See the top rated post in this thread. Click here

Page 2 of 4 FirstFirst 1234 LastLast
Results 14 to 26 of 49

Thread: What is the Tarzan count that beats HO2 w ASC

  1. #14


    Did you find this post helpful? Yes | No
    The executive summary answer is that the canned sims in CVCX used only the I18 indices, where my sims used the full set of HiOpt II Indices.

    The Tarzan sims were the first time I published SCOREs which were generated using my software. In order to demonstrate that my results were not in left field, I also published HiOpt II+ASC SCOREs generated by CV for the same games alongside of my results. I did this by performing new sims myself using CVData+CVCX. The reason for that is that, as Norm just said in another thread, there are many many nuances and interactions which may not be obvious when configuring a sim and also when writing the software to perform the sim.

    CVCX alone only gives you partial control over the sim parameters. By using CVData+CVCX, I was able to match the configurations of both CV and my own software as closely as possible. I also configured all of the sims to conform to the sim parameters of BJA3 Chapter 10, except for things like choice of indices used, since Don and I felt that it would not be fair to compare HiOpt II + ASC to Tarzan without full indices.

    I'm looking at my records and, for the 5/6 S17 DAS game, 1-12 spread from the Tarzan sims. I see my SCORE of 44.19 and the CV SCORE of 44.93 for HiOpt II + ASC. The canned sim SCORE is lower because it used only the I18 indices.

  2. #15


    Did you find this post helpful? Yes | No
    Quote Originally Posted by Gronbog View Post
    The executive summary answer is that the canned sims in CVCX used only the I18 indices, where my sims used the full set of HiOpt II Indices.

    The Tarzan sims were the first time I published SCOREs which were generated using my software. In order to demonstrate that my results were not in left field, I also published HiOpt II+ASC SCOREs generated by CV for the same games alongside of my results. I did this by performing new sims myself using CVData+CVCX. The reason for that is that, as Norm just said in another thread, there are many many nuances and interactions which may not be obvious when configuring a sim and also when writing the software to perform the sim.

    CVCX alone only gives you partial control over the sim parameters. By using CVData+CVCX, I was able to match the configurations of both CV and my own software as closely as possible. I also configured all of the sims to conform to the sim parameters of BJA3 Chapter 10, except for things like choice of indices used, since Don and I felt that it would not be fair to compare HiOpt II + ASC to Tarzan without full indices.

    I'm looking at my records and, for the 5/6 S17 DAS game, 1-12 spread from the Tarzan sims. I see my SCORE of 44.19 and the CV SCORE of 44.93 for HiOpt II + ASC. The canned sim SCORE is lower because it used only the I18 indices.
    I did another simulation for HiOpt II + ASC with full indices and the results came out very close to the canned sim for HiOpt II + ASC. This was done through CVCX.

    Hi-OPT II Full Indice & ASC S17,DAS.jpg

    Canned Simulation for Hi-OPT II Full Indices ASC
    Hi-OPT II Full Indice & ASC.jpg

    The SCORE came out to be really close.
    Last edited by seriousplayer; 01-20-2020 at 09:41 AM.

  3. #16


    Did you find this post helpful? Yes | No
    Quote Originally Posted by Gronbog View Post
    The executive summary answer is that the canned sims in CVCX used only the I18 indices, where my sims used the full set of HiOpt II Indices.
    No, the canned sims in my version give only full indices and don't even offer I18 for Hi-Opt II ASC. So the reason has to lie elsewhere for the discrepancy. And the SCORE above, as seriousplayer mentions, is for full indices.

    Don

  4. #17


    Did you find this post helpful? Yes | No
    seriousplayer, I see a couple of things in your screen shots.

    In order to get a true SCORE, your bankroll must be 10,000 and you must be betting full Kelly optimal bets to the nearest 1. This will cause your risk of ruin to be ~13.53%. In addition to this, although it probably doesn't make much difference, I used the BJA3 chapter 10 convention of 3 basic strategy players to the right of the key player. To make sure this is exactly what was happening, I used the Multi-tracking sim in CVData to run the sim and then used CVCX to display the results.

