My colleagues in Golden Touch Craps Inc., Dominator and Mr. Finesse, are advocates of regression betting as a means of securing wins and avoiding losses. A typical regression play, called the $110 one-hit-and-down, would go like this:

The shooter has established his point. You place $110 inside; that?s $25 on the 5 and 9, $30 on the 6 and 8. The shooter has 18 ways to hit these four numbers. If he does, the payoff is $35 on the first hit. Once that hit takes place, you lower your bet to $22 across, securing a profit of $13. Now, the rest of this shooter?s roll is pure profit.

The thinking here is that the player has a 50-percent chance of winning this wager on any given roll as there are 36 possible combinations of which the ?one-hit-and-down? bettor has 18 of them favoring him. How long would you leave your $110 at risk? Some players would leave it up until it hits or loses to the 7. Others will give themselves a few shots at it before pulling it back.

Is this a good way to wager? Actually, yes, it is, well, kind of.

In a random game of craps, where the 7 is always a 1 in 6 probability, the house edge is a fixed mark, in the case of placing the ?inside numbers? of 5, 6, 8, and 9, that mark is about 2.6 percent. What that means is that the player making the ?inside numbers? bet will lose 2.6 percent of all the money he wagers ? in the long run. If on one bet he has $110 at risk and on other bets he only has $22 at risk, the house edge is working on some average amount between the two extremes.

Of course, when regression betting is working to perfection (meaning it?s violating the probability gods) and an inside number is hit early and often on shooter after shooter; it is a guaranteed one-roll win per shooter. But when the 7 shows before an inside number, the loss is brutal. Once you?ve lost the $110, if you continue to regress on future shooters as you planned will take approximately 10 shooters in a row to get past that first hit or one shooter to have a monster roll as you collect on your $22 across in $7 increments. Yeow!

How often will you be wiped off the board before an inside number can be hit? You?ll win 18 times on your numbers, lose six times on that abominable 7. That means you?ll win 3 times for a win of $35 + $35 +$35 = $105; then you?ll lose once for a loss of $110. You?re $5 in the hole. You?ll have the same effects once you regress down to $22 across. You?ll win $7 + $7 +$7 = $21 but you?ll lose the fourth bet for $22. You?ll be down $1.

The math of craps is annoying because it won?t go away; it?s implacable, like a dreaded in-law. While regression betting does reduce the house?s take on your money, it does so because you are betting less, not because you have come up with a way to beat the game or reduce the house edge.

Still, 75 percent of the time using the above regression will result in being ahead of the game on a particular shooter. If you come to a casino and jump out ahead on the first few shooters, you?re now in the catbird?s seat. You might even consider changing your attack to stop going up on that $110 inside and instead continue with $22 inside on subsequent shooters.

Not all regressions are as drastic as the $110 and down that I?ve just explained. You could, for example, put $44 inside and when it hits once, go to a six dollar 6 and 8. A single hit on $44 inside would see you win $14. Now with just $12 at risk on the 6 and 8, you are ahead $2. Again such a betting strategy cannot turn a negative into a positive; the house has its damnable edge on each and every bet you make in a random game betting those inside numbers.

Is regression betting worth considering? Yes, it is, if you can stomach putting a lot down and losing it to a ?point-seven out? which will happen often enough to make even those with the strongest stomachs queasy.

Do I enjoy regression betting? No, I do not. As a general rule, once I?ve gone with a shooter (after assiduously applying the Captain?s 5-Count), I?m usually on that shooter come hell or high water. (Is high water considered good in this expression?) I prefer to take the risk to go for the bigger win by keeping my initial bets at their initial levels. But in math terms, against random rollers, I am giving the house a better whack at me than had I regressed.

For controlled shooters, regression betting is probably a more powerful strategy than it is on random shooters. Why? Because controlled shooters will have a tendency to hit certain numbers more than their probabilities indicate (most merely attempt to avoid the 7) ? and they will tend to have somewhat longer than average rolls. The longer a controlled shooter rolls, the better the chance that fatigue or distraction will alter his throw, either making the game random once again or, worse, precipitating the appearance of the 7. There is an intriguing speculation amongst the brain trust of Golden Touch that controlled shooters, when they are the least off, tend to hit more 7s than normal because they are still changing the nature of the game but, being off slightly, they are changing it for the worse! A regression bettor can capitalize on that and lock up wins before the 7 shows.

Regression betting has its advocates and its opponents. Some people swear by it; some swear at it. In the final analysis, against random shooters, you?re betting less; against controlled shooters, you might be getting in for some significant hits before coming down. If you like this as an idea, try it out on your next casino trip and see if you like it as much in reality.

Some of Frank's most popular books can be purchased by clicking one of the image
links below: