"The trap you have sprung on yourself is classic! You have taken just one set of outcomes and fiddled and faked until you found a way to beat them."

_
That is just the start of a fairly typical response from an academic mathematician determined to defend the conventional wisdom that games of chance with a house edge cannot be beaten in the long run.

It goes on...

"You might think you are defying gravity and making fools out of some of the greatest mathematical minds in human history, but the only fool is you.

"Very smart people do not uphold the status quo for any other reason than that it is the truth.

"There's no agenda here, no conspiracy with the gambling industry to stop people from winning, just honest mathematics at work.

"If you were honest, you would accept that it is possible to take a large data set and try different betting methods until they win against that sample of outcomes and no other.

"If you were honest (and perhaps you are lying to yourself without knowing it) you would see that the very next time those rules are tried against objective outcomes, your betting method will fail to overcome negative expectation."

I have to admire the way my critics are able to convince themselves that they are the only honest ones, and I am out to challenge them by fraud.

There would be no sense in that. But that is hardly an adequate defense since the mythematicians who take occasional pot-shots at target betting say there is no sense in any of my ideas!

All I can do is repeat my offer to discuss my methodology and provide details of all my results to anyone without a transparent pro-casino agenda. That would enable a rigorous scrutiny of how I have managed to accurately (and honestly) end up with data that are diametrically in conflict with negative expectation, otherwise known as the conventional wisdom.

I have always been aware of the risk of inadvertently manipulating an existing data set until it gives me the results that I want, and that is why I place great importance in the ongoing blackjack trial that has been the subject of most of the posts on this blog and will be covered in many more.

It's not that I have "massaged" anything, but simply that I am aware of the pitfalls that have tripped others who have tried to beat games of chance. I work constantly to avoid the traps, and can disprove any observer's perception that I might be deluding myself.

Blackjack is, in my experience, the best of all the house games for target betting, because of its minimal house edge when disciplined rules of play are applied.

It is also one of the few popular table games in which player decisions can affect the outcome of a wager (pai-gow and 3-card poker can be disregarded because they do not offer an option to draw additional or substitute cards and change a hand).

In most games, you bet, and then you wait. Blackjack requires that you play a greater role in the process than just pushing out a pile of chips and crossing your fingers.

The honest truth is that the results you see in this blog are supposed to be impossible.

Words like anecdotal and subjective and non-representative are routinely bandied about by people who sometimes see themselves as "systems debunkers," out so save the world from snake-oil salesmen and other shameless charlatans.

The honest truth is that the house advantage is only unbeatable if gamblers bet the way the house wants them to, randomly, emotionally, erratically and irrationally.

Stop gambling and start playing to win, and the house is in trouble.

I made one little error in the previous post which I'm going to 'fess up to here to show readers how honest I am.

I said that target betting versus the baccarat data set had resulted in FIVE busts in 55,000 recoveries, for a win rate of 99.989%.

Truth is that combining Player-only and Banker-plays plays, there was ONE bust in more than 118,000 recoveries, for a rather more encouraging win rate of 99.9983%.

Then along came the Wobble set, which on its own had FIVE busts in 57,763 (not 36,720) recoveries, dropping its WR all the way down to 99.9913%.

I posted the pattern charts for four of the five WBL crash n burns, and suggested that in real play, target betting rules would have required the player to bail out of each of those cliff-drop sequences before serious damage could be done. (The fifth "worst case scenario" is attached to this entry).

Combining PO, BO and WBL into a theoretical team-play result that would have required three guys to sit side by side for three years and bet 4,200 shoes three different ways would have seen SIX busts in 175,940 recoveries. That's a WR of 99.9966%.

Simply squeezing a break-even out of the baccarat set by playing WBL with target betting and the 5% skim was not a great success, and it suggests that skimming is only a good idea if you expect to win more bets than you lose over the long haul.

There is only one option in a casino that makes that promise: betting Banker only at baccarat.

As for the fiddling and faking charge I quoted, there is only one fundamental target betting rule, and that is that after one or more losing bets, followed by a win, the next bet (NB) must be increased to cover the loss to date (LTD) plus the pre-set win target.

