2D HiLo Index Numbers + Odd Question

My humblest apologies for the duplicate post (I asked a similar question a couple of months ago), but I have yet to get a complete set of 2D indices.

I have Professional Blackjack, Blackjack Attack, 200-Proof Blackjack, and Griffin's Theory of Blackjack--all of which are very helpful, but none of which contains a complete table for 2D.

Specifically, the rules are 2D, H17, DAS, RSA.

I know the I18 for this set of rules and am currently interpolating the 1D and 4D indices for the rest. This is past the point of diminishing marginal returns, I realize, but it still wouldn't hurt to have the complete, non-interpolated indices. The way I see it, extra memorization (even to the point of composition-dependent numbers) isn't difficult and so is worth even a marginal gain to me.

I'm writing this on my Palm Pilot as I sit in my casino hotel room, so , unfortunately, telling me to get Casino Verite and running sims won't do me much good at the moment.

An odd question:
In completely unrelated matters, could someone explain why splitting 9-9 v. ace is correct with 1D, H17, and DAS? The data on p. 75 of Griffin's ToB doesn't support this decision. Even when the effects of removing two 9s and one ace are taken into account, splitting appears to be worse than standing by a small margin. I would never question Griffin's analysis, so I wonder what I'm missing.

In any event, thanks to everyone who posts to this board. I've almost always found the answer to whatever BJ questions I've had in the past.

An odd question:
In completely unrelated matters, could someone explain why splitting 9-9 v. ace is correct with 1D, H17, and DAS? The data on p. 75 of Griffin's ToB doesn't support this decision. Even when the effects of removing two 9s and one ace are taken into account, splitting appears to be worse than standing by a small margin. I would never question Griffin's analysis, so I wonder what I'm missing.

You are not missing anything. According to Cacarulo's charts on www.bjmath.com, the correct decision is to stand:

Stand: -0.18613
Split: -0.19123

However, this makes for an interesting cover play. The frequency of this hand is 0.000362, therefore the overall affect to the house advantage is only 0.003%. On the other hand, splitting increases your variance.

I don't have the index for this play with me, but is it possible that it is 0 and you think that makes it basic strategy? If that's the case, you have to consider that the count would be -1 because of the ace.

Thanks for the response--I really appreciate it.

Griffin's ToB Basic Strategy Chart (p. 176, Note "a") says to split 9-9 v. ace with 1D, DAS, H17.

Blackjack Attack 3rd Edition (p. 481) says to split with these rules as well.

I don't doubt that this is the correct decision with this specific rule set. My question is why, when the data on p. 75 of ToB seems to suggest otherwise, it is the correct play.

Many basic strategy charts give recommendations based on an amalgam of rules (such is those at www.bjmath.com), so are not always correct for every combination. Griffin's and Schlesinger's are very precise, so splitting 9-9 v ace here is almost definitely correct. I just want to understand why.

I think an answer to my question would require someone who has Griffin's ToB to consult.

The tables on bjmath don't have DAS for 1D, so my previous comments were incorrect. (I've made this same mistake before, you would think I would have learned.)

I finally bit the bullet and created a fairly complete high-low strategy chart for 2D, H17, DAS.

I used Schlesinger's I18 when possible. For other indices, I took the mean of the relevant 1D and 4D indices from the appendix of the 1981 edition of Professional Blackjack. If PB gave the same values for 1D and 4D and differed from the I18, I used PB. If rounding was necessary, I leaned towards the more conservative decision (don't double, don't split, towards 0 for hit/stand).

The table is for high-low with no ace adjustment, with values from -3 to +8. It is primarily total-dependent, but I included some composition-dependent numbers when I (subjectively) thought they were warranted.

+0 may be treated like +0.5. -0 may be treated as -0.5.

The table may be found at http://www.gorillablackjack.com

Any comments--especially corrections--would be welcome.

David

When'd you start posting again? Good to see you!

