Don Schlesinger: BJA2 Corrections

[Don Schlesinger]

How's this for service??!

Some of my comments are actually directed to Arnold Snyder, hence the conversational tone.

Most of the technical, typographical stuff you won't care about at all. But, mixed in are some important things.

I'm going to post this to Don's Domain, as well.

Thanks for the suggestion.

Corrections to BJA2:

1. In the Table of Contents, the Preface should be p. xvii, and not p. xiv.

2. I think you may not have fixed the kerning on the y's in the very early pages, before the first chapter. If you look at the Acknowledgments and other sections, you'll see some examples of the old problem.

3. p. xix. "Highly-regarded Chapter 10" doesn't need the hyphen. I eliminated it in such constructions on other pages.

4. P. 139. Picture is too high.

5. P. 140. The negative sign in the exponent of the first formula should be an en dash, not a hyphen.

6. P. 195. Next time John changes the chart, we need en dashes for all the negative signs.

7. Marcus's BJ 678 actually has raised dots between the 67 and the 78. That's the way he writes it. I used hyphens (6-7-8) on pp. 191 and 342, and then, to compound the problem, you put 6,7,8 in the Index! Remember, we even commented on how it looked funny, coming after the word "Blackjack"? I should have picked up on that, but I let it go by.

8. P. 295. This is a bad one. It wasn't in the original article, but when you retyped the charts for SCORE, Table 11.21, second line, reads "TC?RC," instead of "TC/RC."
You hit the caps key, and the "?" is on top of the "/." But, I missed it, in proofreading, and so did everyone else.

9. P. 312. There seem to be some stray marks in the last full paragraph. Wonder if this is just my copy or everywhere.

10. P. 23. The (h-1) terms, to the right, should be en dashes, and all of the TCs, to the left, should be en dashes.

11. P. 50. Bad one. Next-to-last line of first paragraph: "your true count" doesn't start with a capital!!

12. P. 153. All my fault, since the ORIGINAL article in BJF! Hard to believe. Wrong in the article, wrong in both editions, and NO ONE saw it. Last paragraph, right in the middle: "At the end of 1,000 hours" should be "1,000 HANDS"! How could I have possibly missed that????

13. P. 317. 12 v. 5 index, for the EV, first column, lost its negative sign!!! It was there, in the original.

14. Pp. 3, 6, 12, and 13. All have small y's.

15. P. 116. Picture is too large.

16. P. 306. En dashes for bet spreads.

17. In BJF article, Fall 2000, p. 19, change Carnegie Mellon Institute to University.

18. Dedication. Extra space: "memory of four"?

19. P. 34. Change rounded percentages to 6.55% and 10.07%.

20. P. 275. Page references are 18 numbers too high (weird!).

21. Chapter 10, 8-deck corrections from Norm Wattenberger:

There appear to be a few minor errors in the Chapter 10 eight deck tables.

For example, on page 254 if you take the TC frequency column, multiply by
the %W/L column and multiply by the sim bets, divide by 100, you should get
the average %W/L. This works for 1, 2 and 6 deck tables but not eight deck
tables. The problem appears to be one number in each table. The EV %W/L for
TC <=0 is incorrect. The number that you have is the %W/L for a TC of 0, not
TC <=0. The results all appear to be correct.

There also appears to be a problem with the tables on pages 268 and 269. The
average bets don't come out correctly. It appears that the values that you
have for a spread of 1-8 are the values for 1-10 and the values for 1-10 are
the values for 1-12. I don't know about the other results. Fortunately,
eight deck H17, DAS, LS has got to be a rare game.

22. P. 345, surrender pages should be 71-73, not 71-74.

Of course, the above are in addition to the things we already know, to wit:

1. "way" instead of "Way," on spine of book cover.

2. Need to center spades on back cover, and, in my opinion, change the blurbs font, which is very hard to read (too narrow).

3. If we redo Chapter 10 charts, use en dashes and correct "greater than or equal to" and "less than or equal to" signs everywhere.

