Sunday, August 23, 2009

My right thumb taps past $625,000 at 1T21, and I get a refresher course in the wisdom of people who are too young to gamble!

I'm not planning to put up any more screen shots of my efforts against the iPod blackjack freebie until I either hit $750,000 or go bust.

Pretty much everyone who doesn't know much about gambling assumes I'm on a crazy winning streak and will crash and burn any minute.

The assumption that no one can win, however smart they think they are, was bolstered with last week's revelation that the MIT team of card counters who famously beat Las Vegas out of $800,000 and was lionized in the Kevin Spacey movie "21" did not get to keep their winnings.

Every cent and then some was lost when the tables inevitably turned on the irreverent upstarts.

The Spacey movie, and the book upon which it was based, failed to mention that.

Personally, I don't know what to believe. I abandoned card counting years ago and have considered it over-rated ever since.

It can do very well in the short term, but must ultimately be undone by the fact that a deck that is "rich" in high-value cards is as helpful to the house as it is to the player.

I am hoping I can wean myself away from 1T21 when my BR hits $1,000,000, which should happen before the end of the month.

In the meantime, I have been enjoying the challenge of explaining my betting strategy to two inquisitive teenagers who both rate regular A's in their math classes and do not hesitate to ask questions when they do not understand a concept or detect even the faintest whiff of BS.

The boys have seen me fine-tuning target betting for years but have only recently started taking a real interest.

I am not about to run the risk of corrupting minors, so this summer I reinforced the most important message of all, which is that gambling is for suckers, but that done right, betting can reward a smart player's time and trouble very well.

The lads quickly cottoned on to the idea that it is in the best interests of the gambling industry to warn that casino house games cannot be beaten in the long run, while spending millions to convince losers that today's their day to be a winner for once.

Promoting gambling as a wise investment requires doublespeak that rivals the worst that the tobacco companies could muster back in the days when cigarette ads were unrestricted and butane-lit coffin nails were widely seen as cool accessories.

Reiterating "Target Betting 101" for Chris and his brother Tim was a reminder that the worst thing you can do to kids is to talk to them as if they

Chris suspected that blind luck was the secret of my success until he downloaded MobilityWare's ad-sponsored blackjack app to his iPod and lost his opening bankroll in short order.

(I have no connection with MW, I should perhaps add at this point, beyond nagging them to fix the split-aces glitch that made it possible to draw two naturals in a single hand!)

I'm repeating myself, but for those of you new to this blog, here's what I told the lads:

