Results 1 to 13 of 13

Thread: Alexost: RPC TC precision

  1. #1
    Alexost
    Guest

    Alexost: RPC TC precision

    I've been using CVCX simming RPC with different numbers of indices etc. (mostly for single deck). When I ran sims with 1/2 deck precision and 1/4 deck precision to compare how much efficiency is lost using 1/2 instead of the recommended 1/4 deck precision for single deck... I get a lower score using 1/4 deck precision. My guess is since RPC indices in PBAAB are calculated by dividing the RC by 1/2 decks remaining for TC conversion, applying the 1/4 deck precision is not as effective. So, is it better to use LESS precision (1/2 deck) when using RPC? Another reason I have to ask is because in PBJAAB and the blue supplement Revere clearly shows calculating TC when one deck remains with 1/4 deck precision. Is this a mistake in his publications? Hope this makes sense.

    Thanks!

  2. #2
    Cacarulo
    Guest

    Cacarulo: Re: RPC TC precision

    > I've been using CVCX simming RPC with different
    > numbers of indices etc. (mostly for single deck). When
    > I ran sims with 1/2 deck precision and 1/4 deck
    > precision to compare how much efficiency is lost using
    > 1/2 instead of the recommended 1/4 deck precision for
    > single deck... I get a lower score using 1/4 deck
    > precision. My guess is since RPC indices in PBAAB are
    > calculated by dividing the RC by 1/2 decks remaining
    > for TC conversion, applying the 1/4 deck precision is
    > not as effective. So, is it better to use LESS
    > precision (1/2 deck) when using RPC? Another reason I
    > have to ask is because in PBJAAB and the blue
    > supplement Revere clearly shows calculating TC when
    > one deck remains with 1/4 deck precision. Is this a
    > mistake in his publications? Hope this makes sense.

    Are you calculating the indices with the same precision for each case (1/2 and 1/4)?

    Sincerely,
    Cac

  3. #3
    Alexost
    Guest

    Alexost: Re: RPC TC precision

    > Are you calculating the indices with the same
    > precision for each case (1/2 and 1/4)?

    No, I am using indices calculated by dividing by half
    deck-and then applying 1/2 deck and 1/4 deck precision
    when converting to TC.

    My assumption is that this is OK to do, at least when
    comparing to Hi-Lo sims. Even though I generated Hi-Lo
    indices reckoned by full decks, the score still
    increases every time I used more precision to convert
    the RC to the TC.

    In other words, even though standard Hi-Lo indices are generated using full deck reckoning, score increases the more precise you convert from RC to TC. ie: Divide by full decks, half decks, quarter decks, exact deck resolution.

    So my thought process was that no matter how you
    calculate the indices, (dividing by full or half deck
    most common) you would still get a better score the
    more precise you were when converting the RC to the
    TC.

    Does this help? Thanks!


  4. #4
    Cacarulo
    Guest

    Cacarulo: Re: RPC TC precision

    > No, I am using indices calculated by dividing by half
    > deck-and then applying 1/2 deck and 1/4 deck precision
    > when converting to TC.

    > My assumption is that this is OK to do, at least when
    > comparing to Hi-Lo sims. Even though I generated Hi-Lo
    > indices reckoned by full decks, the score still
    > increases every time I used more precision to convert
    > the RC to the TC.

    > In other words, even though standard Hi-Lo indices are
    > generated using full deck reckoning, score increases
    > the more precise you convert from RC to TC. ie: Divide
    > by full decks, half decks, quarter decks, exact deck
    > resolution.

    > So my thought process was that no matter how you
    > calculate the indices, (dividing by full or half deck
    > most common) you would still get a better score the
    > more precise you were when converting the RC to the
    > TC.

    > Does this help? Thanks!

    Yes, it helps but I don't fully agree with your thought process. Indices should be calculated in the same way you're going to use them.
    However, there are some interesting posts about using a higher number of decks that you may find in the main forum in a thread that compares Zen vs Halves. It has to do with what Norman calls "TC compression". Maybe you should do a search there.