4. Pp. 328-9. Center title of "Spread 2" in all boxes, by spacing once to the right.

5. For Chapters 6 and 10, need quotes around the chapter titles on each odd-numbered page.

6. P. 270. Picture much too dark.

When I tell you that dozens upon dozens of hours were spent by numerous people proofreading, you can imagine my frustration. After all of the above is corrected, let's hope the third edition will be relatively error-free (anyone wanna bet?!).

Don

[Bob]

Thanks Don, I'll make the "Important" corrections, and it was entertaining reading some of the others.

I'm not sure but I think I found (at least) one more....

Page 317, Table 12.10, Float Win/100 column:
Bottom right corner entry under Float Win/100, "$61.55" doesn't seem to be correct. Shouldn't it be higher than "$61.90" under EV Maximizing 4 lines above....Reversed Maybe???

Also, same Table...Risk Averse, Back Count, Float Bank Number "$9450"...shouldn't it be less than $9310 two lines below.

If I'm wrong...no long explanations necessary!
Take Care, Bob

> How's this for service??!

> Some of my comments are actually directed
> to Arnold Snyder, hence the conversational
> tone.

> Most of the technical, typographical stuff
> you won't care about at all. But, mixed in
> are some important things.

> I'm going to post this to Don's Domain, as
> well.

> Thanks for the suggestion.

> Corrections to BJA2: 1. In the Table of
> Contents, the Preface should be p. xvii, and
> not p. xiv.

> 2. I think you may not have fixed the
> kerning on the y's in the very early pages,
> before the first chapter. If you look at the
> Acknowledgments and other sections, you'll
> see some examples of the old problem.

> 3. p. xix. "Highly-regarded Chapter
> 10" doesn't need the hyphen. I
> eliminated it in such constructions on other
> pages.

> 4. P. 139. Picture is too high.

> 5. P. 140. The negative sign in the
> exponent of the first formula should be an
> en dash, not a hyphen.

> 6. P. 195. Next time John changes the
> chart, we need en dashes for all the
> negative signs.

> 7. Marcus's BJ 678 actually has raised dots
> between the 67 and the 78. That's the way he
> writes it. I used hyphens (6-7-8) on pp. 191
> and 342, and then, to compound the problem,
> you put 6,7,8 in the Index! Remember, we
> even commented on how it looked funny,
> coming after the word "Blackjack"?
> I should have picked up on that, but I let
> it go by.

> 8. P. 295. This is a bad one. It wasn't in
> the original article, but when you retyped
> the charts for SCORE, Table 11.21, second
> line, reads "TC?RC," instead of
> "TC/RC."
> You hit the caps key, and the "?"
> is on top of the "/." But, I
> missed it, in proofreading, and so did
> everyone else.

> 9. P. 312. There seem to be some stray
> marks in the last full paragraph. Wonder if
> this is just my copy or everywhere.

> 10. P. 23. The (h-1) terms, to the right,
> should be en dashes, and all of the TCs, to
> the left, should be en dashes.

> 11. P. 50. Bad one. Next-to-last line of
> first paragraph: "your true count"
> doesn't start with a capital!!

> 12. P. 153. All my fault, since the
> ORIGINAL article in BJF! Hard to believe.
> Wrong in the article, wrong in both
> editions, and NO ONE saw it. Last paragraph,
> right in the middle: "At the end of
> 1,000 hours" should be "1,000
> HANDS"! How could I have possibly
> missed that????

> 13. P. 317. 12 v. 5 index, for the EV,
> first column, lost its negative sign!!! It
> was there, in the original.

> 14. Pp. 3, 6, 12, and 13. All have small
> y's.

> 15. P. 116. Picture is too large.

> 16. P. 306. En dashes for bet spreads.

> 17. In BJF article, Fall 2000, p. 19,
> change Carnegie Mellon Institute to
> University.

> 18. Dedication. Extra space: "memory
> of four"?

> 19. P. 34. Change rounded percentages to
> 6.55% and 10.07%.

> 20. P. 275. Page references are 18 numbers
> too high (weird!).

> 21. Chapter 10, 8-deck corrections from
> Norm Wattenberger:

> There appear to be a few minor errors in
> the Chapter 10 eight deck tables.