    I found the archive of the sim that ran at the time. Here is a screen shot of the CVCX display:

    HiOptII+ASC-6D-Pen5-S17-DAS-PA.opt.1-12.jpg

  5. #18


    Did you find this post helpful? Yes | No

    From what I see the Tarzan count requires four levels of complexiity

    Tarzan's count system uses three columns of cards and an Ace side count. Thus there are four levels of complexity with the Tarzan count. With KO w 5m7c and AA89mTc there are only two side counts and the primary count so three levels of complexity.

    Forum readers were complaining about how complicated KO w 5m7c and AA89mTc is to use.

    So my question is why did forum readers give me such a hard time on the KO w 5m7c and AA89mTc being so difficult when they had absoutely no complaints about the complexity of Tarzan's count?

    Finally I showed in a previous post Gronbog's sims of KO w 5m7c and AA89mTc compared to Tarzan's best count and that for the back counted game, KO w 5m7c and AA89mTc beats the best Tarzan count.

    If a third side count were added to KO w 5m7c and AA89mTc, to make it four levels of complexity like Tarzan's count, my system would absoutely beat Tarzan's best system and if 45m79c were used instead of 5m7c the improvement would be even greater.

    Last edited by bjanalyst; 01-20-2020 at 03:32 PM.

  6. #19


    Did you find this post helpful? Yes | No
    Quote Originally Posted by Gronbog View Post
    seriousplayer, I see a couple of things in your screen shots.

    In order to get a true SCORE, your bankroll must be 10,000 and you must be betting full Kelly optimal bets to the nearest 1. This will cause your risk of ruin to be ~13.53%.
    Gronbog is right. And to truly compare apple to apple in the meaning way, we need to stick to the rules of $10,000 bankroll and bet full Kelly optimal bets. If you change the baselines, the end result can only be called c-Score, custom-Score, quasi-Score etc.

  7. #20


    1 out of 1 members found this post helpful. Did you find this post helpful? Yes | No
    Quote Originally Posted by bjanalyst View Post
    Tarzan's count system uses three columns of cards and an Ace side count. Thus there are four levels of complexity with the Tarzan count. With KO w 5m7c and AA89mTc there are only two side counts and the primary count so three levels of complexity.

    Forum readers were complaining about how complicated KO w 5m7c and AA89mTc is to use.

    So my question is why did forum readers give me such a hard time on the KO w 5m7c and AA89mTc being so difficult when they had absoutely no complaints about the complexity of Tarzan's count?

    Finally I showed in a previous post Gronbog's sims of KO w 5m7c and AA89mTc compared to Tarzan's best count and that for the back counted game, KO w 5m7c and AA89mTc beats the best Tarzan count.

    If a third side count were added to KO w 5m7c and AA89mTc, to make it four levels of complexity like Tarzan's count, my system would absoutely beat Tarzan's best system and if 45m79c were used instead of 5m7c the improvement would be even greater.

    1. Because you don't know your Blackjack and trying to be an expert at it when you are not.

    2. It doesn't take three or four secondary side count to beat Hi-OPT II w ASC, one secondary count is enough. Adding more secondary side count makes your count system harder.

  8. #21


    Did you find this post helpful? Yes | No
    Quote Originally Posted by BJGenius007 View Post
    Gronbog is right. And to truly compare apple to apple in the meaning way, we need to stick to the rules of $10,000 bankroll and bet full Kelly optimal bets. If you change the baselines, the end result can only be called c-Score, custom-Score, quasi-Score etc.
    No, sorry, guys that isn't the case. The SCORE value at the bottom is independent of any settings you use for bankroll. Ditto for Kelly fraction. The values for not respecting the strict SCORE criteria show up bottom left, but don't affect the SCORE value, which is derived from the traditional inputs.

    Don

  9. #22


    Did you find this post helpful? Yes | No
    Quote Originally Posted by seriousplayer View Post
    1. Because you don't know your Blackjack and trying to be an expert at it when you are not.

    2. It doesn't take three or four secondary side count to beat Hi-OPT II w ASC, one secondary count is enough. Adding more secondary side count makes your count system harder.
    You did not answer my questions. My question was not about beating HO2 w ASC.

