When I launched this new online venture, I chose to let it evolve in chronological order, with my (occasional, but inevitable!) mistakes doing their part to comply with the warts-and-all nature of blogs.
The problem arose in the batch of BST blackjack sessions that just concluded because compliance with the BST 1-200 ($5-$1,000) table limit was not tracked as fairly as it should have been.
One of the big differences between target betting and tight-spread play (accepting that for many people, a 1-200 spread is far from tight!) is that the strategy player betting in real time will never abandon a series before recovery.
He will often suspend play in mid-series, but EOS is always required before the bet value can fall back to the minimum to signal a new contest against the house edge.
My models are all set up to "play" the same way, and that caused a problem for BST limit tracking because with a low table limit, recoveries took forever.
That, of course, is the whole purpose of house-mandated spread limits. The idea is that maximum bets should be kept low so that a player is likely to get into trouble sooner, and then after a while, it will be impossible for him to get "out of the hole" if bets are maintained at that level.
In BST #15, for example, the $1,000 limit player slipped deep into the red at round 323 and did not get out of it for 1,950 bets (about 13 hours of play!). The recovery was a minor miracle in itself, but the good times did not roll for long: at round 3,300, trouble returned, and when the set wrapped up at round 7,834, BST TL was more than 163 maximum bets behind.
It would take more than a minor miracle for recovery to be achieved again under those circumstances.
I corrected the anomaly by permitting BST TL to call it quits at the end of each session, win or lose, just as a hamstrung limit player would have to do in real play (if Peter Punter gets his clock cleaned to the tune of say $20,000, he goes home and licks his wounds, then starts afresh the next time he's dumb enough to go back to the casino for a repeat thrashing).
The effect of the correction was to make BST TL a less dramatic loser than before:-
Thanks to target betting, the $1,000 limit bettor (let's call him Max Grand) came out of the series of blackjack sessions with a loss that was just 0.19% of his total action, improving considerably on the net house edge of 2.0%.
Max had to make 1,989 bets of $1,000 plus ("plus" being acceptable because of splits and double-downs) vs. high-limit target betting's 398. In other words, the BST table limit did its job and forced Max to risk more overall before his inescapable end-of-session loss.
Max's total action was fractionally less than TB's, indicating that although TB ventured into high-roller territory only one fifth as often as Max, when he did so, his bets were bigger on average.
The critical information from the above summary is that Max lost a little more than $5,000. TB won $148,000, or 5.25% of his total action vs. Max's -0.19%.
The summary below explains how TB was able to win while Max tanked:-
According to gross negative expectation of -5.82%, TB "should have" suffered a loss of $164,000. Net AV was a hair worse than -2.0% indicating a best result for target betting of -2.01% x action of $2,817,105 or a LOSS pretty much equal to Max's: -$5,665.
Didn't happen. And the reason it didn't happen is that TB's lesser number of wins (3,357) had an average value of $442 while his greater number of losses (3,813) had an average value of $350.
The whole point of target betting, as you know, is to Win more when you win than you lose when you lose so that losing more bets than you win won't hurt you.
TB's average win value exceeded his average loss value by +26% vs. a gross house edge of 5.82% and a net house edge of 2.01%.
I think we can all agree that 26% greatly exceeds both 5.82% and 2.01%.
Result, happiness for TB and misery for poor old Max.
As always, skeptics will point to those average bet values and scream blue bloody murder about the "big" numbers. None of the "experts" out there seem to have heard that in a casino (as in the rest of the world) it takes money to make money.
What's the best fiscal policy: risk less and lose almost every time, or venture a little more up front and win every time? You decide!
I am not about to apply the BST TL revision to all the other BST blackjack sessions (1-14) because the job has to be done manually and all I will end up with is a repeat of what we saw at the end of Batch #15.
The spread summaries remain fair and accurate because their purpose was to explore what would happen at different spread limits, starting with Peter Punter's standard 1-5 and working up the ladder.
Here's the #15 summary:-
Again we see that a spread range of less than 1-1,000 is doomed to fail in the long run.
Casinos know that, which is why no layout exists that permits bets from $5 to $5,000.
It is certainly possible to spread that wide if a player is willing to start out with a minimum bet in one place and hop up the ladder by as many rungs as recovery requires, switching games as needed.
Very few gamblers do that.
A prolonged negative turn will bring out a stubborn streak in most players that glues them in place, when the smart thing to do would be to turn tail and RUN.
It will be a while before I have all of the baccarat spread data, but right now, my hope that baccarat's lack of doubles and splits might make it a little less of a challenge than blackjack is NOT being borne out (with just 5 of the 45 baccarat sets processed, 1-1,000 remains the magic number).
In light of all of the data I have collected and analyzed in 30+ years of target betting, I don't see how self-styled gambling experts can pretend objective honesty when they claim that progressive betting cannot win in the end.
Never mind that no other betting method (fixed or random betting being the only alternatives!) has a prayer against the house advantage.
Here's how an old-fashioned Martingale would have done against the #15 outcomes:-
In case this is your first visit to the TB blog, I will say again that I do not recommend double-up as a long term antidote to the house advantage, primarily because casino pit staff will always spot it quickly and step in to interfere with its use.
That does not mean that the method is sure to lose. Far from it. My SM+ variation boosts the EOS profit from a single $5 unit to multiples that, er, multiply as a slow recovery drags on. The rule is -5, -10, -25 (not -20), -50, -100, -250, and so on.
As you can see, neither Martingale version comes close to achieving target betting's win rate.
Finally, the all-bets-to-date summary at the end of #15:-
The data above are self-explanatory, I hope, but I will pick out some highlights:
- Target betting won a total of $1.78 million or 5.12% of total action.
- The gross AV for 78,000 rounds of blackjack was -5.33% and the net AV was -0.78%
- Given the AV numbers show, target betting "should have" lost between $270,755 and $1.85 million. It didn't.
- Target betting's average bet throughout was $445.
- The strategy's average win value was $518 and its average loss value was $402, achieving its goal to "win more when you win than you lose when you lose."
- Target betting had a funny-money bankroll of $1 million that was less than 20% exposed.
- The strategy made 28,328 attempted recovery, EOS or turnaround bets, and LOST them 50.7% of the time, in keeping with the house edge for the game.
- EOS was achieved with a single win more than 65% of the time (mostly thanks to the OSL and MSL rules, but sometimes because of well-timed naturals, double-downs and splits).
- Bets of $1,000 or more were necessary less than 4.5% of the time through 78,000 rounds of betting.
- Target betting rocks...
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.
_
No comments:
Post a Comment