> For example, on page 254 if you take the TC
> frequency column, multiply by
> the %W/L column and multiply by the sim
> bets, divide by 100, you should get
> the average %W/L. This works for 1, 2 and 6
> deck tables but not eight deck
> tables. The problem appears to be one
> number in each table. The EV %W/L for
> TC TC There also appears to be a
> problem with the tables on pages 268 and
> 269. The
> average bets don't come out correctly. It
> appears that the values that you
> have for a spread of 1-8 are the values for
> 1-10 and the values for 1-10 are
> the values for 1-12. I don't know about the
> other results. Fortunately,
> eight deck H17, DAS, LS has got to be a
> rare game.

> 22. P. 345, surrender pages should be
> 71-73, not 71-74.

> Of course, the above are in addition to the
> things we already know, to wit:

> 1. "way" instead of
> "Way," on spine of book cover.

> 2. Need to center spades on back cover,
> and, in my opinion, change the blurbs font,
> which is very hard to read (too narrow).

> 3. If we redo Chapter 10 charts, use en
> dashes and correct "greater than or
> equal to" and "less than or equal
> to" signs everywhere.

> 4. Pp. 328-9. Center title of "Spread
> 2" in all boxes, by spacing once to the
> right.

> 5. For Chapters 6 and 10, need quotes
> around the chapter titles on each
> odd-numbered page.

> 6. P. 270. Picture much too dark.

> When I tell you that dozens upon dozens of
> hours were spent by numerous people
> proofreading, you can imagine my
> frustration. After all of the above is
> corrected, let's hope the third edition will
> be relatively error-free (anyone wanna
> bet?!).

> Don

[Don Schlesinger]

> Page 317, Table 12.10, Float Win/100
> column:
> Bottom right corner entry under Float
> Win/100, "$61.55" doesn't seem to
> be correct. Shouldn't it be higher than
> "$61.90" under EV Maximizing 4
> lines above....Reversed Maybe???

Lower bank and lower ROR for the former. Like comparing apples to oranges. We couldn't make the numbers comeout perfectly, so we had to alter the standard practice of keeping bank and ROR identical. the point is that, with r-a indices, you can win almost the same amount with less bank and lower ROR, or you can win more with same bank and ROR. We chose to display the former.

> Also, same Table...Risk Averse, Back Count,
> Float Bank Number
> "$9450"...shouldn't it be less
> than $9310 two lines below.

> If I'm wrong...no long explanations
> necessary!

Same idea as above. Admittedly, it's confusing and less than ideal.

I very much appreciate the feedback.

Don

[Bob]

That makes sense...I agree that there is no good "intuitive" way to list the results.

5 more questions, for February anyway...

Tables 12.1 & 12.2 are very applicable to the way that I Hi-Lo back-count 6 deck games, and look for other shuffling tables, but I rarely play where LS is offered. I am using your optimum departure numbers religiously to leave tables. Are the max departures of TC=-0.300 for 4.5/6.0 and TC=-0.839 for 5.0/6.0 very sensitive to the lack of surrender? What about x-8 bet spread vs. x-12 spread?

A general question, to see if I grasped the SCORE concept correctly, and how to apply it....I understand using SCORE to choose one count vs. another, a certain table/game vs. another, and that it could be used where heat from bet spread might be a problem. Other than that though, generally, is there more to what you meant by your statement, "I'm going to change my own approach"....(to the way you Wong games), under implications on page 283?

With my $100 max bet, my act and talkative personality allows my to back-count and play x-8,10 spreads without a problem in shoe games, while earning comps. The biggest playing application of SCORE for me is when to bounce into $25 minimum tables (x-4) and $50 tables (x-2) with my max bet of $100.

Using your example on pages 282-283, from the book alone, i.e. without software, I would have calculated your unit bet size as follows: 10,000/400 = 25, so from optimum bet/400 bank column Table 10.18 on page 205, $25*1.58= $40, not $43, and for 1-2 spread, entering at TC+2, $25*3.54 = $89, not $116. Is my ROR lower here than the 13.5% for your scenario, and is this causing your higher unit sizing? In any event, it is not clear to me when you would go to the 2x ($232) bet. I am not too cheap to buy the software, but I trust your book's interpretations and I don't like to generate numbers myself that I will second guess, or might be inaccurate. Is their a "best way" to use the SCORE information in the book, for x-2 and x-4 unit bet sizing and true count bet sizing, without using software, i.e. with Chapter 10? I would have assumed my "Optimum Bet" example above would yield a ROR of 13.5% or so.