    My question was

    Tarzan's count system uses three columns of cards and an Ace side count. Thus there are four levels of complexity with the Tarzan count. With KO w 5m7c and AA89mTc there are only two side counts and the primary count so three levels of complexity.

    So why did forum readers complain about how complex KO w 5m7c and AA89mTc was with three levels of complexity but everyone is fine and no complaints about Tarzan's count with four levels of complexity, especially when KO w 5m7c and AA89mTc beats Tarzans best for the back counted game as shown by Gronbog's sims.

    Please answer the questions that I asked and do not get side tracked on other topics I did not ask.

    Last edited by bjanalyst; 01-20-2020 at 05:45 PM.

  10. #23


    Did you find this post helpful? Yes | No
    Quote Originally Posted by seriousplayer View Post
    It doesn't take three or four secondary side count to beat Hi-OPT II w ASC, one secondary count is enough. Adding more secondary side count makes your count system harder.
    In this post I am going to show that your statement above is incorrect.

    Here are the counts you mentioned that beat HO2 w ASC and none of them is a level one primary count with one side count that beats HO2 w ASC so none of them are simpler than KO w AA89mTc and 5m7c.

    They all have a level of complexity in them more difficult than KO w 5m7c and AA89mTc.

    They are all either level 2 or level 3 primary counts with a side count of Aces (I do not like single card side counts like a side count of Aces - the Aces played is always increasing and you have to estimate decks played so the ASC is APPROXIMATE and much more difficult to keep track of than plus/minus side counts which are EXACT and fluctuate about their mean of zero) or a primary count with multiple side counts (like Gordon that you mentioned below) to beat HO2 w ASC.

    Tarzan Count
    4 levels of complexity

    Gordon Count with side counts
    Level one primary count but have to keep multiple single card side counts which is exactly opposite of what you said when you said you needed only one side count.

    BRH II combined with a secondary count to get BRH I for betting.
    Level 2 primary count with side count

    USTON APC w ASC.
    Level 3 primary count with side count of Aces

    VICTOR APC w ASC

    Level 3 primary count with side count of Aces

    So every one of your suggested counts that beats the HO2 w ASC are either a level 2 or higher primacy count with a side count or the level one Gordon count with multiple side counts and they are all more complex than KO w 5m7c and AA89mTc and even your very best count, the Tarzan count, does not beat KO w 5m7c and AA89mTc for the backcounted scenarios.

    If 45m79c were used instead of 5m7c with KO and AA89mTc giving you KO w 45m79c and AA89mTc then the advantage over the very best count you have, the extremely complex Tarzan count which you are totally OK with and have no complaints about its complexity, would be even greater with three levels of complexity as compared to Tarzan's four levels of complexity.

    A level one primary count, like the KO, and plus/minus side counts, are the easiest to keep. I do not like level 2 or 3 primary counts and I do not like side counts of individual cards like an Ace side count for example.
    Last edited by bjanalyst; 01-20-2020 at 05:44 PM.

  11. #24


    Did you find this post helpful? Yes | No
    Quote Originally Posted by DSchles View Post
    No, sorry, guys that isn't the case. The SCORE value at the bottom is independent of any settings you use for bankroll. Ditto for Kelly fraction. The values for not respecting the strict SCORE criteria show up bottom left, but don't affect the SCORE value, which is derived from the traditional inputs.
    Very interesting. I did not realize this, but playing around with it now, it's definitely true.

    So we're left with my SCORE:44.19 (4 players), CVData+CVCX SCORE=44.93 (4 players) and the CVCX canned sim SCORE=41.07 (heads up), 42.72 (4 players).

    The CVCX canned SCOREs show that 4 players vs 1 makes a difference, perhaps due to slightly higher effective penetration, but still not in the 44+ range generated by the CVData multi-tracking simulator and my own software.

    Here is a screenshot of the canned CVCX sim for 4 players:
    HiOptII+ASC-6D-Pen5-S17-DAS-PA.opt.1-12.4players.jpgCompare that to my CVData+CVCX sim for 4 players:
    HiOptII+ASC-6D-Pen5-S17-DAS-PA.opt.1-12.4players.CVData.jpg
    I am in no position to comment on why, but the are clear differences in the base data collected by the two simulators and, in particular, the true count frequencies and EVs. The difference in the range of the negative true counts can be ignored since they are all assigned the minimum bet. The differences are enough that a higher minimum bet was calculated for the CVData sim and are enough to lead to a higher SCORE.

