newtobj: The backcounter strikes

[OCD]

> So when one has a high positive count, he
> doesn't have a greater probability of
> winning? It's still the .42 win, .49 lose,
> & .09 push regardless of a high positive
> count?

Seconding the question: How exactly does the advantage play out? Which text explains this the best, with examples? My empirical experience(no sims) on the tables has been maddeningly packed with pushed 20s, pushed bjs, and busted stiffs at high counts. Perhaps I should just buy some simulator software. Which ones are Mac-friendly?
In addition, my girlfriend is also a fledgling AP, but is unwilling to tolerate the same variance as I am, and so only spreads 4-1 at most, and she has an uncannily winning record. I'm ahead more dollars (but I play a bigger unit) but as far as percentage increase in bankroll over our 50-ish hours of real AP play on the tables, she's way ahead of me. Is this just "luck"? The SD falling in her favor?
game specs:
both: High-low, using illustrious 18+fab 4 indices, plus whichever others I can remember (about 30 indices total for me, her pretty much just illustrious 18)
either 1D H17 no DAS nsr 50-65% pen (her: 1-3 me 1-4)
and
her only: 6D H17 DAS DOA RSA lsr 60-75% pen play-all spread 1-3
me only: 6D S17 DOA DAS RSA lsr 65-75% pen backcounting or play-all sitting out TC ?1, spreading 1-8.
over the last 20-ish sessions she's doubled her 100$ buy-in usually at least 3 out of 5 times, the other she loses most or all of it and walks.
It is difficult to argue with her concrete dollar successes. Again, perhaps I should just get some software.
Thank you all so much for your time when I post. It is so gratifying to effectively have a community of elders to look to for guidance in this trying new adventure of mine.

[Parker]

One of the hardest things to accept about advantage play is the fact that, until you have been playing for several years, your personal results are essentially meaningless.

The fact that you see more pushed 20's and BJ's in high counts is actually a Good Thing - it means that you are counting correctly.

Most of the advantage in high counts comes from blackjacks and successful doubles. As others have noted, we still lose more hands than we win.

Your girlfriend has been experiencing some positive variance. Instead of wondering why you're not winning, think about this: Had she been spreading 1-8 like you, she'd likely be way, way ahead.

Also, it sounds as if your girlfriend is using a stop-loss, since she quits if she loses most or all of her initial $100 buy-in.

Stop-losses and stop-wins have no effect on your overall win rate, but they can affect the distribution of session wins and losses. This is simply another reason why session results are meaningless.

Sadly, most of the simulation software is only available for PC's. As a Linux user, I feel your pain. However, computers are so cheap these days that it is actually worthwhile to buy a PC just to run blackjack software. That is the only reason that I still have a Windows system, and a Windows partition on my laptop.

[newtobj]

> One of the hardest things to accept about
> advantage play is the fact that, until you
> have been playing for several years, your
> personal results are essentially
> meaningless.

> The fact that you see more pushed 20's and
> BJ's in high counts is actually a Good Thing
> - it means that you are counting correctly.

> Most of the advantage in high counts comes
> from blackjacks and successful doubles. As
> others have noted, we still lose more hands
> than we win.

> Your girlfriend has been experiencing some
> positive variance. Instead of wondering why
> you're not winning, think about this: Had
> she been spreading 1-8 like you, she'd
> likely be way, way ahead.

> Also, it sounds as if your girlfriend is
> using a stop-loss, since she quits if she
> loses most or all of her initial $100
> buy-in.

> Stop-losses and stop-wins have no effect on
> your overall win rate, but they can affect
> the distribution of session wins and losses.
> This is simply another reason why session
> results are meaningless.

> Sadly, most of the simulation software is
> only available for PC's. As a Linux user, I
> feel your pain. However, computers are so
> cheap these days that it is actually
> worthwhile to buy a PC just to run blackjack
> software. That is the only reason that I
> still have a Windows system, and a Windows
> partition on my laptop.

So, the advantage comes from more blackjacks, more successful doubles and splits, and dealers busting more of their stiffs. It does seem that if this is being accomplished, one is winning more hands than he/she is losing. I do think I read somewhere that at certain TCs one actually does win more hands than one loses. I think it happens at around TC >= 4.

[newtobj]

> Understand that there is no free lunch. If
> you don't put the money down, you can't pick
> it up. So, your swings will be less drastic,
> but your profits will suffer commensurately.

> Your choice!

> Don

Then, I am probably using a too conservative bet spread. What would be a good bet spread to use for backcounting versus play all?

[Don Schlesinger]

> Then, I am probably using a too conservative
> bet spread. What would be a good bet spread
> to use for backcounting versus play all?

The spread doesn't have to be wide for back-counting, but the unit size has to be much bigger than for play all. The Chapter 10 charts in BJA3 will give you optimal unit sizes for different spreads.

Or, you can use CVCX or BJRM.

Don

[Don Schlesinger]

> I do think I read
> somewhere that at certain TCs one actually
> does win more hands than one loses. I think
> it happens at around TC >= 4.

Depends on the game and rules. It may happen at +4 for SD, but I think it's +6 or so for multi-deck, and, of course, if surrender is offered, we never win 50% of the hands.

Don

[newtobj]

> The spread doesn't have to be wide for
> back-counting, but the unit size has to be
> much bigger than for play all. The Chapter
> 10 charts in BJA3 will give you optimal unit
> sizes for different spreads.

> Or, you can use CVCX or BJRM.

> Don

Thank you Don.