Showing posts with label house edge. Show all posts
Showing posts with label house edge. Show all posts

Wednesday, April 1, 2009

Confidence, consistency, cold cash and commitment: The keys to winning at table games can be summed up in two letters, P & S, Progression and Spread.

_
Years ago, a Las Vegas resident who promoted himself as the world's #1 expert on casinos and gambling told me that no betting method ever devised could beat the games in his city.

"Even if it could be done, the casinos would find a way to stop it," he said. "All they would have to do is limit spreads."

He was right. The casinos have the edge against 99.9% of all players, reporting a win percentage that is several multiples of the negative expectation at the toughest table game, roulette (5.26%).

Meaning that most players make such dumb choices when they are in a casino having "fun" that they guarantee that they will go home with a lighter wallet.

But the house advantage and the billions in easy money it brings in every year is not enough for the gambling industry.

Whenever casino personnel identify a player who they believe is winning "too often" they first satisfy themselves that he's not cheating...then treat him as if he is.

When mathematician Ed Thorp published his card-counting best seller, Beat the Dealer more than 40 years ago, Las Vegas casinos over-reacted by changing the rules of blackjack to block counters, prompting howls of protest from customers who didn't know how to count cards and didn't care to learn.

Double-downs and splits were cut back, and most annoying of all, shuffles were increased, slowing the game to a crawl.

Pretty soon, the gambling industry realized it had used a sledgehammer to crack a nut.

Players willing to study how the game could be beaten and able to apply their new knowledge with skill and consistency were in a tiny minority, and blocking them meant losing revenue from everyone else.

Perhaps more than anywhere else in the world, time is money in Las Vegas, and the faster hands are dealt, the more chips are raked into the dealer's tray.

In a matter of months, blackjack's rules returned to normal, and the crisis that wasn't a crisis at all was over.

Running a game is for the most part a passive process that permits the casino to provide the layout (the rope) and let the players do the work (hang themselves).

The house's influence is subtle, limiting the ratio of the highest permitted bet to the table minimum, to reduce the players' ability to recover prior losses while knowing that most punters limit themselves far more severely.

Spread limits are effective only against players who believe that the only way to combat a downturn in a given game is to keep on betting in the same location, hoping that sometime soon the tide will turn.

But a surprising number of gamblers play that way. Badly, in other word.

Players with a plan know where there's a table limit, there is a higher one not far away. If the next bet needed exceeds the table limit, there is a simple solution: Go play somewhere else.

It is an indisputable fact that both flat and random betting must eventually fall foul of the house edge at all games of chance.

It is also true that most players who are "in the hole" will fall into a pattern of robotic fixed betting, or will randomly hurl good money after bad by cranking up their wagers.

The hole will get deeper and deeper and the ladder (the bankroll) will get shorter and shorter, until eventually there is no way out.

Haphazard betting won't work. Flat betting won't either.

That makes progressive betting the only alternative.

And the realities of gambling also ensure that progressive betting will be relentlessly targeted for derision by anyone whose livelihood depends on promulgating the conventional wisdom that in the end, you can't win.

Blind luck can make you a winner for a while, but as we all know, luck never stays around for long.

Never forget that losing is a progressive process, because with each loss, the amount you are "behind" gets larger.

Betting the same amount each round against a negative trend will not enable you to recover, unless of course the win-loss pattern (WLP) changes and you are able to win more bets than you lose.

Betting less won't help either because even a prolonged winning streak will not get you out of trouble.

Again: Progressive betting is the only viable alternative. And the irony is that as long as you apply a disciplined, consistent plan, progressive betting is also less risky than any other option.

Say what?

It's like this: The vast majority of players tackle casino table games with far less money than arithmetic indicates is necessary to prevail against the house edge.

So in spite of the fact that the house advantage has to be relatively tiny in order to make occasional overall wins by players possible (without them there would soon be no customers) its relentless grind against an inadequate bankroll almost always proves fatal.

"Underfunded" players lose. Again and again. And losing is a very expensive way to have fun.

Progressive betting demands more up-front money, for sure.

But it very quickly builds the bankroll with recovery after recovery.

And as time goes by, ruin becomes a rapidly-diminishing threat.

Look at it this way: If you can't afford to win, you shouldn't play. Why even think about taking on the house if you know you probably don't have enough money to get the job done?

All of the BST logs posted here demonstrate that table limits serve the house, not the player.

No big surprise to some of you, but many people do not think about the fact that the quicker you get to your green ceiling, the more likely you are to be a victim of the house edge.

In blackjack session after session, playing up to the table limit and never beyond resulted in up to twenty times the number of high-value bets that unrestricted target betting required, and in consequence, far higher average bet and overall action values.