Thanks Again Don.

> Lower bank and lower ROR for the former.
> Like comparing apples to oranges. We
> couldn't make the numbers comeout perfectly,
> so we had to alter the standard practice of
> keeping bank and ROR identical. the point is
> that, with r-a indices, you can win almost
> the same amount with less bank and lower
> ROR, or you can win more with same bank and
> ROR. We chose to display the former.

> Same idea as above. Admittedly, it's
> confusing and less than ideal.

> I very much appreciate the feedback.

> Don

[Don Schlesinger]

A little busy today and getting ready to watch Super Bowl stuff. Will try my best to answer you tomorrow.

Don

[Don Schlesinger]

> 5 more questions, for February anyway...

Hmmm. Not sure what the allowable quota is! :-)

> Tables 12.1 & 12.2 are very applicable
> to the way that I Hi-Lo back-count 6 deck
> games, and look for other shuffling tables,
> but I rarely play where LS is offered. I am
> using your optimum departure numbers
> religiously to leave tables. Are the max
> departures of TC=-0.300 for 4.5/6.0 and
> TC=-0.839 for 5.0/6.0 very sensitive to the
> lack of surrender? What about x-8 bet spread
> vs. x-12 spread?

They are slightly sensitive to rules variations, but not at all sensitive to spreads, as I mentioned in the article (see p. 306, "Appreciation of the Data"). I wouldn't worry too much about the differences.

> A general question, to see if I grasped the
> SCORE concept correctly, and how to apply
> it....I understand using SCORE to choose one
> count vs. another, a certain table/game vs.
> another, and that it could be used where
> heat from bet spread might be a problem.

So far, so good!

> Other than that though, generally, is there
> more to what you meant by your statement,
> "I'm going to change my own
> approach"....(to the way you Wong
> games), under implications on page 283?

Simply this. One way counters are detected is by a large spread. For play-all, it's an unfortunate prerequisite. But, many back-counters employ the same bet spread, once Wonging, that they do when playing-all. And, the data show that SCOREs for much smaller bet spreads (with progressively higher units, of course) do almost as well. So, the suggestion is to give up a little of the EV, by not spreading as widely, as a tradeoff for better longevity.

> With my $100 max bet, my act and talkative
> personality allows me to back-count and play
> x-8,10 spreads without a problem in shoe
> games, while earning comps. The biggest
> playing application of SCORE for me is when
> to bounce into $25 minimum tables (x-4) and
> $50 tables (x-2) with my max bet of $100.

Understood. In general, I think thew latter would be a very good idea.

> Using your example on pages 282-283, from
> the book alone, i.e. without software, I
> would have calculated your unit bet size as
> follows: 10,000/400 = 25, so from optimum
> bet/400 bank column Table 10.18 on page 205,
> $25*1.58= $40, not $43, and for 1-2 spread,
> entering at TC+2, $25*3.54 = $89, not $116.
> Is my ROR lower here than the 13.5% for your
> scenario, and is this causing your higher
> unit sizing? In any event, it is not clear
> to me when you would go to the 2x ($232)
> bet. I am not too cheap to buy the software,
> but I trust your book's interpretations and
> I don't like to generate numbers myself that
> I will second guess, or might be inaccurate.

Yes, as no units were given in the article, and as you get lower EVs with your $25 unit, you can assume that you would have a lower ROR and that the units used for the book illustrations (optimal betting) are much higher than $25. In fact, for the 1-4 spread, the correct unit (using BJRM) seems to be almost exactly . . . $100!! so, you bet $100 at +2, and eventually top out at $400 at +8. At +5, you're betting only about $250, so, for all intents and purposes, most of the time, this 1-4 spread is really looking much more like a very tame 1-2.5 spread!