    Sincerely,
    Cac

  5. #5
    Alexost
    Guest

    Alexost: Hmmm.

    I guess I'm not really explaining myself very well. I have read all of the posts on TC compression and I have a good understanding of TC conversion. What I'm asking I dont think has anything to do with TC compression.

    I know yourself and Don are more than capable of understanding what I'm getting at, so it must be my explanations.

    "Indices should be calculated in the same way you're going to use them"

    If the above statement is true, then why do indices for standard Hi-Lo (not Hi-Lo lite) that are generated by full decks have better scores when more precision is used when converting to TC? If the above statement were correct, we would be advising people who use Hi-Lo to never use more than full deck precision when converting to TC.

    Examples:

    Hi-Opt II- full deck generated indices. The more precision when converting to TC, the higher the score.

    Hi-Lo- full deck generated indices. The more precision when converrting to TC, the higher the score.

    RPC- half deck generated indices. The more precision when converting to TC, the *lower* the score.

    That is my dilemma.


  6. #6
    Don Schlesinger
    Guest

    Don Schlesinger: I know the problem

    > I guess I'm not really explaining myself very well. I
    > have read all of the posts on TC compression and I
    > have a good understanding of TC conversion. What I'm
    > asking I dont think has anything to do with TC
    > compression.

    Maybe not. But I know what it does have to do with, and we've been down this road a million times.

    > I know you and Don are more than capable of
    > understanding what I'm getting at, so it must be my
    > explanations.

    It's terminology.

    > "Indices should be calculated in the same way
    > you're going to use them"

    Yes, they should.

    > If the above statement is true,

    It is.

    > then why do indices
    > for standard Hi-Lo (not Hi-Lo lite) that are generated
    > by full decks have better scores when more precision
    > is used when converting to TC?

    "Full decks" can mean dividing by one (full) deck, 1.5 (full decks), two (full) decks, etc. What it does not mean, in the above cases, is dividing by two half decks, three half decks, or four half decks. So, the indices can be reckoned by whole decks, and then, when you play, your whole-deck precision can be by full decks or half decks. Half-deck precision (of WHOLE decks!) will be better.

    > If the above statement
    > were correct, we would be advising people who use
    > Hi-Lo to never use more than full deck precision when
    > converting to TC.

    Not true. See above.

    > Examples:

    > Hi-Opt II- full deck generated indices.

    What, precisely, does this mean?

    > The more
    > precision when converting to TC, the higher the score.

    But, you don't mention the following: You generate by full decks. You use increasingly more precision for TC reckoning, and the SCORE (capitals, please!) improves. Fine. Now, suppose you match the generation process to the playing process. You haven't checked to see that the SCOREs are even better now, right?

    > Hi-Lo- full deck generated indices. The more precision
    > when converrting to TC, the higher the score.

    I would imagine.

    > RPC- half deck generated indices. The more precision
    > when converting to TC, the *lower* the score.

    Because this is DIFFERENT. You're now dividing by quarter-decks instead of half-decks. You need to stay with half-decks! Now, you might do something weird like, say, 3.5 half decks remain, but that is NOT the same thing as switching to quarter decks. And if you do what I said, you will have greater accuracy than using integral numbers of half-decks only -- but it will drive you crazy in the process.

    One day, I'm going to write an article on this. It's a terribly misunderstood concept. I proposed to Cacarulo and Norm to write just such an article a few months ago, but they talked me out of it. I'm going to do it anyway, one day!

    Don

  7. #7
    Alexost
    Guest

    Alexost: Re: I know the problem

    Sorry about the SCORE not capitalized.

    I do understand the article that you are going to write!
    I guess the terminology I'm using is conveying a different message.

    > Hi-Opt II- full deck generated indices.
    What, precisely, does this mean?

    This means generating strategy departure indices using full decks instead of half decks. If we were to generate Hi Opt II indices with half decks, the integers would be compressed.

    For example, I know that when generating strategy departure indices on a level 1 system using 1/2 decks, the indices end up "compressed" into a narrower range.

