Since the DI crowd are going to apparently read DI into every point I try to make about Craps, I thought I should quit trying to random walk around the subject. So here is what I think based on a lot of time spent on the subject. In fact it was Dice Control on Beau's website www.dicecoach.com that revived my interest in Craps. I started out determined to debunk the entire subject and reveal it for the toxic snake oil it obviously was. But it did not turn out that way.  
 
First let's define terms. I hate the term Dice Influence. Sounds like telekinesis. Now I am  
much in favor of telekinesis and would love for it to exist, but sadly the reproducible scientific evidence for it has proven evasive. So let's mark that off the list of things we are talking about.  
 
So I am going to call it Dice Control.  
 
There is a well-established method of Dice Control known as the "Army Blanket Toss". It is cheating. I would not be surprised if you could be prosecuted for it. Most likely it would be detected in any casino almost immediately and you would at least be ejected. One or both dice are not actually thrown, but skimmed across the surface so they just slide down in the same orientation they started. Can be learned with practice, is effective, but cheating. So let's mark that off the list of things we are talking about.  
 
At this point I spent some time working on a simple concept. If Dice Control exists, there must be some physical explanation for how it works. Surprisingly, I came up with one.  
 
It should be possible to throw the dice in such a way that they proceed through the air in formation, without yaw, perhaps with axial rotation around a single common horizontal axis, strike the felt squarely enough and hit the pyramid rubber soft enough to preserve some information about how they were initially orientated, some percentage of the time.  
 
The numbers on the horizontal axis would thus be "disallowed" and the 36 possible outcomes would be skewed some percentage of the time.  
 
The next question was "would this matter?" Would the effect be uselessly scattered about, or would it focus on a useful subset of numbers? Would the level of control required to give a useful percentage enhancement be within the realm of reason, something a person could learn and master?  
 
After some time and lots of spreadsheets and tests and further thought and C programs, I came to the conclusion that simple axial dice control was a learnable skill. I was lead to believe dice control could be mastered to a level where it would usefully influence certain aspects of the distribution of the 36 possible outcomes in a potentially exploitable way.  
 
Overall I figure one can be looking at 1% to 2% de-randomization of the results. Given that house advantage on the Pass Line is about 1.41%, that is worth thinking about.  
 
Next time I will give examples of calculating given levels of dice control on some example dice sets. It is tricky to go from a statement like "I can control the dice one out of every ten rolls" and translate that into its effect on the distribution of the numbers that really count, the total on the top of two dice. For now suffice it to say that when I am talking about 1% to 2% de-randomization of the results, it is going to take a lot more than controlling 1 or 2 rolls successfully out of every 100 throws. It is not that easy.  
 
From here people get a bit crazy on the entire subject. Sets and their mystical properties abound, along with facial control and more voodoo than your average new age crystal and aromatherapy store.  
 
I was prompted to work on set analysis because much of what I saw written about dice sets and their properties seems to just be made up and not really computed out based on reasonable assumptions. It is very important to know exactly what a given dice set is going to do with a given amount of control. If you don't, you have no way of knowing how to adjust your betting pattern to profit from such control.  
 
But, clearly, Dice Control has a potentially sound physical basis. It is clearly a learnable skill. It can de-randomize the 36 total possibilities in ways that can be exploited, probably in the 1% to 2% range. Even allowing for all the problems with points digging into the felt and sending dice wildly spinning and flipping off into space, impacts with chips and overly enthusiastic encounters with the pyramid rubber, it is still possible to believe that, some percentage of the time, the dice will be de-randomized.  
 
So let's take that as read.  
 
One thing you must realize is even the best Dice Control artist is a random roller. Viewed over time, their performance from roll to roll will be a "random walk" around a baseline same as everyone else's.  
 
What do we mean by a "Random Walk"?  
 