More action always means more risk. That's a gambling given.

Here are data from the current BST blackjack set (#15):

(Click on the image to enlarge it)

What we see here is that a betting spread tighter than 1-1,000 (the BST limit is 1-200) is certain to lose, over the long haul.

The two optimum spread ratios are $10-$25,000 (1-2,500) and $5-$25,000 (1-5,000).

$25,000 bets? That's madness! No, it's not.

It was never my hope, or my intention, to come up with a betting method that would permit a weekend punter on a shoestring budget to go home with his pockets bulging with casino cash.

That just is not possible, unless blind luck intervenes.

A successful player (not a gambler) has to think like the house, with big bucks up front and complete confidence that disciplined betting and The Math between them will make him a winner.

Look at the action numbers in the summary above and you will see far higher action/average bet figures for narrower spreads than 1-5,000, with a 1-500 spread losing after requiring risk that was almost ten times greater than the bet levels that optimum target betting needed to WIN!

A fluke, say the skeptics. A sample of just 7,300 hands is "not representative"! The data shown are anecdotal, meaning that they apply to this set of outcomes and have no relevance to future bets.

OK, so let's look at the summary for more than ten times as many outcomes (all the data sets in the current BST trial).

(Click on the image to enlarge it)

Bigger numbers, same story.

In coming posts, I will provide data from the two extended baccarat trials of target betting, the Jones outcomes and the Rodriguez sets covering more than 315,000 rounds over which I had no control, and that are verifiable at their independent sources.

Not representative? Of course not! One thing you learn when you challenge the status quo is that any sample of data, however large, that upends the YCW (you can't win) conventional wisdom is always non-representative and anecdotal.

Here's more BST fun and games...



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.
_

Tuesday, March 31, 2009

Beating the house is like safe driving. You have to pay attention, be alert, and steer, brake or accelerate as conditions require.

_
Runaway sims may be enough to satisfy mythematicians that the house edge in casino games is unbeatable, but they have very little to do with actual play.

Take camouflage, for example.

The double-up method of progressive betting (call it the Small Martingale, if you like) is unplayable because after a handful of redoubled bets at the same layout, you might as well pull out a klaxon horn and blast it in the pit boss's ear.

Double-up makes money, for sure. But it also makes waves. Big ones.

Consistent winning is, as far as the casinos are concerned, a crime akin to outright cheating, so if you want to play on undisturbed (and of course, you do) you have to dodge and weave a little from time to time.

The core of target betting is the rule that you must at all times know your loss to date (LTD) and be ready to cover it with an LTD+ bet when the right "trigger" comes along.

All the other rules, profitable as they may be, are essentially window dressing, in place to distract and befuddle the enemy.

That means that you can vary application of those rules as you see fit, tapping the "brake" or the "gas pedal" when you feel the moment is right (Cialis costs extra!).

All of the target betting spreadsheet models are required by the rules of fair testing to maintain the same set of parameters or protocols throughout. That means that like runaway sims, they do not tell the whole story, although unlike sims, they at least tell no lies.

Human beings playing with real money in real time against a real game quickly develop a sense of when things are going well and might continue to do so, and the reverse. And so, like a driver on a country road, they know when to speed up or slow down.

All the target betting models are based upon a minimum bet of $5 and a "house limit" of $25,000 and the method is backed by a bankroll of $1 million in make-believe moolah.

My models permit doubling the bet up to an added $100 in response to a push, and all the other target betting rules are there, including open and second loss responses, win progressions and the mid-series loss "do over" option (see the rules if this is all Greek to you!).

Missing from the mix are my recommended responses to a dealer natural (NB=PBx2) and a dealer 3+ (three cards or more) 21 (same as a push), both of them useful ways to improve profitability without adding too much extra risk.

It's worth noting that the "push rule" alone adds about $11,500 to the overall "win" from the current blackjack trial, which as I type this stands at 6,695 rounds and 1,130 recoveries unblemished by a single bust. You get a 10% increase in profits along with a 22% increase in action, which can be interpreted as potential added risk. Is it worth it? That's for a human player to decide, something a runaway sim cannot do.

Other real-life complications that are tricky to model in a simulation include "round-ups," meaning the conversion of a clumsy number into whatever units are currently in play, a variable value.

For example, when a $500 win leaves you with an LTD of $45, you would be more likely to push out a $100 chip than a $25 and 4 $5s, and a $15 LTD calls for a $25 bet. Sure, I can do that in my spreadsheet models, but why bother? (the money isn't real anyway!).