    "Because this is DIFFERENT. You're now dividing by quarter-decks instead of half-decks. You need to stay with half-decks!"

    Even though on pg. 137 of PBJAAB Reveres' chart shows to multiply the RC by 2/3 when 3/4 deck is remaining, and then multiply by 2 when 1/4 deck is remaining?

    Isn?t Revere also making the mistake of using quarter deck precision on a system where he generated indices using 1/2 decks?

    So when I?m playing 1 deck games, I should divide by 2 until I get to one half deck remaining, and then divide by 1 when one half deck remains? And skip the 3/4 deck and 1/4 deck remaining calculations?

    (I know, it is rare to find 1 deck games where I'll be making decisions at 1/4 deck remaining but I do find them in Reno)

    Thanks for you time Don and Cac.


  8. #8
    Don Schlesinger
    Guest

    Don Schlesinger: Re: I know the problem

    > Sorry about the SCORE not capitalized.

    You're forgiven.

    > This means generating strategy departure indices using
    > full decks instead of half decks.

    Yet again, I desperately ask: When you used the FULL decks, did you allow for half-deck PRECISION, such as 1.5 FULL decks, 2.5 FULL decks, etc.?

    > If we were to
    > generate Hi Opt II indices with half decks, the
    > integers would be compressed.

    Hi-Opt II is a level-2 count similar to the RPC. It is perfectly acceptable to use half-deck generation of indices and TC reckoning when you play.

    > For example, I know that when generating strategy
    > departure indices on a level 1 system using 1/2 decks,
    > the indices end up "compressed" into a
    > narrower range.

    Right. And that's bad. It's not bad for level-2 counts, as this is the equivalent of using full decks for level-1 counts.

    > "Because this is DIFFERENT. You're now dividing
    > by quarter-decks instead of half-decks. You need to
    > stay with half-decks!"

    > Even though on pg. 137 of PBJAAB Revere's chart shows
    > to multiply the RC by 2/3 when 3/4 deck is remaining,
    > and then multiply by 2 when 1/4 deck is remaining?

    The above are HALF-DECK conversions!! 3/4 of a deck remaining are 3/2 HALF decks. So you divide by 3/2, which means multiply by 2/3. You're making the same mistake again. If you divide by WHOLE decks, you're allowed to divide by 2.5 such WHOLE decks. If you divide by HALF decks, you're allowed to divide by, say, 1.5 such HALF decks. You desperately need to understand this.

    > Isn?t Revere also making the mistake of using quarter
    > deck precision on a system where he generated indices
    > using 1/2 decks?

    NO! See above. Quarter-deck precision is different from GENERATING the indices by using quarter-decks! He used half-decks to generate them 9with 1/4-deck precision), and he plays the game by using half-decks to reckon the TC, but with quarter-deck precision.

    > So when I?m playing 1 deck games, I should divide by 2
    > until I get to one half deck remaining, and then
    > divide by 1 when one half deck remains? And skip the
    > 3/4 deck and 1/4 deck remaining calculations?

    No. Do them as outlined above. (I can't write that article soon enough! :-))

    Don

  9. #9
    Alexost
    Guest

    Alexost: Re: I know the problem

    > You're forgiven.

    > Yet again, I desperately ask: When you used the FULL
    > decks, did you allow for half-deck PRECISION, such as
    > 1.5 FULL decks, 2.5 FULL decks, etc.?

    Sorry, I'm not sure how to 'allow' for different levels of precision when generating indices.. (I've been using PBA and just set it to generate the indices reckoned by full or half decks)

    > The above are HALF-DECK conversions!! 3/4 of a deck
    > remaining are 3/2 HALF decks. So you divide by 3/2,
    > which means multiply by 2/3. You're making the same
    > mistake again. If you divide by WHOLE decks, you're
    > allowed to divide by 2.5 such WHOLE decks. If you
    > divide by HALF decks, you're allowed to divide by,
    > say, 1.5 such HALF decks. You desperately need to
    > understand this.

