thall: index #'s

[thall]

just curious, when the index for 12 vs dealer 3 shows (2) two, does it mean stand at 2 and above, or stand above 2. Also, for true count do we divide by number of decks left or the number of 1/2 decks left? (i've heard both)

[gorilla player]

> just curious, when the index for 12 vs
> dealer 3 shows (2) two, does it mean stand
> at 2 and above, or stand above 2. Also, for
> true count do we divide by number of decks
> left or the number of 1/2 decks left? (i've
> heard both)

That is an index I use. It means "stand at 2 or higher, hit at < 2...

I personally use deck accuracy in 6d games. I use 1/2 deck accuracy in 2d games unless the penetration is really good, once the first full deck is gone, I use the normal 1/4 deck accuracy for the remaining single deck, although many games won't go very far into that remaining deck before shuffling.

I don't think you would lose _that_ much by using deck accuracy until you get into the last deck, although I find multiplying by 2/3 to be not that hard if you have 1.5 decks left. Probably depends more on how well you can do the division/multiplication without screwing up your counting...

I do it that way because I find estimating to the nearest whole deck easy in a 6d game, but to the nearest 1/2 deck it gets pretty hard (for me) anyway. But once I get to 2d, going to 1/2 deck resolution is not hard, and for one deck, 1/4 deck accuracy is also not hard. A little practice will go a long way.

I have a bunch of estimation decks. I took 2 decks, clamped them together with wax paper between the cards and the vise, and touched each edge of the deck with a drop of thin CA glue (available at your local hobby show, they have thick and thin, you want thin.) It will turn that 2 deck stack into a solid "block. Do this for 1d, 2d, 3d, 4d, etc. Then for 1.5, and finally for .75, .5 and .25. I got a bunch of decks from a friend that is a security supervisor at a nameless casino. Now I have all these "blocks" that are accurately sized. I made sure to put a deuce on the bottom so I have plenty of room to use a big marker to write "4" or "1.5". Now I can pull 'em out of a sack, one at a time, lay them down and then "name those decks". You can get pretty quick and accurate with not much work.

One warning, make sure your wife doesn't see one of the decks and pick it up to play cards. They will _not_ separate after the thin CA glue fires, which takes just a few seconds...

[Don Schlesinger]

> Also, for
> true count do we divide by number of decks
> left or the number of 1/2 decks left? (i've
> heard both)

I think, of all the questions that we get over and over again, and that are the most misunderstood, this, along with "how many hands should I play?" is the most common.

Despite "gorilla player's" answer, above, you need to understand the following: If there are, say, 2.5 decks left, and we reckon the TC by dividing by 2.5, we are dividing by WHOLE decks, not half decks!! Had we been dividing by number of half-decks remaining in that situation, we'd be dividing by 5, not 2.5. Do you understand the difference?

Reckoning TC by whole decks remaining can be done with whole-deck accuracy or half-deck accuracy, but neither is the same as reckoning TC by the number of half-decks remaining. Too many people just don't understand the difference.

For most level-one counts, you reckon TC by whole decks (although you still might make divisions by 1.5, 2.5, 3.5, etc.). For some level-two counts (such as RPC), you divide by half-decks remaining.

Don

[Sun Runner]

Short, quick, and concise; an excellent repsonse!

It is so mis-understood because the explanation you just gave is very hard to find in print and stated that clearly.

Thall should also know that if he does truly decide to count half deck, he will need to adjust his indexes down -'roughly' by half.

Snyder's HiLo Lite (which is true counted by half decks, and rounded) would give that 12v3 index a value of zero, not two. Add'ly 12v2 would carry an index of two, not four.

[Don Schlesinger]

> Short, quick, and concise; an excellent
> repsonse!

> It is so mis-understood because the
> explanation you just gave is very hard to
> find in print and stated that clearly.

Shamefully, it isn't even in BJA3! I should add it to the last chapter, where there are random questions. I'll see if we have room (looks like I might be able to squeeze it onto p. 386).

> Thall should also know that if he does truly
> decide to count half deck, he will need to
> adjust his indexes down -'roughly' by half.

> Snyder's HiLo Lite (which is true counted by
> half decks, and rounded) would give that
> 12v3 index a value of zero, not two. Add'ly
> 12v2 would carry an index of two, not four.

This is why half-deck TC reckoning is usually reserved for level-two and higher point counts. You lose some accuracy when you divide the smaller, level-one running counts by the larger, half-deck integers. Also, this is why indices for the RPC and Hi-Lo are so similar. RPC tags are basically just double those of Hi-Lo, and the TC divisor is also doubled (half decks instead of whole decks), leaving the indices relatively the same.

Don

[thall]

> Thall should also know that if he does truly
> decide to count half deck, he will need to
> adjust his indexes down -'roughly' by half.

> Snyder's HiLo Lite (which is true counted by
> half decks, and rounded) would give that
> 12v3 index a value of zero, not two. Add'ly
> 12v2 would carry an index of two, not four.

Now I am confused. I use RPC count(level 2), are you saying my indexes are doubled. Ex. 8 vs dealer 6 index says 6/ been using this index for years, are you saying it should be 3(please say NOT)...