> Is their a "best way" to use the
> SCORE information in the book, for x-2 and
> x-4 unit bet sizing and true count bet
> sizing, without using software, i.e. with
> Chapter 10? I would have assumed my
> "Optimum Bet" example above would
> yield a ROR of 13.5% or so.

Optimal betting does yield a 13.5% ROR, but unless you're lucky enough to have one of the schemes in Chapter 10 match up closely to give you that 13.5% ROR, you can't always surmise from the charts what the correct bet ramp is. Unashamedly, I admit that this was done by design so that John could -- deservedly so -- sell some software. You really owe it to yourself to get BJRM, or CVCX, or both.

> Thanks Again Don.

You're very welcome.

Don

[leepfrog]

> Hmmm. Not sure what the allowable quota is!
> :-)

> They are slightly sensitive to rules
> variations, but not at all sensitive to
> spreads, as I mentioned in the article (see
> p. 306, "Appreciation of the
> Data"). I wouldn't worry too much about
> the differences.

> So far, so good!

> Simply this. One way counters are detected
> is by a large spread. For play-all, it's an
> unfortunate prerequisite. But, many
> back-counters employ the same bet spread,
> once Wonging, that they do when playing-all.
> And, the data show that SCOREs for much
> smaller bet spreads (with progressively
> higher units, of course) do almost as well.
> So, the suggestion is to give up a little of
> the EV, by not spreading as widely, as a
> tradeoff for better longevity.

> Understood. In general, I think thew latter
> would be a very good idea.

> Yes, as no units were given in the article,
> and as you get lower EVs with your $25 unit,
> you can assume that you would have a lower
> ROR and that the units used for the book
> illustrations (optimal betting) are much
> higher than $25. In fact, for the 1-4
> spread, the correct unit (using BJRM) seems
> to be almost exactly . . . $100!! so, you
> bet $100 at +2, and eventually top out at
> $400 at +8. At +5, you're betting only about
> $250, so, for all intents and purposes, most
> of the time, this 1-4 spread is really
> looking much more like a very tame 1-2.5
> spread!

> Optimal betting does yield a 13.5% ROR, but
> unless you're lucky enough to have one of
> the schemes in Chapter 10 match up closely
> to give you that 13.5% ROR, you can't always
> surmise from the charts what the correct bet
> ramp is. Unashamedly, I admit that this was
> done by design so that John could --
> deservedly so -- sell some software. You
> really owe it to yourself to get BJRM, or
> CVCX, or both.

> You're very welcome.

> Don
HI CAN ANYBODY GIVE ME SOME REFERENCE ON CLUMPING THEORY ? THANKS

[Robert V. Lux]

Hi

My name is Robert V. Lux. What is the exact advantage when playing against different bj games, such as 1, 2, 4, 6 and 8 dekcgames, if you play perfect basicstrategy and play according to Las Vegas Strip rules?

Best regards; Robert V. Lux
Mail: [email protected]

[Don Schlesinger]

> My name is Robert V. Lux. What is the exact
> advantage when playing against different bj
> games, such as 1, 2, 4, 6 and 8 dekcgames,
> if you play perfect basicstrategy and play
> according to Las Vegas Strip rules?

We need the following information for each game: s17 or h17 (probably s17); das or ndas (probably das for 4, 6, and 8 only); lsr or ns (probably lsr on 4, 6, and 8 only).

Write back with exactly what you want, and we'll post the edges.

Don

[Bob]

Don,

Thanks for the information; the answers were clear. I will consider one or both of the software progams. About 5 years ago I stopped in GBC while visiting Las Vegas and tried to use Stanford's BJCA program and it gave me fits....I have all of Stanford's books and liked them all, so I was disappointed and dismissed my own software simulation use, instead trusting the books. Since then, I have been hesitant of all programs due to internal errors, user-friendly issues, and the like. Your book (1st edition) and Chapter 10 simulations rekindled a more serious interest and "trust" in counting, I have been a consistent winner ever since.

I'll consider the programs you mention. Where does SBA fit into the mix?