All the other rules (OL, or opening loss, 2L, 3L, the win progression or WP, and the MSL or mid-series "do over" rule) help to boost the EOS (end of series) profit but are dispensable.

So the charge that I have been twiddling and tweaking for years to defeat a unique sample of outcomes is false, as is the claim that the house edge is unbeatable.

The baccarat data set came to me independently from Lee Jones and Lorenzo Rodriguez long after the principles of target betting (also known as Turnaround) were published on the Internet in 1997.

The ongoing BST blackjack trials began in mid-2008, and as I type this, target betting is ahead $1.83 million in funny-money against 80,457 rounds with an overall gross HA of 4.73% (0.63% net). The win is more than 5.0% of total action. There have been 14,366 recoveries so far, for an average EOS of 5.6, and NO busts. The blackjack AWB/ALB number is 127%, which is dramatically higher than the 109% achieved in the baccarat set. The expected result (ER) given negative expectation of -4.73% is a LOSS of $1.7 million, not a WIN of $1.8m. Using the net HA (-0.63%), the result "should have been" a loss of $226,000.

What we see from both trials is that it is not only possible to win more when you win than you lose when you lose, but absolutely necessary.

Over the years, I have dealt with gambling experts who look at my results, back off from insisting that they are either fraudulent or impossible, and focus instead on the size of the bets required.

All I can say to that is that I never expected to be able to discover a betting strategy that could turn a shoestring bankroll into a fat wad of cash.

"It takes money to make money" is a cliche not just for gamblers but for investment in any commodity from which a substantial return is expected.

I have never pretended that target betting is easy or cheap, just that it enables a disciplined player to win consistently, making a steady profit over the long haul.

Money alone is not enough to beat the house. Countless former millionaires who went broke trying can testify to that.

But together, money and target betting make a very powerful antidote to the house edge.

Here are more nails in the coffin of the house advantage. The red numbers, rare as they are, show that once in a very great while, target betting can get into trouble.

(Click on an image to enlarge it)
The question is, does the strategy win often enough to recover those occasional busts and get back on the road to overall profit?

The answer is clear.

Even with five busts in the WBL run-through, target betting was able to win back each loss before ending up with a result that was slightly better than a break-even.

I have included the WBL data even though it is clear to me that betting that way to offset the commission gouge may not be a good idea.

I believe that WBL may be a viable way to go if the player is alert and sensible, like a human being should be. But I can't prove it, so I have to let my readers judge for themselves.



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.
_

It would be great to win all the time, but the gambling industry would hate us for it. So how about a 99.989% win rate?

_
Last time, I put up data that showed target betting flipping a $3 million LOSS betting Banker only into a $3 million WIN simply by compensating for the damage done by the 5% commission on Banker wins.

We have already seen that backing Player only against the same data set, avoiding commission entirely, would have been very profitable using the target betting technique.

So how about "the Wobble," my Americanized name for a bet-selection method the French call avant dernier (thanks, Google, for helping me discover the correct word order and spelling for "before last" en Francais)...?

The idea is that you catch streaks either way in baccarat or roulette by betting on whichever even-money option came up in the hand or spin before the last won, so P,P would call for a bet on Player, and B,R would indicate a bet on black.

It's not a bad way to go, because there is nothing more frustrating than zigging when the shoe or wheel zags (or vice versa!).

Given a six-bet Player streak at baccarat that was preceded by P,B you will catch 5 of 6 wins, or 4 of 6 if the streak was preceded by B,B.

Being right two-thirds of the time is certainly better than missing every switch in the kind of ping-pong pattern that is common in both baccarat and roulette.

Wobble comes from win before last (WBL), and it is a useful escape from brain strain for those players who need a break from blackjack or (horrors!) refuse to play it at all.

My reluctance to allow the house to withhold 5% of the win that just turned around a losing sequence and put me back in the black has kept me away from Banker bets for years.

I now accept that that was probably a mistake.