[Sun Runner]

> Now I am confused. I use RPC count(level 2),
> are you saying my indexes are doubled. Ex. 8
> vs dealer 6 index says 6/ been using this
> index for years, are you saying it should be
> 3(please say NOT)...

I don't have a working knowledge of RPC, Don certainly does, but I'll guess that waiting to double 8v6 @ TC of 6 seems high to me -unless you are TC'ing in whole decks (divisor equals 1.5, 2.0, 2.5, etc). The Hilo index for 8v6 is 2.

If RPC should be TC'ed in half decks (and I'm thinking it does), the same divisor above s/b 3,4, or 5.

Question .. to arrive at your TC, if 3 decks are left unseen, what have you been using as your divisor? Also, where did you get your indices?

Hopefully someone more learned than I will pitch in for you.

[Don Schlesinger]

> Now I am confused. I use RPC count(level 2),
> are you saying my indexes are doubled.

Your indexes are what they are. where did you get them from?

> Ex. 8
> vs dealer 6 index says 6/ been using this
> index for years, are you saying it should be
> 3(please say NOT)...

If you are dividing by whole decks to get TC, which isn't the way the RPC is designed to be used, you may have 6, although 4 or 5 would make more sense. You didn't specify how many decks were being shuffled.

For TC by half decks, I use +2.

Don

[thall]

> Your indexes are what they are. where did
> you get them from?

> If you are dividing by whole decks to get
> TC, which isn't the way the RPC is designed
> to be used, you may have 6, although 4 or 5
> would make more sense. You didn't specify
> how many decks were being shuffled.

My apology, I was mistakinly using my HiOpt index # in my RPC example. RPC index shows 3 with mult decks

> Don

[thall]

> Your indexes are what they are. where did
> you get them from?

I am looking at an older BJA of mine and see Don's Illustrious 18 and Fab 4. Are these index #'s based on whole or half deck? I'm guesing half...

[Don Schlesinger]

> I am looking at an older BJA of mine and see
> Don's Illustrious 18 and Fab 4. Are these
> index #'s based on whole or half deck? I'm
> guesing half...

Those are standard, typical Hi-Lo indexes, and Hi-Lo indexes are always constructed and intended to be used with whole-deck conversion of RC to TC.

Don

[thall]

> Those are standard, typical Hi-Lo indexes,
> and Hi-Lo indexes are always constructed and
> intended to be used with whole-deck
> conversion of RC to TC.
Thanks, now for a question I asked on the main page. Would Casino Verite give me the correct indices for RPC while correcting errors in BS, BS deviations, count, and bets. I'm needing some re-assurance after a long lay-off from bj. Leaving on a trip to play in about 10 days and intend to order today if it does.

[gorilla player]

> I think, of all the questions that we get
> over and over again, and that are the most
> misunderstood, this, along with "how
> many hands should I play?" is the most
> common.

> Despite "gorilla player's" answer,
> above, you need to understand the following:
> If there are, say, 2.5 decks left, and we
> reckon the TC by dividing by 2.5, we are
> dividing by WHOLE decks, not half decks!!
> Had we been dividing by number of half-decks
> remaining in that situation, we'd be
> dividing by 5, not 2.5. Do you understand
> the difference?

My answer was based on the idea "TC = running count divided by decks remaining."

My answer was also based on the idea of "when you divide by decks remaining, do you use whole-deck resolution or half-deck resolution."

IE since I am a "HiLo'er" my perception is based on the HiLo TC concept. Never tried any others, and so have never thought about the idea that some counts might divide by number of half-decks as opposed to dividing by number of whole-decks remaining.

If my answer was imprecise, it was more due to ignorance of other counting systems that have a different concept of "true count" (a really bad idea IMHO as it leads to different definitions of the same term, something that I occasionally see in computer science and something which I generally hate because it causes more confusion than good...

When I play a 6d game, I divide RC by 6,5,4,3, or 2, in general. If I am lucky enough to find a 6d game with 1d cut off, I might do the last two decks to 1/2 deck resolution and divide by 2, 3/2 or 1. Obviously for DD I use 1/2D resolution as I mentioned. I haven't played SD in a long while now thanks to 6:5, but on those few occasions where I did earlier, I divided by 1, 3/4 or 1/2 and would love to have the opportunity to divide by 1/4. And if a casino would make the offer, I'd even try 1/8 and learn to estimate that accurately.

> Reckoning TC by whole decks remaining can be
> done with whole-deck accuracy or half-deck
> accuracy, but neither is the same as
> reckoning TC by the number of half-decks
> remaining. Too many people just don't
> understand the difference.

Some of us didn't even know there +was+ a difference. IE I have read lots of BJ books, but I have _not_ tried to read every different BJ strategy book. IE I have Canfield's book and quickly got tired of being told how it was so much better than all the other counting systems, and how a narrow spread was good enough to beat a 6D shoe, etc...

Fortunately BJA didn't try to do that.

> For most level-one counts, you reckon TC by
> whole decks (although you still might make
> divisions by 1.5, 2.5, 3.5, etc.). For some
> level-two counts (such as RPC), you divide
> by half-decks remaining.

> Don

Is there any significant gain in using 1/2 deck resolution in (say) the first half of a 6d shoe? Not that I want to divide by 11/2 of course... But I find that not very hard (mult by 2, divide by 11, using what a computer scientist would call "integer math" which truncates...)