Thursday, March 21, 2013

Introducing 3-Play. It's not as much fun as, um, foreplay, but (for most of us) it's a whole lot more profitable. And by keeping wagers in check, it makes Target-style progressive betting just a little less scary.

OK, I confess I was planning to keep the "3-Play" modification of the Target rules to myself, since I'm not selling anything on these pages (or anywhere else!) and too many people going home with house money might force those sneaky casinos to make their games tougher to beat.

But then I saw the above bit of mythematical sleight of tongue, and decided to let everyone join in the fun.

Let me first of all make it quite clear that the analysis referenced by Imspirit in yet another of his stupefying attempts to spread gambling industry disinformation was not Target.

My version of progressive betting is perfectly capable of running in the green for more than 54,000 automated and human-free shoes of baccarat before being "debunked" by a one-in-a-zillion losing streak, but the author of the strategy in question was a gentleman whose PhD presumably makes him even smarter than I am.

Imspirit, regular readers will remember, was the guy who changed the rules of my betting strategy specifically to record a "loss" against a simulated baccarat data set that anyone can access online.

It took the guy weeks to agree to make his deliberately fraudulent work available for my analysis, and when I pointed out his multiple errors in methodology, he pronounced himself "not interested" in running his evaluation of Target again with the correct rules in place.

Oh, well, so much for accuracy and integrity.  Or being "just"...

The Lim Method Imspirit examined, probably without properly applying the rules, ran profitably for the equivalent of more than forty years of playing baccarat five hours a day, five days a week, for 50 weeks every year, before a spectacular crash and burn.

Imspirit was then delighted to pronounce the method catastrophic on the basis of two anomalous losing streaks that it is both logical and scientific to conclude probably would not have had a chance to prove fatal if a real player betting real money in real time had been on the receiving end of real cards dealt from a real shoe.

All simulations (my own included) assume the participation of a suicidal player with limitless funds on one side of the table, and on the other side of the game, a casino willing to accept bets from one to thousands of units at the same layout.  Neither element exists in real life, obviously, but "sims" save time and money so we're stuck with them.

To the nitty-gritty, finally:

Progressive betting is a strategy that recognizes and exploits the mathematical given that since in the long run no one can hope to win more often than he or she loses, the only way to consistently beat the house advantage is to win more money when you win than you lose when you lose.

It takes arithmetic to beat arithmetic is another way to put it.

The problem with progressive betting is that it can push bet values very high indeed before recovery or turnaround is achieved, and a losing session or series of bets is at last profitably concluded.

Target mitigates the draw-down problem to some extent by limiting increases in wagers from one bet to the next to a handful at a time (unlike, for example, a Martingale, which keeps doubling the bet until a win recovers all prior losses, and a new series begins with a minimum wager).

After telling Target students to be flexible and to tailor the strategy to their own style and resources, I usually recommend at least a 3x bet after an opening loss, followed by at least one double-up (-1, -3, -6) before freezing the bet until a win ends a losing streak.

Better yet, x5 after an opening loss, then x2 twice (-1, -5, -10, -20), but that needs a hefty bankroll and gonads to match.

3-Play is a more cautious way to go, one that's based on the mathematical certainty that over time, more than 85% of all streaks either way (for the player or for the house) will end in three bets or fewer.

The math is simple enough: In a relatively even game (blackjack or baccarat, for example), half of all outcomes will be followed by an opposite outcome (+1, -1), 25% will extend to one more win (+1, +1, -1, either way) and half as many again will be three like outcomes, followed by an opposite (+1, +1, +1, -1).

Add 50% to 25% to 12% and you will see that in the long run 87% of all house winning streaks will end in three bets or fewer.

The same applies to player winning streaks, of course, but we will get to that in a moment.

At a full-size baccarat layout with multiple seats filled, it's possible to apply 3-Play by sitting out after three consecutive opening losses in a new series (say, -1, -5, -10).

Assuming you're too smart to swallow the casino crap about Banker being the only smart bet in the game of baccarat, you would be backing Player and a nine-round Banker winning streak would look like this: -1, -5, -10, 0, 0, 0, 0, 0, 0, (0).  And after missing out on the Player win that ended the losing run, your Target bet in response to the trigger would be 26u.

As always, your chances of winning your hoped-for turnaround bet are no better or worse than for any other bet, namely a pip or two above a 49% probability at baccarat.

