counter trucker: question for mr. Schlesinger

[counter trucker]

Mr. Schlesinger
I`ve discovered some differences between index numbers in Professional Blackjack by S.Wong (p.46thru 49 tables of the last edition) and yours Illustrious 18. Namely:
Pro BJ: 13 against 2 is 0

12 against 5 is -1 for both h/s 17

12 against 6 is 0 for s17, -3 for h/17

BJ Attack: 13 against 2 is -1

12 against 5 is -2

12 against 6 is -1
Then, as a source of index numbers, you wrote that these come from Pro. BJ. Which indexes are correct, please?

[Don Schlesinger]

> Mr. Schlesinger
> I`ve discovered some differences between
> index numbers in Professional Blackjack by
> S.Wong (p.46thru 49 tables of the last
> edition) and yours Illustrious 18. Namely:
> Pro BJ: 13 against 2 is 0

> 12 against 5 is -1 for both h/s 17

> 12 against 6 is 0 for s17, -3 for h/17

> BJ Attack: 13 against 2 is -1

> 12 against 5 is -2

> 12 against 6 is -1
> Then, as a source of index numbers, you
> wrote that these come from Pro. BJ. Which
> indexes are correct, please?

Wong's switch of his indices, for the 1994 edition of Pro BJ, has caused nothing but confusion and aggravation for readers who were around for the earlier editions of his book. Those first indices were rounded, whereas the later ones were truncated. The BJA3 indices are all floored.

We all wish Wong would have left things the way they once were.

Don

[HC]

> Wong's switch of his indices, for the 1994
> edition of Pro BJ, has caused nothing but
> confusion and aggravation for readers who
> were around for the earlier editions of his
> book. Those first indices were rounded,
> whereas the later ones were truncated. The
> BJA3 indices are all floored.

> We all wish Wong would have left things the
> way they once were.

> Don

So what would be the difference in playing strategy if i follow either one? does it really matter practically?? i am quite confused because i am currently using Wong's Pro BJ indices.

[Don Schlesinger]

> So what would be the difference in playing
> strategy if i follow either one? does it
> really matter practically?? i am quite
> confused because i am currently using Wong's
> Pro BJ indices.

The important thing is to USE the indices in the same manner that they've been generated. So, if Wong truncated, then YOU have to truncate when calculating the TC. If your RC is -3, and two decks remain, your TC of -1.5 has to be called -1 (truncated) and not -2 (floored).

Someone using indices that are floored would call the TC -2. And, you would compare those TCs to the indices you use and play accordingly.

Don

[MJ]

Why did Wong decide to truncate for the 94 edition? Perhaps he figured it is easier to truncate then it is to round(which is quite true). Is there some type of advantage for truncating the TC as opposed to rounding in terms of EV?

-MJ

> Wong's switch of his indices, for the 1994
> edition of Pro BJ, has caused nothing but
> confusion and aggravation for readers who
> were around for the earlier editions of his
> book. Those first indices were rounded,
> whereas the later ones were truncated. The
> BJA3 indices are all floored.

> We all wish Wong would have left things the
> way they once were.

> Don

[Don Schlesinger]

> Why did Wong decide to truncate for the 94
> edition?

You'll have to ask him. I think he must regret it now.

> Perhaps he figured it is easier to
> truncate than it is to round (which is quite
> true).

Maybe. Not sure.

> Is there some type of advantage for
> truncating the TC as opposed to rounding in
> terms of EV?

No. And, it creates a double interval around zero, which contains as much as 45% of all the hands, in a shoe game, which is not good at all.

Don

[MJ]

> No. And, it creates a double interval around
> zero, which contains as much as 45% of all
> the hands, in a shoe game, which is not good
> at all.

I see what you mean with the double interval around zero(-.99 to .99 truncated is TC = 0 ). However, the counter will only have his minimum bet out in this interval anyhow. So, in the grand scheme of things, it doesn't seem as though this would affect SCORE very much(although I could be wrong).

Perhaps rounding is a little bit more powerful then truncating(at least in theory). But how much of a difference can it really make to a counter's bottom line? That is the question which any Hi-Lo counter should ask. At least with truncation the division can be peformed much faster then rounding! It just makes Hi-Lo that much easier to use. If it turns out I need to sacrifice a few hundreths of a percentage point for this ease of use, it is well worth the trade off.

Who knows in ACTUAL casino play truncation might even have a HIGHER EV as fewer mistakes will be made and less mental fatigue will occur then would be with rounding the TC to the nearest integer.

In closing, Stanford Wong must have had a good reason for his decision to truncate as opposed to round, for SW is a smart man! If there was any significant difference in EV, I'm sure he would have stuck with rounding or at least provided the indices for both approaches and left it up to the reader to decide which method to use.

It would be great if somebody ran a SIM and compared the PB 94 indices vs the BJA3 indices using each author's respective approach. That would settle it once and for all which approach yields higher EV(flooring or truncating).

-MJ

[Don Schlesinger]

> It would be great if somebody ran a SIM and
> compared the PB 94 indices vs the BJA3
> indices using each author's respective
> approach. That would settle it once and for
> all which approach yields higher EV (flooring
> or truncating).

With all due respect, we did it our way for a reason. And, you ought to be able to guess what that reason is, no?

Don

[Norm Wattenberger]

1. The 1981 version of PBJ used flooring, not rounding. It was changed to truncation in 1994. The 1994 version clearly states truncation is used. The 1981 book does not explain how to calculate the TC. (In fact few BJ books do.) I don't know the reason behind the 1981 use of flooring; but it may simply have been an error.

2. PBJ indexes were generated using 'representative subsets' and assuming infinite decks. This is a very fast method, but not accurate. Computer time was a lot more expensive then. The errors introduced by this methodology far outweigh the TC calculation technique.

3. Despite this, the indexes are not that bad. The gain from using newly generated EV maximized indexes would be about 1% in a good penetration SD game.

4. Of course you could add to that gain by generating RA indexes.

5. Personally, I prefer to round. I simply find it easier and the difference between rounding and flooring, assuming the indexes were generated in the respective manner, is tiny.

Serious Blackjack Software