# A sound betting system for craps

Aug 10, 2004 3:50 AM

In last week’s article on craps, I put forth my rationale for picking sides — do or don’t. Whether it be systematic or by whim, I felt the choice of sides was not particularly relevant. Past patterns, cycles, trends, etc., seemed of little consequence as the next roll was relatively a 50/50 proposition and could go either way. The player had no control over strictly random circumstances.

If one is a flat bettor (consistently the same wager), then one has given up control of his only viable choice option. He is strictly at the mercy to the whims of the dice. The ability to vary bet amounts is the only course not to be a total slave of chaotic dice behavior.

Obviously, some type of betting progression is in order. There are two types: negative and positive. Initially, the novice succumbs to the seemingly invincible nature of the negative progression. This is the continual raising of the bet size on a loss until the eventual (and inevitable) winner occurs and one recoups all losses plus a small profit. It has destroyed many a fortune and maybe even prompted a few suicides.

Given a bottomless pocket, the negative progression could eventually beat a casino. However, the maximum bet limits are made to thwart such plans. The upper limit also saves many a player from himself should he get carried away.

Positive progressions (raise bet on a win) are the obvious way to go. It seems a majority of the wise players do this. It’s the mode of the successful Kelly Criterion as well. Raise bets and compound winning streaks and cut back and minimize on losing slumps.

Most progressions follow some form of a mathematical or geometrical number sequence rationale — Martingale and Fibonacci as an example. There is a mathematical, logical premise for each ascending number.

I once wrote about the evolutionary betting procedures developed from the four-horse train. That’s three separate parlays started on every play. However, each of the three parlays are of two, three, and four-play sequence duration.

Win or lose, you start the two, three, and four-sequence parlays on each and every play. If your last play lost, then all your previous working parlays are gone and the total next bet is reduced to a simple three units.

Of course, had the last play won, then you would also have some ongoing parlay action to add onto your initial three units, etc. It can become very involved. Using just one type of side selection system, you can have nine parlays working simultaneously — some starting out, some finishing, and some ongoing.

I was looking for a stronger method to cover more eventualities to lessen the bite of the fickle antics of the dice. I did this by playing various side selection systems simultaneously, using the four-horse parlay train series on each independently.

That would be like two guys at the table using the four-parlay series but say one guy betting "follow" and the other avant dernier (time before last). However, you might have a situation where one guy was betting three units on the do and the other seven units on the don’t for the next play.

I added a couple more selection methods, each using the three parlays, and then played them all together simultaneously. The tracking arithmetic became horrendous as the dice were rolled. That’s especially if one player were trying to balance out the action, say of four or more other players.

The more I combined things, the more complex it became internally, but surprisingly, the simpler it became externally in overall combined action. While in development, I had several progressive numbers of unit bets. They eventually boiled down to only two.

These were the numbers three and five. These numbers were not just experimental, arbitrary numbers but were the synthesis of handling up to 36 simultaneous parlays on various side selection methods. Everything combined, it boiled down to these two numbers: Three and five.

In all cases, a three-unit bet was made initially or always after a previous loss. The bet is raised to five units after any win and continues as five-unit bets on successive wins. On a loss, revert back to three units. It’s too simple to believe.

But even the computer helped us here. Why or how, I can’t quite figure it out myself, but mysteriously, it favored the avant dernier side selection method for the 3-5 betting sequence.

The remaining option for the craps player (out of watching not betting the first roll of a session) is when to quit. That’s a "stop loss" on a losing sequence or "take profit" on a winning one. This seems up to the caprice or rationale of the player. Obviously, you’ll have no house-imposed bet limitations, high or low.

So far, I haven’t been able to get the computer to give me a definitive answer on quitting. However, by observing millions of simulations, I’ve set my own standards.

In any single session, set your stop loss at -20 units. That can be \$20, or better yet, a stack of \$5 chips (\$100). Bet the 3-5 method, and if it gets off badly and your stack is gone, then walk away, take the 20-unit loss. Chalk it up to a screwy, aberrant quirk of the chaotic dice.

However, if your betting stack is intact and growing — keep right on betting indefinitely until you’re tired or satisfied with your profit. I have no end profit goal.

But your winning stack will go up and down, but in any session, a peak will be reached. Note and keep track of that point and only quit when you lose back one-third of your highest profit point.