For years, I have had a habit of betting a limited spread at local "low roller" casinos where the table minimum is $5 and the max $200 or $300, accumulating several apparent "busts" to be played off when I can make a visit to Stateline or Reno, where the limits are higher.

The effect of this approach is that while very often a tight spread can be profitable, in the long run I am a loser at my neighborhood haunts. I don't broadcast the news that hanging series are always recovered down the road (literally and timewise!), because no one else needs to know that.

Among the most powerful advantages of target betting is that it tells the player when to stop winning.

Let's say that I take a half dozen incomplete NB/LTDs into a high-rent casino, each of them frozen at 200/500, meaning that I am $500 x 6 "in the hole" and the next bet required by the rules is $200, matching the wager I lost before suspending the series for later recovery.

I am not planning to bet at the $200 level indefinitely, but only until the dangling series have been recovered. If things go well, I will wrap up all six incompletes and then fall back to whatever minimum bet applies at my current location (probably $10-$25).

Without a betting strategy, I would be starting out with $200 bets, winning $3,000, then trying to decide whether it's time to quit while I am ahead, or keep on playing knowing that if I get 15 bets or less behind the house, I am going to be back in the red.

Target betting does not permit that risk. Once the suspended series have been recovered, a new contest begins back on the first rung of the ladder, with uncertainty and greed both erased from the picture.

This is a process you can see repeated over and over again if you scroll down any one of the current set of target betting models.

The average EOS for target betting comes in less than six bets at blackjack, which generally has faster turnarounds than baccarat because of naturals and winning splits or double-downs that happen to coincide with a potential turnaround bet.

In the current batch (#15) there are at this point 1,130 x EOS in 6,695 rounds, making for a 5.9-round average. But only one third of all series drag out to more than 6 bets from opening bet to recovery. That indicates that after resuming play at the suspended betting level, the odds are better than 2-1 in your favor that EOS will come in six bets or less.

Confidence and cold cash are both essential to a target betting player's long term success, but a little luck now and then does not hurt. What is important is that you always know when to quit, something that very few gamblers can claim.

There is a temptation to try for a fatter bottom line by dispensing with the "low roller" phase and starting each new series with a $100 minimum bet.

Don't do it!

Assuming a $5 minimum and an available house limit is $25,000, a $100 opener would slash your spread from 1-5000 to 1-250, and that significant "tightening" will eventually choke you. A 1-1,000 spread is about as tight as you should figure to go, if you want to keep winning without an excess of "brown-trouser moments." It's just arithmetic.

Here are more BST numbers...

(Click on the image to enlarge it)

Important data missing from the summary because of the error I described cover the comparative overall result for target betting and a limited 1-200 spread, as follows:-

bst090330c: AV/HA -9 = -6.08%, TB won $2,000 = 13.66% of $14,640 action, 157 rounds net incl 3-2 naturals, avg bet $99, bets of $1,000+ = 2

BST TL $5-$1,000: LOSS of ($9,000) = -5.73% of $157,000 action, avg bet $1,061, bets of $1,000+ = 148


It's the same old, same old...a loss forced by a tight spread, with far more high bets required and a result closely in line with negative expectation (-5.73% vs. -6.08%) and a very different story for target betting, which delivered a tidy win with a far smaller average bet value and just two bets of $1,000-plus.

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.
_

Monday, March 30, 2009

The million-dollar question: Is damage control just another gambler's fallacy, or can ruin be avoided intuitively?

_
Human intuition or decision making is the critical element that the You Can't Win brigade omits from the runaway simulations they use to "prove" that the house advantage at games of chance is ultimately unbeatable.

They also eliminate cards, dice, wheels, tables, money, dealers and real time, and then pretend that a high-speed computer simulation based on a random number generator (RNG) is "exactly the same as a real game."

Once upon a time, I was able to ask who would be gullible enough to bet actual cash against a game that used RNG output to simulate blackjack, baccarat or roulette, and I was confident that the answer was clear.

Then the casinos came out with giant flat-screen TVs featuring curvaceously UN-flat, life-sized "dealers" and slots for suckers to push their money into. And by golly, some people do play these so-called games, lured perhaps by cleavage that they can drool over without getting their faces slapped.

The mythematical defense of sims is that at some point, a fatal negative pattern of losses and wins will occur in real play that is exactly like output from a RNG, and vice versa, and ruin will be unavoidable.

It is, as I have said before, like arguing that a car that crashes with no driver at the wheel and with the gas pedal and brakes disabled is inherently unsafe.

Sorry, guys, but gambling is a human activity and the human element really does matter.

