Baccarat "tight spread" results are in, and it's not a tale of triumph, just another lesson in the wisdom of wideness!

_
Running target betting under tight restrictions is not something I relish because I know ahead of time that a hobbled horse can't win a race.

Still and all, limiting the spread to 1-200 ($5 to $1,000) resulted in 40,706 bets of $1,000 or more. Target betting spreading wide needed only 4,859...but with most of the recommended rules set aside, it lost anyway.

Baccarat is a frustrating game because somehow it is far "streakier" than blackjack. And since backing Banker is a bad idea because of the 95-100 payout on wins, I find it about as much fun as watching a deep puddle evaporate on a cloudy day!

With the rules fully intact, target betting handily beat a 1.6% house edge, but that was not the object of this particular exercise.

As we know, the target strategy bets more but less often. And the wiggle room it gets from its wide spread wraps up series faster.

Crippling the method with a $1,000 limit brought the final outcome in line with negative expectation (-1.6%) with an actual AV of -1.77%.

Turning off most of target betting's switches but leaving the $25,000 maximum in place gave us far fewer busts (4 vs. 26) but the hobbled version's failures were huge, and its overall AV was more than double expectation at -3.61%.

There was just one "bust" with all the rules in place, and next time I will post charts that again raise the critical question, Would a sentient being stand (or sit still) for a beating this bad...?

Here are the relevant summaries, "tight" first, then "wide." The third shot covers target betting with the recommended rules modified to compensate for baccarat's slower pace and its lack of doubles and splits or bonus payouts.

(Click on an image to enlarge it)



The experts who deny that the house advantage is vulnerable in the long term reject any data that indicate to the contrary, and these will be no exception.

Analysis of past outcomes (the only analysis that is possible!) is not ideal, but as with studying the weather, history is all we have.

That is why I recommend to anyone who wants to learn how to win that they apply target betting's rules to real-time play as soon as possible, preferably without risking any money until they know what they are doing.

Reality contradicts theory and analysis all the time, but that is in target betting's favor (actual conditions are almost invariably easier to handle than situations in which intuition and experience cannot be applied).

For example, table limits and the need to disguise the betting method from prying eyes make prolonged negative trends far less likely. And EOS bets have to be rounded up in real play, sometimes increasing exposure for a while, but always resulting in greater overall profits.

The summaries above all reflect the application of a rule requiring that an EOS bet cannot be less than the value of the successful bet that preceded it. In real play, that would be optional.

One thing you can take to the bank is that if the house edge was truly invulnerable, all of the results posted in this blog would be flat-out impossible.

They are real and accurate because winning more when you win than you lose when you lose is the only antidote to the house advantage.

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.
_

"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.
_

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.
_

Strange to wake up in the morning and find your brain has been working on a problem while you slept!

_
Don't worry, this is not going to be a post about how dead people can help you win at baccarat (see the first blog entry, archived at left!).

But it is something of an experiment as I adjust to the process of allowing target betting to evolve day by day under public scrutiny.

As I have said many times, baccarat has always seemed to me about as exciting as watching grass grow next to blackjack, but now and then I feel obligated to revisit the game because of its huge international status.

As I drifted off to sleep last night, I was noodling the notion that since Banker always wins more hands than Player over the long run, target betting should be able to exploit that advantage by compensating for "The Gouge."

The gouge, of course, is the so-called 5% commission that, in my test, trimmed a $3.3 million funny-money win all the way down to barely $500,000.

My baccarat partner's long-ago suggestion that chips gobbled up by the gouge could be recovered with a post-EOS bet before NB reverts to the minimum did not at the time strike me as workable, and I still have my doubts.

But what if the potential EOS bet itself incorporated commission paid out during the current series, so that the NB equation would be something like NB=1.05*(LTD+Tgt+Comm)?

The effect might be disastrous, pushing bet values and overall action into the stratosphere and risk along with them.

But because Banker always has a solid edge over Player, it might just work.

The experimental aspect here is that I have never before speculated on the effectiveness of a modification ahead of actually applying it.

Given that I have 315,000 verifiable baccarat outcomes to work with, courtesy of Mr. Rodriguez and Mr. Jones, I think it's worth the effort to give it a try.

Once again, the spreadsheet platform rules! Running the test will be straightforward enough, and of course I will post the results here as soon as I am done.

The hard part will be figuring out the conditional formula to make the test accurate and effective!

Watch this space...

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.
_