Suppose you begin a session of Craps and you start making bets. On each throw of the dice you may win, lose or be unaffected. If you plot your results for each roll, one after the other, you will be taking a "step" forward with each roll. If you win, you will step forward and up, if you lose, you will step forward and down. If you are unaffected, you will step straight forward.  
 
After the session when you look at your plot it will be a "random walk". Hopefully you will take more steps upwards toward the profit side than you do downward towards the loss side. You may wander off towards the down side a fair amount, and then turn around and wander back into positive territory. Or you may wander positive a bit, only to take a turn for the worse and start relentlessly marching negative.  
 
Just like someone taking a field sobriety test, you will be trying to walk down a straight line we call the "baseline". That line represents the long term average expectation. If the game were totally square, if all bets were even money, the baseline would be flat. In fact it would be heading off towards infinity right at zero. Such games are called "zero sum".  
 
If there is no dice control involved, the baseline will slope downwards at a rate determined by the type of bets you are making and what their composite house advantage is. If you are just making passline bets without odds, that downward slope will be minus 1.41%. If you walked for ever, infinite steps, you would wander away from this baseline in both positive and negative directions, sometimes just a bit, sometimes rather far away. But when the long term average was computed, you would find, on average, that you have drifted 1.41% negative, to the loss side.  
 
In a good session, you will take a couple of nice steps in the positive direction and get above the baseline. Then you will implement a money management scheme to ensure that if your results start to wander back towards the baseline, giving back all your nice profits, you will quit, while you are ahead.  
 
In a bad session, you will take an unfortunate excess of steps in the negative direction and wander off towards ruin. Again a decent money management scheme will catch you at your preset loss limit and demand that you quit while you can still afford the loss. Leaving you with bankroll to play another, hopefully more positive, day.  
 
The Dice Control artist is a random roller just like everyone else. They will random walk around a baseline. But if they actually have Dice Control, their baseline will not only be determined by the composite house advantage on the bets they are making, but also on how their Dice Control is affecting their personal advantage on the bets they are making.  
 
Thus their baseline may have a flatter slope, or even one that rises slowly from roll to roll. They may even have a steeper downward slope if they have not done their homework and matched their dice set to their personal control, and then matched their betting strategy to how their personal control combines with their dice set selection to change what really matters. What really matters? The distribution of the numbers, the sum of the two dice from roll to roll.  
 
The steeper the positive or negative slope of the baseline, the quicker your random walk is going to drift, on average, positive or negative. The more you risk on each bet, the bigger the steps positive or negative you are going to be taking on each roll.  
 
If you make bets with lousy house advantage, your baseline is going to slope downwards more sharply than someone making bets with lower house advantage. You can still win, but you must overcome a larger negative bias than someone focusing on the lower house advantage bets.  
 
If you can adjust your bankroll to account for the increased average size of your bets, and start taking single or double odds, you can flatten out your baseline relative to just betting the passline by itself. It is entirely possible that changing to taking double odds on your passline bets could flatten out your baseline, remove more of its negative slope, than weak dice control can offer in the way of an advantage.  
 
Plus dice control only works when you are shooting, whereas a change in your betting strategy to change the slope of the baseline by reducing the composite house advantage on the bets you are making is going to work every time you bet.  
 
So let's examine the article on "Free Odds Fallacy". (Dice Setter Newsletter - October 2006 http://www.dicesetter.com/newsletter/2006/newsOctober_06.htm) The point was that when you take odds, you change the amount of money on average which you are putting at risk. Your random walk will take bigger steps. The point I was trying to make was you absolutely had to increase the size of the area you were going to walk in by increasing your bankroll appropriately or you would greatly increase the chances that you would bust out before your fortunes could wander back into decent positive territory.  
 
That is, you have to now allow for bigger negative excursions during a session by having a bigger bankroll to start with. The dip you may have to outlast to get back into winning territory may be deeper.  
 
The Dice Control artist has the same problem. The center or baseline of their random walk may be a bit flatter, but not by much. And if they start taking odds, they will take the very same, larger steps due to the increased size of their average bets. If they do not increase their bankroll appropriately, they will run the same significantly increased chance of busting out before getting ahead.  
 