Simulations are useful only because they run at very high speeds, saving a huge amount of real-time research, and can give some indication of how a new game (or a method of betting) will perform.

But when the Wizard of Odds and other experts sponsored by the gambling industry claim that casinos rely on simulations to evaluate new games or rule changes to old ones, they are being deliberately disingenuous.

Sims have a role to play, but is always a preliminary one.

Casinos know very well that the only tests that really matter have to be conducted out on the floor, and nothing new ever gets the green light until it has been exhaustively evaluated under real-play conditions.

I have gone to great trouble over the years to amass as many blocks of outcomes from actual play as I can, and target betting has so far recorded only one million-dollar crash-and-burn in more than 500,000 rounds of baccarat and blackjack (more than five years of play for a full time gambler, if such an animal still exists).

The pattern that caused target betting to crash and burn looked like this...


If you were to see a "deadman's drop" like this in real play, one that went on to give the house a 33.33% edge in 114 bets in which 34 more wagers went south than didn't, you would not wait for rock bottom.

You know for sure that if you were sitting at a baccarat or blackjack table and the shoe turned that vicious that fast, you would decide it was time to take a break long before your bankroll was in serious jeopardy.

As it happens, this spectacular nosedive followed eight consecutive wins that raked in $1,090 in profits before the ninth bet, for $500, lost.

For years, my rule for blackjack has been to keep redoubling during an opening win progression until the bet hits $200, add $100 each time if the winning streak continues, and accept a loss of $500 or more as EOS, happy to end the series with $590+ in extra chips, and fall back to a minimum opening bet at a different layout.

Blackjack is the perfect game for target betting because "natural" pay-offs that exceed the value of the original bet (adding 50% or I'm not playing!) and well-timed splits and double-downs make it much more profitable than baccarat.

To make up the difference, I set the win progression cut-off much higher at $1,000 vs. baccarat, and that made the rules set vulnerable in just this one instance out of more than 60,000 successful recoveries.

The first baccarat trial, bust-free after 37,062 recoveries, delivered a virtual profit of $1.97 million and retired the original $1 million buy-in. The second baccarat series had one bust in 18,446 series and earned a little more than $110,000 (better than a break-even but not by much!). The current blackjack trials stand at a funny-money win to date of $1.75 million and counting. That's a total win of almost $4 million for five years' work, chickenfeed for a CEO but a big improvement on negative expectation.

It would be easy for me to change the baccarat results to eliminate the "bust" then claim that target betting was unbeaten against hundreds of thousands of real-play rounds.

But that would be cheating, and I don't believe in that (not because I am a saint, but because it is simply not necessary!).

For the record, changing the win progression rule to match my blackjack strategy would have turned a baccarat crash-n-burn into a win equal to 5.0% of the action for that same data set.

House-trained academics have a notion that systems promoters always have an excuse when their method crashes in ruins, and eliminate the problem simply by changing the rules.

So target betting's single loss of its bankroll stands in my book, reducing my method's win rate from 100% to -1/69,272 recovered series = 99.9985564%.

I generally claim a 99.992% win rate for my betting method, equal to favorable odds of 12,500 to 1. If I am in a cautious mood, I fall all the way down to 5,000 to 1 in a player's favor, figuring that even that has to be preferable to negative odds of 495 to 505 if you're a little lucky and play blackjack well.

And then there's the obvious wisdom of not claiming a 100% win rate, but instead encouraging the casinos and their tame experts to go public over and over again with "proof" that target betting cannot beat a runaway sim and is therefore worthless.

Let's face it, if gambling industry brainiacs can show for certain that the strategy is worthless, then they have no legal argument for banning its use in their casinos.

As for the ability of a thinking human being to recognize potential danger before it becomes deadly and take steps to avoid it, "simsters" know well that evasive, self-protecting tactics are difficult (meaning inconvenient) to model.

I have done it with what I call an interactive sim, which relies on the famously unreliable Windows RNG to spew out losses and wins with a randomness that may be questionable, but is at least out of my control beyond my insertion of a clear house edge of 1.4%.

In real play, several factors are likely to prompt a player to quit a series before recovery, then resume betting at the appropriate level elsewhere. A sim can't imitate boredom, hunger, fatigue or a need to pee, among other human responses, but table limits play a key role and they can be factored in.

The way it works is that the sim starts and ends series after series as a "live" player would, but the spreadsheet platform enables a human to scroll down, identify a spread limit trigger, and respond to it.

The job is done by "freezing" all of the outcomes or rounds up to that point, then refreshing the RNG to simulate play continuing at a different layout.