What we're saying in the bet sequence above is that we hope to win a minimum of 1u per round overall, although we won't complain if a "chop" (B, P, B, P, B, B, P, B) pattern persists throughout the shoe, and we end up with an even bigger profit than that.

The standard Target sequence for the above rounds would be -1, -5, -10, -10, -10, -10, -10, -10, -10, +10, making the recovery or turnaround bet worth 76u instead of 3-Play's 26u.

My recommended sequence would be in even bigger trouble after nine consecutive losses: -1, -5, -10, -20, -20, -20, -20, -20, -20, +20 (-(116+10) = 126u, cut to 100u if an NB=PBx5 cap is applied).

The idea is to accept a smaller overall win in exchange for fewer (I'm using the technical term here) brown-trouser moments, and 3-Play does a very good job of that.

However you choose to apply progressive betting, you're going to run into logistical problems from time to time, the most common being that your required next bet exceeds the current table limit, and you have to move to a more expensive layout to get the money down.

You'll also encounter baccarat games where sitting out for one or more rounds is not permitted by the house, in which case, you should simply drop the bet to the table minimum, and proceed that way.

The minimum drop is the only option at blackjack, too, but at craps and roulette you can skip as many rounds as you want without being hassled.  You're facing tougher odds at those two games, but that should not worry you until the NB (next bet) values climb really high.

As I am always saying, I dislike mindless sims.  But at least they help illustrate what can happen when a betting strategy's rules are modified against the same set of random outcomes.

For example, here's a 5,000-round RNG sample with Target backing Player throughout, with 3-Play applied with the sit-out option (-1, -5, -10, 0):

And next, 3-Play applied the same way against the same random data set, but with Target betting BANKER all the way:

We get a tidy profit whichever side we choose to bet, meaning that partners betting opposites independently at the same table would both be winners after 5,000 rounds (roughly 65 shoes, or two weeks of grind for a player with an extremely high boredom threshold!).

A pattern you will see repeated pretty much every time is that, in spite of the fact that the Banker bettor won more often against the same set of outcomes, the so-called "5%" commission saw to it that he or she went home with two-thirds of the Player profits.

Sure, Banker wins more often than it loses.  But the house gouge more than makes up for the difference, over and over again, and if you are a Banker slave (as many baccarat players sadly are) you will pay dearly for those extra wins.

Here's the same data sample with Target staying in the game with 3-Play dropping to a 1-unit bet after three opening losses:

...and Banker against the same outcomes:

Again, "Banker" took home 40% less than "Player" because of that ugly commission rake.

I have on occasion admitted to switching temporarily from Player to Banker after a prolonged losing streak, and I hate to do it.

When I make the hop to the dark side, I hop right back to Player as soon as I rack up a turnaround.  And every potential recovery bet (defined as any bet that will get you out of the hole if it wins) is subject to a boost of at least 10% to cover the "Bank Tax."

Truth is, baccarat is in my view the most numbingly tedious game that isn't played on ice (sorry, curling fans) and its only saving grace is that in most casinos, routine table limits are higher at full-size layouts than they are for blackjack.

So, that's all I have for now on 3-Play, a great way to go for Target players who prefer to trade bigger wins for smaller risks.

And if course, like any other, my RNG can be prodded into showing a "catastrophic" loss for Target.  But in real play, I don't sit still for 27-round losing streaks, and my guess is that neither do you.

Target's overall win rate sits at 99.95% (five losses in 10,000 series is a worst-case scenario) and that's a very considerable improvement over the random bettor's 49.35% win expectation betting Player at baccarat.

Obviously, you won't win any more often with Target: Negative expectation will always see to it that over time, the number of your wins will be exceeded by your number of losses.

But you will leave the casino with more cash than you took to the game over and over and over again because (as in the first screen shot above) your average win (+19) exceeded your average loss (-15) by a percentage far greater at +23% than the actual value (AV) of the house edge for the same 5,000 outcomes (1.16%).

Given the AV, if you had been betting randomly and racked up the same action as Target, you "should have" LOST almost 1,000 units.

Instead, you WON more than 7,000u (almost +8.8% of your action instead of -1.16%).

And that, folks, is the name of the game when you play it with Target on your side.

(Here's the streak analysis for the 5,000 random outcomes that produced the screenshots above):


_ 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 Bovada (formerly Bodog), spend a few minutes and save a lot of money by reading this. _