And one more question, I promise. If I want to memorize one set of Hi-Lo matrix numbers for backcounting shoe games, and very rarely playing 2 deck, should I use the Catch 22 R-A numbers, without exception, or would you go with E-V numbers with small bets out? i.e., should I just always use the R-A numbers or not? Memorizing 2 matrix numbers, especially for 10 vs. 10, is not a problem.

I am going on a cruise in the western caribbean this saturday for 7 days, and should be able to get in some decent time at the tables. With only fair penetration in 6 deck, the most interesting opportunity is a decent tournament that concludes at the end of the week. I took second in one of those last spring, with the help of Stanford's Tournament Strategy Book.

Thanks...Bob

> Hmmm. Not sure what the allowable quota is!
> :-)

> They are slightly sensitive to rules
> variations, but not at all sensitive to
> spreads, as I mentioned in the article (see
> p. 306, "Appreciation of the
> Data"). I wouldn't worry too much about
> the differences.

> So far, so good!

> Simply this. One way counters are detected
> is by a large spread. For play-all, it's an
> unfortunate prerequisite. But, many
> back-counters employ the same bet spread,
> once Wonging, that they do when playing-all.
> And, the data show that SCOREs for much
> smaller bet spreads (with progressively
> higher units, of course) do almost as well.
> So, the suggestion is to give up a little of
> the EV, by not spreading as widely, as a
> tradeoff for better longevity.

> Understood. In general, I think thew latter
> would be a very good idea.

> Yes, as no units were given in the article,
> and as you get lower EVs with your $25 unit,
> you can assume that you would have a lower
> ROR and that the units used for the book
> illustrations (optimal betting) are much
> higher than $25. In fact, for the 1-4
> spread, the correct unit (using BJRM) seems
> to be almost exactly . . . $100!! so, you
> bet $100 at +2, and eventually top out at
> $400 at +8. At +5, you're betting only about
> $250, so, for all intents and purposes, most
> of the time, this 1-4 spread is really
> looking much more like a very tame 1-2.5
> spread!

> Optimal betting does yield a 13.5% ROR, but
> unless you're lucky enough to have one of
> the schemes in Chapter 10 match up closely
> to give you that 13.5% ROR, you can't always
> surmise from the charts what the correct bet
> ramp is. Unashamedly, I admit that this was
> done by design so that John could --
> deservedly so -- sell some software. You
> really owe it to yourself to get BJRM, or
> CVCX, or both.

> You're very welcome.

> Don

[Don Schlesinger]

> I'll consider the programs you mention.
> Where does SBA fit into the mix?

It's a wonderful product, but somewhat harder to use than the others. To say you have the best and most completely accurate simulators and practice tools on the market today, I would suggest CVCX and CVData, SBA, Blackjack 6-7-8, and, of course, BJRM (which, technically,is not a simulator).

> And one more question, I promise. If I want
> to memorize one set of Hi-Lo matrix numbers
> for backcounting shoe games, and very rarely
> playing 2 deck, should I use the Catch 22
> R-A numbers, without exception, or would you
> go with E-V numbers with small bets out?

It really doesn't matter. The p. 195 (or 317) indices are perfectly fine for your needs. If, instead, you use the Catch 22 ones (be careful about the 12 v. 5 EV error on p. 317!), that's OK, too.

> i.e., should I just always use the R-A
> numbers or not? Memorizing 2 matrix numbers,
> especially for 10 vs. 10, is not a problem.

The point is, if you use the R-A ones, you should run sims with them so that you are betting accurately and using the bankroll that makes sense for that set of indices, rather than the EV-maximizing ones. But, as you see from the article, the differences are so slight that it hardly matters.

> I am going on a cruise in the western
> caribbean this saturday for 7 days, and
> should be able to get in some decent time at
> the tables. With only fair penetration in 6
> deck, the most interesting opportunity is a
> decent tournament that concludes at the end
> of the week. I took second in one of those
> last spring, with the help of Stanford's
> Tournament Strategy Book.

Usually, because of the crowds, BJ conditions on ships are pretty poor, so resign yourself to simply having fun, and if you should also happen to win a few bucks, that will be a nice bonus!

Good luck!

Don