Tuesday, January 5, 2016

As this blog approaches its 7th birthday, let's go back to the very beginning and make one thing abundantly clear: Progressive betting is the only way to beat casino games of chance, and anyone who says otherwise is probably working for the house!



Since I started this in March of 2009, I have taken a fair amount of flak for relying on spreadsheets to demonstrate that the conventional wisdom about table games is bollocks, and that disciplined progressive betting with realistic filters applied can consistently win in a casino.


The spreadsheet platform is ideal for this job because it allows me to generate thousands of random outcomes in an instant, and then test my winning algorithm against them one by one without anyone being able to claim that I somehow "cooked" the positive result achieved against a negative data set.


Defenders of the CW may not all be house shills, but many of them become so shrill and shrewish when challenged that I'm reminded of Shakespeare's line, "Methinks the lady doth protest too much."


Their argument that "any amount bet against a negative expectation must ultimately have a negative result" is unshakeable as long as gamblers follow the herd as 99.9% of them do and set their bet values randomly or within a very tight spread.


But the moment rational, restrained progressive betting is applied with confidence and consistency, the house edge becomes irrelevant and fewer wins than losses will deliver a profit instead of the predicted loss over and over again.


I have covered all of this before, of course, and over the years, as spreadsheet programmes have evolved into the high-speed powerhouses they are today, Turnaround/Target has become leaner and meaner and vastly more efficient along with them.


These days I call my strategy Pattern Betting or 3-Play because those names more accurately describe the nitty-gritty of the methodology.


The patterns that matter the most apply to the predictable frequency of paired wins and "three-peat" losses, which comply with the law of large numbers and don't differ much from one representative random data set to the next.


I'm banging the same old drum again because I can't shake the memory of a sceptic who set out to debunk my algorithm, and when he failed to do so against multiple sets of verifiable outcomes, he dispensed with integrity, dignity and honesty and tossed out my rules one by one until he got the negative result he wanted.


The same weasel came up with the half-baked axiom that all bets of the same value are subject to the negative expectation that applies to the sample as a whole, and it's that bit of nonsense that I am keen to address here, along with some other deliberate disinformation promoted by casinos.


First, some pertinent outtakes from a spreadsheet that in some ways supersedes all the thousands of models that preceded it:







Doubters abound among my audience, and in some cases, they hint that perhaps I get positive results by doing as "Imspirit" did with my algorithm, manipulating the data until I get the big wins that support my thesis (although, of course, Dastardly Dave needed losses to earn his paycheck!).


With one exception, the data sets in play in the above screen summaries are readily verifiable: the two sets of "real shoe" outcomes in logs from Zumma Publishing, and two 1,000-shoe baccarat simulations posted online by the best-known house shill of them all, the Wizard of Odds.


The first column of outcomes are logged rounds played in real time against casino game simulations offered online by Betfair in the UK and by Ken Smith, creator of the Basic Strategy Trainer for blackjack.


It's worth noting that Betfair's online "demo" games are remarkably like the real thing, and permit a spread from £1 to £10,000.  The real-money games look pretty much the same, but limit bets to a far tighter spread and use software that responds defensively to progressive betting.  Don't even think about risking real cash on them!

First, consider the final results shown above.


It is, according to CW-pushers, "axiomatic" that while it may be possible to twiddle and tweak until an algorithm is created to achieve a win against a large negative sample of outcomes, the same method is sure to fail against any other data set of similar size.


So either you are hallucinating, or I'm a crook...


Never mind, the casinos need you to believe that you can't win in the end but losing is tons of fun, so we shouldn't be too hard on them for resorting to outright lies to defend their bottom line.


Pattern betting will get you out of a temporary slump with a single win about 90% of the time, because it is progressive and because most losing streaks will not last longer than two or three consecutive losses.


Check the numbers in the summaries: Around 50% of all player losses (25% of all rounds) are followed by a player win, and roughly the same percentages apply to player wins, most of which are immediately followed by a house win.


Streaks of three or fewer like decisions either way (-1, -2, -3, +1, +2, +3) account for about 85% of all rounds in a session and are about evenly split between the player and the house.


What matters is that progressive betting can show a profit from 1, 2 or 3 consecutive wins while the house needs a streak lasting four bets or longer to do damage, which is invariably temporary because of the natural rhythm of swings in each direction, which looks like this:




No single outcome can be predicted (whatever hocus-pocus charlatans might try to sell you) but once we analyse large data sets, useful patterns can be seen and profitably exploited.


Pattern betting has about a 43% chance of encountering the wins it needs to end a recovery series with a profit (+1, +2, +3) while threats (-3, -4 and worse) are about a 12% probability.


Even the limited number of specimen rounds shown in the screen snaps above confirm that if you bet right (progressively!) you don't need more wins than losses to get ahead.


An important element in long-term casino play is to avoid betting exactly the same way every time, as long as my algorithm's essential recovery rules are faithfully followed, and the methodology applied to the above sets was a simple variation on an ancient tune.


Basically, after an opening loss in a new series or sequence of rounds, we bet 5x, and do the same again after a second consecutive loss, halving the bet after a third loss in a row.