It is certainly true that when Player is getting hammered for hand after hand, the voice in my ear that whispers "just bet Banker" gets louder and louder until the only alternative is to stand up and walk away from the game.

I would like to be able to report that adding the 5% skim to target betting gave WBL a solid win against the baccarat data set, but it did not happen, and here are four good reasons why:

(Click on the image to enlarge it)
When I first posted my ideas on an Internet gambling forum a dozen years ago, target betting was quite a bit more complicated than it is today and I was repeatedly "flamed" (such a '90s term!) to a crisp when I suggested that players with more than half a brain do not stick around when the dice, deck, shoe, wheel or whatever turns against them.

Damage control is a non-viable concept, said the experts, because no one can ever know if backing away from a downturn will be jumping from the frying pan into the fire, or if the negative trend would have reversed itself if the evasion tactic had been ignored.

Overall, they say, it's a wash: a little less than half the time, damage control will help, and a little more than half the time, it won't.

The "experts," I want you to know, are full of it.

For a fixed-bet or random player, they may actually be right. But for a target bettor (a progression player is another way to describe him) they have it all wrong.

Take another look at the fatal four-pack above.

The experts tell us that when a prolonged spike in the house's favor occurs, pushing the HA far beyond standard negative expectation for the game, there is no reason to hope that it will be substantively offset in time to enable recovery of prior losses.

Again, that may (or may not) be true for a random punter. But for a player who can get out of the hole with two consecutive wins, and more than half the time can make a profit with just ONE win, ducking a potentially deadly downturn will prove right far more often than not.

Each of the killer streaks above was followed by offsets that would have handed target betting a turnaround in a heartbeat. They just didn't come soon enough for the robot player who opted to keep taking punches on the nose instead of picking up his chips and walking away.

Runaway sims "beat" any betting method ever devised by requiring the preposterous assumption that a player would not respond to a downturn like the ones you see above.

He would just keep on betting until his bankroll ran out.

Common sense does not support that conceit any more than science does.

Because of spread limits and the constant need to maintain a low profile to avoid casino interference with a winning method, a target player does not have to trust his own judgment to figure out the right time to bail out of a downturn. The strategy rules do the job for him.

I recommend a 1-5 spread limit in any one location, except at an opening layout (one where you started with a minimum bet), where 1-20 is probably OK.

In the "killer" series above, target betting would have been out of the first one 11 gets in with a $300 loss and a ($575) LTD.

The fatal series in #13 began with seven consecutive losses, and a move would have come before the 11th bet at -200 (-375).

In #20, we would probably have been out at -100 (-100) after four consecutive opening wins, and certainly after the next loss (-200).

Finally, #23's black hole would have ended after at most seven bets at -200 (-375).

Applying those strict rules would of course have affected countless other series and required a whole lot of healthy exercise and worn out shoe leather.

But one thing we know for sure is that on average, recoveries come in less than SIX bets, and series dragging on for TEN bets or more are substantially less than a 10% probability.

And let's face it, probability is what gambling in a casino is all about.

You have to know the numbers, and how to exploit them for consistent profit.

Usually, I figure at least a 99.992% win rate for target betting.

This time, the strategy delivered 99.989%, darn it.

Against a 1.35% HA game, a random punter's win rate will be 49.335%, giving him an expected 5,068 losses in 10,000 bets.

Target betting's performance betting the Wobble against the baccarat data set was ONE loss in 10,000 series (about 55,000 bets!).

You choose: 1 loss in 55,000 or 27,870 in 55,000!

OK, it's a trick question, because of course target betting will lose as many individual bets as any other method. The house advantage will see to that.

The 1 loss in 55,000 rounds is not just a loss, it is ruin, crash and burn, wipe-out or whatever else you might choose to call it.

What matters is that with target betting, more lost bets than won bets will not result in an overall loss of money 99.989% of the time.

The reverse, in fact.

Except once in 55,000 rounds IF you are dumb enough to sit still and suck it up when the house edge goes haywire.

Here are the data for the WBL "skim" run-through:


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.
_