Whatever the negative expectation for a game might be, it applies equally to every hand or round.
That means that if you bet the same amount every time, or if you vary bet values randomly, you must eventually lose a percentage of your overall action (the value of all bets, wins and losses, combined) that pretty much matches the known house edge for the game.
So if your action or churn against a game with a 2.0% negative expectation is $1 million, you will probably (but not certainly) lose around two percent of a million bucks ($20,000)...or your entire BR if it is less than twenty grand.
Experts in the field of gambling like to talk about the "laws" of mathematics that are relevant to it, chiefly the law of independence of trials and the law of large numbers.
Independence of trials covers the first item in our list (the notion that negative expectation applies equally to every bet, and that the outcome of any bet is unaffected by outcomes that preceded it).
The law of large numbers essentially says that the more bets there are, the more likely it is that their collective result will match negative expectation.
Most gamblers have been brainwashed to think in "small picture" terms, treating each bet in isolation except for occasional bet-the-farm moments when they try to recover all their prior losses in a single mad wager. (A BTF wager is in essence random because it cannot be repeated if it loses, and we have already established that random or fixed-sum bets cannot prevail in the long run).
A "big picture" approach to betting that connects rounds or wagers into series in which the value of each is determined by what preceded it poses a serious threat to the validity or relevance of negative expectation.
A Martingale, for example, makes the value of a bet double that of a losing wager that preceded it (NB=PBx2), permitting a player to profit from a series in which he lost the great majority of his wagers.
No betting strategy can overcome the absolute certainties that in a table game with a house advantage (and there is no other kind!), even the cleverest or luckiest player will lose more bets than he wins, and that the outcome of any bet cannot be known ahead of time.
It therefore follows that the ONLY way to win consistently is to bet in such a way that the average value of a lesser number of winning bets exceeds the average value of a greater number of losing bets by a percentage that is substantially higher than the percentage value of negative expectation.
For example, if the average value of 1,000 bets is $10 and 2% more bets are lost than won, then the sum of losing bets (510 x $10) vs. the sum of winning bets (490 x $10) will result in a loss (-$200) that is 2% of the total action of $10,000. However, if the value of the average winning bet (AWB = $11) were to exceed the value of the average losing bet (ALB) by 10% (ALB = $10), the final outcome would be $5,390 WON minus $5,100 LOST = +$290/$10,490 = +2.76% in spite of expectation of -2.0%.
A Martingale or double-up strategy depends on the observable truth that while the odds of the player winning any given bet in a -2.0% game are (at 49-51) less than 50-50, the odds of the house winning two consecutive bets are (at 74-26) in effect more favorable to the player, enabling linked or targeted bets to repeatedly overcome negative expectation.
The primary problems with a Martingale are, first, that bets will often increase in value very rapidly indeed, and, second, that the 2x pattern of betting is almost immediately obvious to the house, potentially singling out the player for obstructive or defensive measures. (Table limits are not a problem, assuming a viable BR, because a blocked wager can be moved to another layout or a different casino whenever necessary, until the house limit is reached).
Target betting usually freezes the bet after a mid-series loss, requiring a very large increase in the NB value only in response to a mid-recovery win and thus making it much less easy to spot and obstruct than a Martingale, but generally no less effective.
Chris's assumption that a casino could easily block target betting is, I am happy to report, unfounded. Independence of trials means that if a bet is ever potentially problematic because it would mean "breaking cover" (or because the house has refused it!), suspending a series will not adversely affect the very high probability of an ultimate win. (Likewise, if one player is barred from a casino before he can win his turnaround or EOS bet, he can hand the bet off to another player without a negative impact on "The Math").
One important caveat: Target betting may not deliver consistent profits if the payout on any win is less than 100%, as in baccarat, where winning bets on Banker are subject to a 5% "commission." The baccarat solution: only bet on Player. Otherwise, stay clear of pai-gow poker, and don't risk big money on roulette, field bets at craps or 3-card poker, all of which can be profitable for a while but have a house edge that is three to ten times greater than that for single-deck blackjack. The rule: When in doubt, play the safest bet in the house (blackjack with a 3 to 2 payout on naturals).

This week I heard from a regular reader (and frequent critic!) in China who plays the Macau casinos and also bets big on foreign exchange and other ventures that are completely beyond my ken.

He had an idea for target betting that I will not divulge here, but it reminded me that when wins consistently outnumber losses or when the return on a bet is less than even money, a betting strategy can actually be a liability.

Over the years, I have tested my strategy against such sports-book staples as horse racing and college/pro games, and every time, I have retreated to the blackjack tables.

For a while, I had a great time applying the principles of target betting to wagers on horses paying 6-5 or better, but was often tripped up by the fact that odds will sometimes shorten substantially after a bet has been placed.

Paired and longer winning streaks are also much harder to achieve betting the nags than at any casino table game.

As for pro and college games, my downfall was lack of expertise in the field and an inability to grasp why anyone would bet $100 to win, say, $110 without an absolute cast-iron guarantee that their pick was going to come out on top (and for some reason, bookies never go for that idea!).

According to many of my critics, I'm a simpleton.

As such, I am happiest sticking with what I know.

What I know is that if you can win more when you win than you lose when you lose, then losing more often than you win won't hurt you.

Have I said that before?

No graphics, but here's where I stood with 1T21 earlier today...

6317 wins, 6704 losses, 1135 pushes, HA 2.73% to 2.97% including pushes.
Win to date $645,420 against negative expectation that indicates a LOSS of +/- $382,500 (precise overall action not known).

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.