"I think an answer to my question would require someone who has Griffin's ToB to consult."

The answer is very simple, and I'm surprised it didn't occur to you. Griffin's EORs aren't for DAS! The ones you're looking at, on p. 75, are for NDAS.

To find the DAS ones, see BJA3, p. 516, and consult the second set of splitting tables, below the NDAS one.

Don

You can get free software that generates strategy indices (I think using Griffin's EORs) at http://www.blackjackforumonline.com . You can compare numbers.

I don't know what you mean by "+0 may be treated like +0.5. -0 may be treated as -0.5." Why wouldn't you treat zero as zero? If you are truncating, so both +0.5 andd -0.5 are treated as zero, that might not be the best choice.

ETF

Thank you very much for the link--I really appreciate it.

Regarding +0 and -0, when interpolating some of the index numbers, I would sometimes get a result of +0.5 or -0.5. Though I could have truncated, floored, or rounded these values, I thought slightly more accuracy (probably not enough to make any difference) would be achieved by keeping them as +0.5 and -0.5. Since decimal values looked ugly on the chart, I just used +0 or -0, with the understanding that +0 was a little more than 0 and -0 was a little less than 0.

Thanks again for response.

David

That explains it--I'll finally be able to sleep at night.

The thought had crossed my mind, but looking at how small the gains are from doubling 11 against ace, and the fact that it would only be an opportunity on 4/49 of split 9s, made me discount it as the reason, much to my chagrin and embarrassment now.

Thanks also for writing such an outstanding book. I just got Blackjack Attack a couple of weeks ago and have already found it indispensable. Particularly useful so far are the sections on SCORE and the effects of deviating from correct play. The "Ultimate Gambit" in Burning the Tables in Las Vegas certainly seems to be influenced by your work. I also started reading the chapter on the floating advantage, and, though I'm not sure it will make me any money, it is fascinating.

By the way, any idea of the effect of RSA on the SCORE for 2D (62/104), H17, DAS? I realize it's not a particularly valuable rule, especially with one or two decks, but I am curious.

The last time I posted something, bj21's hard disk crashed and everything was lost!

"That explains it--I'll finally be able to sleep at night."

And, I'm cheaper than Ambien! :-)

"The thought had crossed my mind, but looking at how small the gains are from doubling 11 against ace, and the fact that it would only be an opportunity on 4/49 of split 9s, made me discount it as the reason, much to my chagrin and embarrassment now."

No problem.

"Thanks also for writing such an outstanding book. I just got Blackjack Attack a couple of weeks ago and have already found it indispensable."

So, what were you doing before last week? :-)

"Particularly useful so far are the sections on SCORE and the effects of deviating from correct play. The "Ultimate Gambit" in Burning the Tables in Las Vegas certainly seems to be influenced by your work."

Yes, Ian consulted that section and incorporated some of the findings into his UG. He was always very generous, however, with attributions, as I always try to be, and I was happy to be cited in his work.

"I also started reading the chapter on the floating advantage, and, though I'm not sure it will make me any money, it is fascinating."

Well, at the time, it was a monumental project, involving the biggest computer study (John Gwynn) that had ever been done to date to analyze a specific BJ problem. As you see, the results were more of theoretical value than practical, but, every once in a while, we find a very deeply dealt game, and the FA kicks in, so it's still good to be aware of the concept.

"By the way, any idea of the effect of RSA on the SCORE for 2D (62/104), H17, DAS?

Well, the DAS part doesn't enter in, because you can't double after you resplit; you still receive only one card.

"I realize it's not a particularly valuable rule, especially with one or two decks, but I am curious."

It's more valuable than you may imagine, especially with larger spreads. Although you don't get to make the play very frequently, the ability, especially with a max bet, to break up your 2 or 12 and turn it into two max bets of starting 11, as if you were doubling the original hand, instead of starting with 2 or 12, is quite valuable.

Can't answer your question, however, because it depends very heavily on the spread being used.

Don