Remember Dice Control artists only get to control the dice when they are shooting. And there are typically more than a few people around a Craps table. So they only get to use their talent a percentage of the time in a given Craps session. When they do use their talent it only changes the slope of the baseline around which they are random walking. And not by much. Even with Dice Control, Craps is still a game of chance.  
 
Everyone should learn Dice Control  
 
However, I am prepared to state flatly that every serious Craps player should learn dice control. They should learn the technique and practice a decent throw. Preferably they should be shown and coached by an expert at least once.  
 
It is essential to have a smooth, practiced and reproducible throw. It is worthwhile to think about each throw of the dice. It is worthwhile to observe other shooters to see if they show signs of having learned this.  
 
It keeps your head in the game. It keeps you focused on business. And that alone is an important thing.  
 
But, and this is a big BUT, it is vastly more important to learn and use money management including setting win goals, loss limits, using reserves and lockups to ratchet forward any winning session.  
 
It is more important to know house advantages on individual bets and composite house advantages on positions (groups of bets made in parallel). All other things being equal, a strategy based on lower house advantage bets will usually financially do better.  
 
Of course the Dice Control specialist would point out that when they are shooting they get an even flatter baseline or even one that rises slightly due to their de-randomization of the dice. I can agree with that. But again, they only get to shoot a percentage of the time, so it is more important to take care of the larger issues first.  
 
It is more important to be involved in the flow of the game and disciplined enough to quit if a table or session is going south on you. Above all, you must never go on tilt and stray from a tight grip on your predetermined strategy.  
 
If you don't do this, you will become worse than a random roller, you will become a Casino's favorite thing in the entire world. "A random gambler".  
 
Dice Control is like a good golf swing. Every golfer needs to learn a decent swing. I cannot believe some of the hacking I have seen on the golf course! I swear you cannot tell if they are trying to hit the ball or kill a rattlesnake. Expensive clubs, fancy golf carts, and fat green's fees, they have. But the ability to swing in a somewhat reasonable way at a golf ball is a thing they somehow failed to invest in, along the way.  
 
And now my favorite golf joke:  
 
Three golfers were out playing on an Oklahoma course one Sunday morning when a terrible thunder and lightning storm came up. Lightning was hitting all around.  
 
Two of the golfers dove into a shallow ditch and curled up far away from their bag of metal-shafted clubs. Suddenly they noticed the third member of their party standing in the middle of the fairway holding a long iron straight up directly over his head.  
 
"Are you crazy? What are you doing?" they yelled.  
 
"I am perfectly safe," their friend said calmly. "Even God can't hit a one iron."  
 
Back on track. Just like a decent golf swing is just a part of playing par or better golf, a decent dice throw, with or without significant dice control, is just a part of the overall approach required for Craps.  
 
Perhaps that is a bad example. A decent golf swing is fairly essential to a decent golf game. Fortunately, being able to hit a one iron is not.  
 
A decent dice throw is rather far down on the list. Behind money management, knowledge of house advantage and betting strategies, monitoring game flow and its associated discipline to act on what you see and feel happening around you.  
 
As I tried to show in my article, Distance Between Sevens, (published October 2006, Playing 4 Keeps Newsletter) a given Craps session is very likely to be aberrant and not follow long term probabilities. It will be a small chunk of a random walk. It is very important to try to figure out which way it is headed, and "what is wrong with this picture". Then adapt your play accordingly, to match the direction that the game is headed.  
 
Clearly there are lots of people on both sides of the Dice Influencing question. But show me just ONE Craps shooter that does not believe in trends, hot streaks and cold tables.  
Or even (cringe) Sevens being due...  
 
PS: I am absolutely mathematically certain there is life on other planets.  
 
Copyright 2006 Mike In Hawaii  

Click Here to return to the list of prior articles ...