Friday, April 17, 2009

Gambling is not sensible, logical or rational. Mathematics is nothing but. So what am I trying to do here? Win sensibly, that's what!

_
I have spent most of this week testing my idea that the crippling effect of the baccarat gouge (the so-called 5% commission on Banker wins) can be overcome by arithmetic.

I plowed into the problem well aware that every baccarat book ever written says the game cannot be beaten, and that no less a gambling luminary than Prof. Edward O. Thorp gave up trying way back in the 1980s.

I actually met the man at a backgammon bash in Pebble Beach, Calif., more than 30 years ago, and the author of the best-selling blackjack bible of all time told me then that he had baccarat in his sights as his next gambling target.

Pity it didn't work out for him...

First, a summary of the results achieved when what I now think of as "skim theory" is applied to the 315,000 baccarat outcomes mentioned in earlier posts.

(Click on the image to enlarge it)


The highlighted columns contain results from individual data sets for Player only bets, Banker bets without the "skim" modification, and Banker bets with the anti-gouge LTD adjustment applied.

The surprise to me is that the new method turns out to be an even better bet than staying away from commission wagers entirely (+2.22% of overall action vs. +1.53%).

But I should be used to surprises by now: the good ones help me to do a better job, and the bad ones teach valuable lessons before I bury them forever.

Academic mathematicians and others who oppose the idea that the house advantage in casino games of chance is vulnerable will reject these results. That's their loss, because there is a lot they could learn here if they would only open their minds!

They will doubtless argue that 315,000 outcomes is not a representative sample, or that I somehow managed to massage that huge data set until I came up with a rules combination that beat over 4,000 shoes of baccarat. Sorry, guys.

As always, my methodology has been scrupulously careful, fair and accurate. Any other approach to the challenge would be a waste of time and toil.

The bottom line here is that backing Banker all the way does not have to be a losing proposition in spite of the fact that every win pays less than even money.

Beating one of the most artfully camouflaged rip-offs on the casino floor does not require any change to the rules of target betting other than the deduction of 5% from the value of each win before it is applied to the loss to date (LTD).

My erstwhile baccarat buddy would probably tell me that this new approach is a no-brainer and that he does it already. But unless he has changed his ways in recent years, he doesn't and he won't. His "method" is to apply target betting rules to a point, then abandon them at the time he needs them the most. And as a result, he is a certain long-term loser, with no hope of recovery.

The effect of the "skim" is that a recovery series continues until enough has been won to offset prior losses PLUS all commission due on winning bets throughout that series.

And the rules modification is greatly helped by the fact that over time, there will always be more wins for Banker than for Player (betting under those conditions is a unique experience for me!).

It sounds simple, if not simple-minded, but consistency is the key to beating the house advantage, and that is a very uncomplicated concept.

Hard as I try to get everything right, the one thing I did not build into this week's test is a 105% boost to each potential EOS bet to offset commission due after recovery. I'll get to that later...

Note that the total action required for the modified method to win $6.3 million before commission is fractionally less than the action necessary to make Player a $2 million winner.

And I don't doubt for a moment that even better results could be achieved by letting "The Math" identify a more efficient antidote to the gouge.

I don't want to get into that here because of the academic prohibition against "twiddling and tweaking."

As things stand, the models apply bets that would be rounded up in real play, inflating overall action but hopefully also benefiting the final win. That is one problem with a mathematically accurate, nit-picky approach: precision must sometimes be compromised to a degree to reflect reality, but only after a theory has been proved.

Speaking of reality, here's a chart of the win-loss pattern (WLP) that was disastrous for Player-only bets against the Rodriguez/Jones baccarat sets and led to the one "bust" in more than 50,000 successful recoveries:-


Bear with me a moment and ask yourself if you would stick around for an over-the-cliff plunge like this if you were playing baccarat for real money and a shoe went this far south this fast.

If your answer is No, as mine sure is, the next question that's begged is, Would it make any difference in the long run if you walked away from every egregious "house spike" in the hope of finding a friendlier WLP elsewhere?

The academic defenders of the status quo say no, absolutely not. But both mathematics and common sense say they are wrong when gambling is set aside and target betting replaces it in the equation.

For a fixed or random bettor there will often come a point where the hole he is in is too deep for the longest ladder (in the form of a partial reversal of a negative trend). Once he is 20 or 30 bets behind, as in the pattern depicted here, he's dead meat.

But the target strategist is only seriously threatened when he reaches his maximum bet value, which has nothing to do with table limits and all to do with how much money he has behind him.

Until then, two wins in succession will get him out of the hole...a mighty short ladder by any definition. And more often than not, a single win will save his bacon.

So while the mythematicians who manipulate arithmetic in the service of the gambling industry are not exactly telling a lie when they say you can't run away from bad luck, their "truth" does not apply to target betting.

I will be posting some more "over a cliff" sequences for your amusement from time to time. And I use the cliff analogy with good reason, recognizing that seeing into the future is a little easier on the road than it is in a casino.

Imagine that you are driving down a mountain road and you see a sharp curve coming up and what could be a 1,000ft drop straight ahead. Do you apply the brakes and turn the wheel, or do you keep going? The answer (Duh!) is so intuitive that it now has a place in the Merriam-Webster American Dictionary: used derisively to indicate that something just stated is all too obvious or self-evident.

And that takes me back to the heading at the top of this post.

Mathematics is not disconnected from logic and good sense, however frustrating and impractical its precision may sometimes be.

Gambling, the way most punters interpret and apply it, is illogical, irrational, imprecise and ultimately chaotic and hopeless (which is not to say that its practitioners do not keep doing the same things over and over again while hoping for a happier result!).

Arithmetic can beat the house, as we see here. Luck cannot.

The runaway sims that mythematicians use to "prove" that the house edge is ultimately unbeatable rely on rare patterns like the one above to undo a betting method.

A car driven at high speed with the steering and brakes disabled will almost certainly crash and burn.

A safety-conscious, sensible driver knows how to navigate the worst roads and weather conditions and get home (or to the bank) alive.

Which ride would you prefer?

An important reminder: The only person likely to make money out of this blog is you, Dear Reader. There's nothing to buy, ever, and your soul is safe (from me, at least). Test my ideas and use them or don't. It's up to you.
_