After a mid-recovery win, we bet up to 5x the prior bet (PB), dropping to 2x in response to subsequent wins.  Once we're in striking distance of a recovery, meaning that the loss to date (LTD) for the current series is less than PB, we bet PBx1.


I played a 1-2,500 spread here, and anything above 1-100 usually prompts at least one reader to wail, "They won't let you bet that much!"


Of course they will, but not necessarily at the same layout or game where you placed the first bet in the current recovery series.


Table limits are in place for one reason only, and that is to thwart progressive betting.


You will hear all sorts of other explanation for restrictive house rules, but they will all be blowing smoke: every casino operator knows progressive betting can consistently beat table games, and telling big fat fibs is the least of the tricks they will deploy to protect their bottom line.


It's also the reason a progressive bettor won't win against online "casino" games: The software that runs them recognizes betting patterns and responds to them in ways that an honest dealer cannot.


I have been back in England for quite a while now, developing a pattern betting algorithm for horse-racing, and it is a big disappointment that the UK Gambling Commission ("Keeping gambling fair and safe for all") not only fails to rein in dishonest online games, but does not monitor them at all once licensing fees have been collected.


Then again, the British government has a notoriously irresponsible attitude to gambling.  Nowhere else in the world is it possible to step from any town centre street into a mini-casino where almost £20,000 can be lost against rigged games in less than an hour.


Even the greedy USA recognises that high-stakes casino gambling can cause all manner of societal ills and takes care to corral big money games into specific locations far, far away from residential and retail zones. 


That means that a loser has to really want to throw his money away, traveling miles for the privilege, rather than dumping the family budget on a whim on his wobbly way home from the pub.


In my spreadsheet models, random rounds have to be played out consecutively without reference to bet values, and their purpose is to prove again and again that the "axiom" that games with a negative expectation are unbeatable is nothing but casino propaganda.


As for the nonsense that all bets of the same value face the same long-term negative odds, take a look at the max-bet summaries above.


The mathematical truth is that once the max has been reached, the house edge in the current data set (in this case a series or sequence of bets) is so far out of whack that the law of large numbers will exert a correction at some point.


We don't know exactly WHEN the tide will turn because we don't believe in predictive betting, but we do know that at the end of the day, or the week, or the month, ALL of our profits will have been derived from high-stakes bets, making table minimums all but irrelevant.


Only one of the five verifiable data sets used in this trial showed a positive flat-bet outcome (+0.15%!) and that case, as in every other, pattern betting steam-rollered the actual value (AV) of the sample and delivered an overall win that represented a significant percentage of the total action.


The data that matters is at the top of each snap above, but here's a quick look:


Sample B, 278 max wins vs. 278 max losses for a max bet win of 478,000 units (the all-bets win was 387,204)
Sample D, 135 max wins vs. 119 max losses (+200,000/248,357)
Sample F, 133 max wins, 109 max losses (+275,000/279,630)
Sample H, 343 max wins, 331 max losses (+150,000/247,132)
Sample J, 242 max wins, 248 max losses (+175,000/245,295)


I will not claim that these numbers are definitive and of course they will not be precisely repeated, any more than they can guarantee that any sample of any size will see more winning max bets than losers.


What matters is that we are looking at six separate samples of 5,500 rounds apiece with exactly the same algorithm in play against each of them.


According to the CW and all those house-propagated "axioms" the results you see cannot have been achieved.


The model I am showcasing here permits me to randomly "play" against large blocks of archived results that in most cases long predate the development of the Turnound/Target/Pattern algorithm and are set in stone in the spreadsheet, all 322,276 of them!


In the summaries above, you will see that my algorithm delivered steady profits from more than 13,000 series or sequences in which the house won more bets than we did.


According to the CW, each one of those should have ended with a loss, and even the 12,000 or so series in which we won more often than we lost "should have" delivered only small profits.


The mathematical foundation for all this is that is when you win you win more than you lose when you lose (meaning that your average winning bet exceeds your average losing bet by a percentage far greater than the prevailing house edge) you will always win when you "should have" been soundly beaten.


There's no room for flim-flam here, although I cannot flat-out guarantee there will always be green ink because any model or simulation creates an essentially unreal betting market.


For one thing, no one can bet from £5 to £12,500 against the same table layout, and my max is higher than the house limits at most tin-pot casinos, which know the arithmetic inside out and accordingly limit their punters to spreads of 1-500 at best (1-100 is more common).


Sims also require that the "player" take no defensive action whatsoever, even when he has just lost 12 bets in succession and is thousands of units in the hole, a distorting factor that I refer to as "the inertia fallacy."


In real play, we either switch to a higher-stakes layout when a table limit is reached, or we suspend betting in a prolonged recovery sequence, fall back to a minimum bet, and reactivate the postponed PB/LTD values when we can.


Probabilities are unaffected in the long term by frequent moves from one layout to another or one game to another or one game to the next, and if you like you can visualise an empty casino with 200 open blackjack or baccarat tables and imagine yourself playing out pattern betting with just one wager at each layout.


You'd win, and get some very healthy exercise along the way!