    I do understand this, I guess what it comes down to is my NOT understanding how to run CVCX properly. I think all along I have been confusing myself because of the figures I got out of CVCX, because I set up CVCX wrong. Maybe Norm can read this thread and make heads or tails of my ineptitude?

    > NO! See above. Quarter-deck precision is different
    > from GENERATING the indices by using quarter-decks! He
    > used half-decks to generate them 9with 1/4-deck
    > precision), and he plays the game by using half-decks
    > to reckon the TC, but with quarter-deck precision.

    I understand the difference between generating indices and applying more precision to true count calculation.

    But I still don't understand how anyone 'allows' for more precision, or 'doesn't allow' for more precision when they are generating the indices. So I guess that there is something I can learn from that article you are going to write!

    > No. Do them as outlined above. (I can't write that
    > article soon enough! :-))

    Good, because I want to read more of your literature, please keep writing more... And while you're at it- Please write a pamphlet on ace prediction available to people who can be confirmed as non-casino affiliated; and they must sign a contract using their soul as collateral that they won't sell it to casinos. I want to see ace prediction from your perspective.

    And whats more, please write a semi-non fictional book like "The Counter" about your adventures as a cardcounter, team member, team leader etc.

    Thanks again Don!


  10. #10
    Alexost
    Guest

    Alexost: Re: RPC TC precision with CVCX question

    I found the Halves vs Zen dispute, very interesting and it helps show me how index generation and TC calculation terminology is used by the experts. As I mentioned below in response to Don, I think my confusion has to do with how I set up CVCX. I set CVCX strategy tables with RPC half deck reckoned indices (as published in PBJAAB) and set TC calculation to 1/4 deck resolution thinking that this is how to get 1/4 deck TC precision. I guess the proper way of setting CVCX (with RPC 1/2 deck indices) is 1/2 deck resolution which gives me 1/4 deck precision (resolution) because half of a half deck IS a quarter deck. (We can use resolution to mean the same thing as precision cant we?)

    1) When generating system indices we can reckon the them to full, 1/2 or even 1/4 if we want to.

    (I also read that an old version of Hi-Lo from Thorpe reckoned the indices by 2 decks)

    2) Once we have generated the indices with any of the above reckoning methods we can then apply the level of precision (resolution) we wish: full deck, half deck, quarter deck, exact deck.

    3) But when I use CVCX and set the resolution to 1/2 deck using half deck RECKONED indices I'm actually seeing 1/4 deck precision/resolution because half of a half deck is a quarter deck.

    Is that right Norm?

    Thanks to all-


  11. #11
    Norm Wattenberger
    Guest

    Norm Wattenberger: Re: I know the problem

    CVData allows you to set different levels of precision in different parts of the shoe when generating indexes.

  12. #12
    Norm Wattenberger
    Guest

    Norm Wattenberger: Re: RPC TC precision with CVCX question

    > 3) But when I use CVCX and set the resolution to 1/2
    > deck using half deck RECKONED indices I'm actually
    > seeing 1/4 deck precision/resolution because half of a
    > half deck is a quarter deck.

    No, they are not related.

  13. #13
    Alexost
    Guest

    Alexost: Figured my mistake- my fault.

    Sorry for the confusion to everybody who helped me out. In the midst of doing all these sims I neglected to double check whether I kept the flooring, rounding, or truncating the same for all the sims that I did.

    So, it turns out that when I elected to sim at 1/4 deck precision, I had a different TC method selected (floor, round, truncate)than when I did 1/2 deck precision sims- This is what made the SCORE lower even though it should have been better at the 1/4 deck precision..

    I knew I had indice generation and precision fully understood, and I was even using the software correctly. It was my lack of attention to detail during that sim that sent me on this goosechase.

    Thanks to all for their time-

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •  

About Blackjack: The Forum

BJTF is an advantage player site based on the principles of comity. That is, civil and considerate behavior for the mutual benefit of all involved. The goal of advantage play is the legal extraction of funds from gaming establishments by gaining a mathematic advantage and developing the skills required to use that advantage. To maximize our success, it is important to understand that we are all on the same side. Personal conflicts simply get in the way of our goals.