observe that there haven't been many 10s hitting the felt so far in the shoe, you CAN create a betting strategy to exploit this later in the shoe. That's what card counting is all about.

The reason it works for blackjack is the dependent trials part. Craps resets every time the dice are rolled.

Quote: Asswhoopermcdaddy

Is there a change in optimal betting strategy under the following circumstances:

1.) We know the normal distribution of 2 fair sided dice. We have the belief that over time, through many simulations we will approach this normal distribution.

2.) In our current limited number of trials, our actual distributions are either slightly skewed to the left or right of the average.

For example: we have a distribution with a low number of 7's and a higher number of 5's and 6's or 8's and 9's.

Is there an optimal betting strategy that can be created that is geared toward assuming more iterations will mean revert back to the norm versus another strategy that continues the actual observed distribution of 5's and 6's or 8's and 9's?

Sure, don't play craps.

If you are looking for a mathematical way to win at craps the answer is there is no mathematical way to win at craps. And that goes for betting too.

if you have a short-term anomaly, like that there have been excessive 7s, the long-term will indeed correct that. Unfortunately, the dice don't know that they must roll less 7s. So instead the anomaly simply gets buried under the force of larger numbers, the dice in the short-term perhaps continuing excessive 7s.

no clue what you are asking here.

let me take a stab at what im guessing you are talking about. If you are hitting more 5s/6s then bet more on the 5 and the 6, i would guess the same for the 8 and 9 as well? LOL ;)

Quote: nezbit

no clue what you are asking here.

I'm guessing also, but I think the thinking here is this: there is a point in every gambler's life when he goes, "wait a minute, 6s [or whatever] rolled 10 times more often than normal in the last 200 rolls won't stand. In the long run, the percentage will return close to 5/36. I should be able to cash in".

Not realizing it doesn't mean the dice even it out by 'trying not to roll' sixes.

I think a strategy towards the "anti-bias" is a dumb one.

Dice or roulette tables don't know the meaning of the word "normal". As the number of rolls (spins) increase, you expect to see a more "normal" distribution, but that in itself is a statistical anomaly. In an unbiased result set, PAST results have no meaning over future results. That is, if you see 10 8s in a row and the dice is truly unbiased, the odds of seeing the next 8 is 5/36.

If you believe in bias, then the best way to bet in roulette or craps is with the trend, not against it. In the worst case scenario, a bias doesn't exist and you will see the expected results over the NEXT number of rolls. The true odds of seeing an 8 is 5/36 despite the fact that you just saw 8 of them. In the best case scenario, the bias does exist (whether it be flawed dice, dice influence, God, magnetic forces, pip weight, whatever), and you should bet the numbers you see most often to take advantage of the pattern or bias that you think you see. If bias are biased, you will see them more. If dice are not biased, then you will see a random distribution. An anti-bias cannot exist because dice have no memory. They do have a distribution pattern however.

Mind you, you can't be dumb about it. If you believe that box cars come up 1 time in 31 rolls (instead of 1 in 36) you'd still be stupid to bet them because they only pay 30.

Ok I take your point since the outcomes are independent of future events. Suppose as you say, we are dealing with biased dice then. Boymimbo, you suggest betting with the trend.

Is there an optimal trend betting strategy that is geared towards 5's and 6's or 8's and 9's w/ a lower than usual occurrence of 7's? Can we take advantage of such a scenario and still hedge off the impact of rolling a 7's?

For example, a Don't Pass bet that is taken down on a higher frequency point? Followed by place/buy bets on the higher frequency numbers?

there is no trend betting system that works, none because there are no trends you can predict.

If there were 5 -6's in a row, you can say that there had just been 5 6's in a row, that has
nothing what so ever to do with the next roll.

So called trends are a reflection of only what has just happened. That last shooter was hot, the
table has been good..... you cant say the next shooter will be hot or the table is going to be good.
JUst ask the people at Borgata after Demauro had her 154 roll, it was reported that the craps pit
did well that night even after losing $280,000 on that roll.,.... why...trend betting....

More money is lost on so called trend betting than you can count.

If you play, use the best bets you can. The pass line bet with odds, place the 6 & 8. Buy
a 4 or 10 if the vig is paid if you hit. The field bet if they pay triple on 2 & 12 is a much
better bet than a 5 or 9.

NO bet on the table is in your favor, some are just worse than others.

Dicesetter

Concentrate on the odds bet is our best chance of winning, there is no better bet in the casino. All the other bets just tempt us to bet more numbers which hurts our chances overall. Do not be tempted by the myriad of other bets as I too often fail to heed myself.

I like to increase my odds bet as inside numbers are rolled after the point is established in 3 steps, start with $20 odds, then $40, then $60 per point sequence as inside numbers are rolled, using a $1200 per day bankroll. It works surprisingly well overall, per 17 day sessions played thusfar.