Certainly, no one in their right mind keeps battling against a bad run, even when they know that statistically, bailing out mid-series and decamping to another game does not guarantee that they won't be jumping from the frying pan into the fire.


Statistical analysis does tell us that prolonged runs either way are relatively rare (15% or less) so the odds favour defensive action in a prolonged downturn.


There is no way to know for sure that things will get better if you flee from a bad run, or that they would not have immediately improved if you had stayed put.


Just take comfort from the numbers: threats, defined as prolonged streaks for the house, occur only 15% of the time, so there is a good chance that the next game will provide you with the paired or triple wins you need to recover your losses.


You can't know that the losing streak you ran away from was going to get worse, just that if it did, it would hurt you.


And making a move instead of being a sitting duck in a fusillade will make you feel better.


"Chase" systems, as they are often sneeringly named, are derided by gambling mythematicians (and even by casino pit bosses if you will lend them an ear) because they push stakes ever higher until they finally "get lucky" and recover prior losses plus a small profit.


But it is their aggressive nature that enables them to succeed where all other methods fail.


For example, given a routine sequence that runs -1, -1, +1, +1 the flat-bet or random outcome would be a break-even, whereas a standard Martingale would be -1, -2, +4, +1 and pattern betting with the variation described above would be -1, -5, +25, +1.


So, the choice is between a wash (0.0%), +2/8 (+25%) and +20/32 (+62%) with the constant double-up Martingale running second.


The problem with double-up is that casinos routinely interfere with its use, blocking it entirely in some shops by refusing mid-shuffle play at blackjack and baccarat.


No matter, shrewd progressive punters simply stay on the move, betting at blackjack and craps where they can (because of occasional paybacks at better than even money) and otherwise making do with brief stops at baccarat and roulette layouts.


Every time I repeat that simple truth, it will be challenged, which tells me that most gamblers either don't pay attention to what is going on around them, or are deluded.


You will often hear that it is madness to risk, say, 128 units in pursuit of a 1-unit profit, and for a split second, the argument seems rational.


But what actually happens when a 128-unit bet comes home is that prior losses from seven consecutive bets have been recovered in a single wager, which makes the house edge in casino table games no threat at all.


And, of course, the smart thing to do is to pad the final win by raising the bet in the early stages of a progression, perhaps -1, -2, -5, -10, -25, -50, +100 = +7/183 (+3.8%) instead of +1/63 (+1.6%). The house edge here was -5/7 (-71.4%) so a win is an achievement to be proud of.


There is nothing "lucky" about progressive betting, for sure.


Its whole purpose is to spin a profit when more bets have been lost than won, and it does it series after series and session after session.


Most of my models track double-up, random and flat betting along with 3-Play against the same set of thousands of random outcomes, and while the Martingale often does well, it's a bust in real play because of logistical hurdles.


Much of the pattern betting algorithm is in place to camouflage the method: halving instead of falling back to a minimum after -n losing rounds, for example.


By all means, vary mid-recovery wagers during a losing streak, but don't mess with the turnaround formula when a win triggers a get-out-of-the-hole: at least the loss to date (LTD) plus the number of rounds in the current series, and never less than the previous bet (PB).


Perhaps the most important advantage of progressive betting aside from its consistent long-term profitability is that it tells you when you have won enough, and it's time to drop back to a minimum bet.


The biggest enemy for most gamblers is their inability to respect enough: They will keep pressing their luck in a winning streak, then refuse to accept that it is over, and in no time at all, their chip stack is back where it started. Or worse.


Forget hold 'em or fold 'em. What matters is heap 'em or keep 'em! You aim to win what you need, then start over.


I have tried to work with gamblers who insist that the only proper response to a prolonged losing streak is to pull back and "protect" the bankroll until a slow recovery turns the red tide back to green.


Stop-loss limits seem oh so sensible in theory, but in truth, scared money never wins.


You have to have enough money (ammunition!) in your arsenal to take you through occasional tough times and out the other side.


Gamblers who cave in when the pressure is on are the casino's best friends.


What we all have to accept is that the bankroll is the bet, and once in a great while, we might have to be ready to put it all on the line.


Certainly, we can't hope to save it by constantly surrendering great chunks of the bankroll and betting scared, because pretty soon there will be nothing left.


Instead, when the going gets tough (and it doesn't happen often) take a break, and remind yourself that if winning were always easy, there would be no losers, and no casinos either.


In my Las Vegas days, I did more walking than playing, and I was grateful whenever a dealer I knew would warn me away from their empty table with a signal that said that they were on a hot streak for the house.


Sims and models may not be the real thing, but they confirm that often, when things go bad, they will go very, very bad before they get better.  And, of course, no one runs away from a a winning streak, however improbable it may be.


Recording or remembering suspended losses and reactivating them in random order when appropriate may seem like hard work, and sticking with a core rule through thick and thin may require iron guts and very deep pockets.


But honest work should not scare anyone, and winning is always a whole lot more fun than losing.


Can we at least agree on that?


_ 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. One more piece of friendly advice: If you are inclined to use target betting with real money against online "casinos" such as Bodog, spend a few minutes and save a lot of money by reading this. _