This is done whenever a spread limit trigger arises, and my assumption is first that at a $5 table, a $100 bet is as high as a real player would be able to go without attracting unwanted attention. After that, the rule is a spread of 1-5: lose a $500 bet after starting out with a minimum wager at a $100 table, and move on, then a $2,500 bet, and so on.

Series very rarely drag on for long (the average EOS comes in less than six bets) so this is not nearly as complicated or tedious a procedure as it sounds.

The point is that flesh-and-blood players do not suffer the kind of downturns that the robot at the heart of a runaway sim is required to ignore in order for the routine to "prove" that no betting method can ever beat the house advantage. It is simply not human nature to take a relentless beating when real money is at stake without taking appropriate evasive action.

Runaway sims are dishonest. But in some ways, that's not a bad thing for a cool, calm and disciplined player who knows how to beat the odds and win consistently.

Here's the latest BST blackjack data (playing the $500 win progression rule!).

(Click on the image to enlarge it)

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.
_

Thursday, March 19, 2009

Give target betting a try where your learner's mistakes won't cost you cash. Then please share your experience with me!

_
I could keep on posting new examples of target betting demolishing a house edge far greater than negative expectation, but there are people out there who will not believe it until they do it themselves.

And that is how it should be.

What I have been working hard to demonstrate here is that my method's consistent success is not a fluke, it is just common sense and arithmetic at work.

The basics of target betting have been around for more than three decades and available on the Internet since 1997, but still I get warnings that my ideas won't work "next time."

How many next times do there have to be, I wonder?

I once had an interesting conversation with the publisher of a now-defunct gambling magazine who summarily dismissed all my claims for target betting, but added: "Even very smart people who ought to know better wish in their hearts that the house advantage could be beaten. And if you show them something new in a casino and start winning real money as they watch, they are very easily impressed."

His point was that personal experience is a more powerful influence than prejudice or theory, adding that the keenest academic minds can be dulled by the thrill of an easy win.

So anyone contributing to this unending series of trials should fight any temptation to rush to judgment, even if their judgment is that target betting is the only way to win.

My publisher friend said that for his part, he would remain officially unconvinced if I showed him solid wins against millions of verifiable outcomes, and he hinted that I had come to the wrong place for support.

And of course he was right.

Most of his advertising came from casinos, and his sponsors would be seriously unamused if he started teaching his readers how to threaten the gambling industry's billions in profits.

I had naively assumed that his magazine put its readers first and would be glad to make gambling more pleasurable for them. But without casino ads, there would be no publication, and sometimes self-interest trumps the greater good (ask any politician!).

Here's the latest BST session, the last for a while (winning is fun, but less fun when there's no real money to take to the cashier's cage!). Hopefully one of these days, I will be posting results that other people have obtained.

Click on the image to enlarge it

A few more words about spread limits...

Given BST's $1,000 table limit against the current batch of blackjack outcomes to date (2,418 bets), with no target betting rules applied, a by-the-book player would have squeaked out a win of $5,388/$642,815 (+0.84%) with an average bet value (ABV) of $266 and 604 bets of $1,000+.

With only target betting's LTD+ rule added, the win would have been $13,553/$1.04m (+1.31%), ABV $428, 943 x $1,000+.

Bump the table limit to $25,000, add the full array of target betting rules, and you get an overall win of $44,635/$786,000 (+5.68%), ABV $325, 131 x $1,000+.

The notion that tight spreads and a low green ceiling constitute a smarter way to bet is exposed for what it truly is: mythematics, devised by the gambling industry for the gambling industry, with player protection not even a passing consideration.

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.
_

Monday, March 9, 2009

Mixed messages: The mythematicians say there's no way to win, while casino ads say, "Come! Play! Win! And isn't it great that losing is such fun!?"

_
(Click on the image to enlarge it)

I aim to try to keep future posts shorter and let new charts and screen shots speak for themselves.

My target betting material repeats a win rate estimate of 99.992% for the TA/T rules, suggesting favorable odds of better than 10,000 to 1.

This is confirmed by the current on-going blackjack trial, which shows TA/T without a "crash and burn" loss in 12,318 series or about 69,000 bets.

My friends the mythematicians like to say that any sample of outcomes that shows the house edge being soundly thrashed is by definition non-representative, and that either fraud was involved or the method that prevailed will be ruined by the next batch of outcomes of similar size.

This back-and-forth has been going on for years, and will probably never end.

Just hold on to the thought that the casinos need players to believe that only luck can make them winners, and so the fate of their bankroll is out of their control.

Poppycock! Know your game, learn the rules of target betting and (in blackjack) play strictly by the book, and the house advantage will become an irrelevance.

(Click on the image to enlarge it)


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.
_