Craps Patterns and Correlations between numbers rolled at the craps table

This post is just for information and discussion purposes. Over the many years of playing craps I have noticed certain ebbs and flows of the game as well as certain patterns and correlations between numbers. I have no proof of what I have observed. These are only my personal observations and based on my experiences. When I teach students I always tell them not to believe a word I say. The reason I state this is because everything I believe is based on my own personal experiences. Each person will have different perceptions and experiences to base their own beliefs. If I tell you something it means nothing. But, if you observe it for yourself, it means everything.

So with that said, one correlation I noticed is that if a point is a 4, 5, 9 or 10 an outside number (4, 5, 9 & 10) will usually be rolled during the shooter's roll. It does not happen all of the time. By my estimation it happens about 60 to 65% of the time. So if the shooter establishes a point of 4, I will bet the outside numbers, including the point. The same applies if a point is the 5, 9 or 10. The caveat is that I am a very conservative player and will take my bets down if the shooter has not rolled an outside number within the first three rolls of the dice. I have missed some wins by taking my bets down too early. By the same token, I have avoided losses by taking my bets down early. It is all a matter of personal preference.

I am not advocating that this bet is right or wrong, good or bad. I am only giving some information. My suggestion is that you look to see if what I am saying is true for yourself. If the point is a 4, see if a 5, 9 or 10 is rolled during that roll. An exception to this is when a shooter establishes a point and immediately sevens out. Since a point seven out or a point number seven out tend to come in bunches, I do not bet on the next shooter until I qualify that table again. As an example, lets say a shooter establishes a point of 6 and then immediately sevens out on the next roll. The next shooter then establishes 4 as the point. I will not place a bet on this shooter. Just my personal preference.

If I can get a slight advantage over the casino and it helps me to win, I will take it.
See for yourself at the craps table if you observe what I observe.
David Medansky

 

The trouble with drawing conclusions based on one's own experience is that it constitutes a very small sample. How many times have you read someone reporting some strategy that they say "has worked for me"? Then you run it in WinCraps for a few million rolls and find it doesn't work like that at all, but could in a small sample.

You know there is a way to calculate the probability of one (or more) number(s) occurring before another, right? For example, there are six ways to roll a seven and five to roll a six. Since the other rolls are irrelevant to this question, the probability of the six coming first is 5/11, or .4545, and that is, indeed, the probability of winning a place bet on 6 or 8. So, if we add up the ways to roll a an outside number (3, 4, 4, 3) we get 14 ways, so the probability of getting one of those numbers before a seven is 14/20, or .7. Of course, these figures apply whether or not the point is an outside number; they apply at any point in time.

That's not the same thing you wrote above; you included the 4. If you restrict it to the outside numbers that are NOT the point (5, 9, 10), you have 11 ways vs. 6, 11/17 or .647.

These numbers apply all the time; there is no advantage over the casino here.
Cheers,
Alan Shank

 

Thanks Alan, and David for contributing to an interesting observation at the tables.

I think David it comes down to what a player believes ... some people core believe is rooted in Math others like you in "feel" or trends.

Its funny - your observation is a little like those players that call there bets "OFF" when it hits the wood or the chips or drink girl, stick change. It's kinda of a "Feel" they have. I always wonder what they know - because you wouldn't believe how often the seven comes in those situations ... yep you guessed it 6/36 times. lol.

Again, "their" experience is the one they remember ... right ... you see they never seem to remember when the place bet comes on that "roll" that they called off and the $$ they missed out on - NO - they will only tell you when the 7 hits ... see - hit the wood ...etc. I kinda laugh inside at them.

But feel players and dice Trends are popular out there. Can you believe there are people selling $3000.00 systems based on "Trending" dice.

Any system at any time can be bang on right - and your experience may well be that every time a five is rolled - and nine comes shortly thereafter.,,or whatever people advocate.

But for me the truth is in the math.

Me (and Alan) ... we believe math over the long term is always right.

 

Question for you Alan,

Based on David's post and others like - is there any "system" of play that has been run through the Hyper Drive on Win Craps that has shown a winning Batting Average?

 

I ran a simulation in WinCraps Pro using a Don't Pass with 20X odds that was ahead after 20,000 sessions. No way do I believe this was indicative of a player advantage, but what happened was that the money bet on the odds dwarfed that on the DP. The bets on the DP did somewhat better than expectation, and the odds also did better, ending up +0.29%. I think I was using a $1600 bankroll and a $600 win goal, and the goal was reached almost 2/3 of the time, the bankroll busted about 1/6 of the sessions. The mean outcome was $15.24, with a standard deviation of $836.

What this illustrates is that, when you jack the variance up with lots of odds, it swamps the ev, so that a just a little bit of luck can overcome the expected loss. Also, if you set a win goal that's less than half of your bankroll, you will hit it quite often.

Bankroll considerations aside, $5 DP laying 20x odds, 60 bets (a couple of hours' play), the ev is just -$4.21, SD $817, which means the probability of breaking even or better is pretty close to .5. Of course, one SD on the wrong side means you lose $821! Way, way too rich for my blood.

One of the coolest things about WinCraps is that it can display the data either in numbers or graphically, so you can see what I call "The Shape of Possible Futures". If you set the bankroll high enough so you never bust, and don't set a win goal, so that each session is terminated by a number of bets or rolls, you will see a nice normal distribution. Put in a loss limit and or win goal that will be hit frequently, and you change the shape of the graph dramatically. I have attached a .pdf file that shows the histogram from this sim. I haven't tried uploading any files before, so hope you can view it.
Cheers,
Alan Shank

 

I know a lot of people that track dice looking for trends and patterns. Entire seminars are created and devoted to teach people how to look for trends with craps dice. And the results? Only tired players who spent a day of dice tracking who did not make much headway in the casinos.

OK, I do track dice. But for different reasons. Mostly to determine if the dice are bias and to track my own shooting as to if i have an edge or not at that table with those dice.

As for betting or not betting - well tracking is no better than using the 5-count or the qualified shooter method. Ideal, you want to avoid the short rolls and bet on the long rolls. The problem is you never know when such a roll will happen, unless you can see in to the future or have some Jedi-like powers to manipulate the dice. So, you guess like the rest of us what shooters will have long rolls and what ones will not. It is just that some people need to feel that their guesses are 'scientific' and more important than the slob next to them who bets on every shooter, or the guy who bets on left handed shooters, or the guy who uses the 5-count.

 

I concur with trends at the table .... I also cringe and SMH when charters ignore the energy of the table ... for instance, a Trender may Lay the 5 because it is considered a Lost number -- which is fine, but not if the Point is the 10 ... even dealers will concur that the 5 and 10 have a special energy relationship