    For what it's worth, here are the sim data as generated by my simulator:
    Code:
    Generated by gamblor master commit fb27f246
    Sun Jan 13 23:02:55 2019 GMT
    Copyright Gronbog, All Rights Reserved
    
    Rules: 6 Decks, 52 Cards Cut Off, S17, BJ 3/2, DOA, DAS, SPL3, SA1, PEEK
    Strategy: withIndices
    Counting System: HiOpt-II+ASC
      True Count Conversion Options
        Deck Size: 52
        Deck Estimation: round to nearest 0.50 decks based on cards remaining
        True Count Conversion: floor to nearest 1.00
      Index Strategy: hiOptII.s17
    Count Unit: 1
    Count Precision: 0 decimal places
    
    Bankroll        $10000.00
    Kelly Fraction     100.00%
    
    Ramps for Spot 4
    Spread      1 To        12
    Min Bet   $14 Max Bet $168
    
    Betting                              Standard
    Count     Frequency        EV        Deviation     Variance     Optimal Bet        
    <=-1       45.77%        -1.26%        1.143         1.306          $14                
      0        17.66%        -0.27%        1.154         1.331          $14                
      1        8.76%          0.08%        1.156         1.336          $14                
      2        7.48%          0.41%        1.158         1.341          $30                
      3        4.56%          0.72%        1.160         1.345          $54                
      4        4.05%          1.06%        1.162         1.350          $78                
      5        2.47%          1.37%        1.164         1.355         $101                
      6        2.32%          1.72%        1.168         1.365         $126                
      7        1.40%          2.04%        1.172         1.375         $148                
      8        1.37%          2.39%        1.179         1.389         $168                
      9        0.81%          2.70%        1.184         1.401         $168                
      10       0.82%          3.09%        1.190         1.416         $168                
      11       0.48%          3.40%        1.195         1.427         $168                
      12       0.50%          3.81%        1.201         1.441         $168                
      13       0.29%          4.09%        1.204         1.448         $168                
      14       0.30%          4.50%        1.209         1.462         $168                
      15       0.18%          4.72%        1.211         1.465         $168                
      16       0.19%          5.22%        1.216         1.479         $168                
      17       0.11%          5.47%        1.217         1.481         $168                
      18       0.11%          5.91%        1.222         1.494         $168                
      19       0.07%          6.13%        1.223         1.495         $168                
    >=20       0.29%          7.78%        1.235         1.526         $168                
             100.00%
    
    Percent Rounds Played               100.00%
    Average Bet (per round played)      $34.75
    Average Bet (per round observed)    $34.75
    Win/Loss% (EV)                        1.25%     $43.51 per 100 rounds observed   $43.51 per 100 rounds played
    Variance                             3.547
    Standard Deviation                   1.883     $654.50 per 100 rounds observed  $654.50 per 100 rounds played
    N0                                  22,630
    Risk of Ruin                         13.12%
    DI                                    6.65
    c-SCORE                              44.19
    Bankroll needed for 1% risk of ruin $22,670.62
    The base statistics (Frequency, EV) more closely resemble those of the CVData results, which is why my SCORE is also closer to that SCORE.
    Attached Images Attached Images

  12. #25


    Did you find this post helpful? Yes | No
    Quote Originally Posted by bjanalyst View Post
    In this post I am going to show that your statement above is incorrect.

    Here are the counts you mentioned that beat HO2 w ASC and none of them is a level one primary count with one side count that beats HO2 w ASC so none of them are simpler than KO w AA89mTc and 5m7c.

    They all have a level of complexity in them more difficult than KO w 5m7c and AA89mTc.

    They are all either level 2 or level 3 primary counts with a side count of Aces (I do not like single card side counts like a side count of Aces - the Aces played is always increasing and you have to estimate decks played so the ASC is APPROXIMATE and much more difficult to keep track of than plus/minus side counts which are EXACT and fluctuate about their mean of zero) or a primary count with multiple side counts (like Gordon that you mentioned below) to beat HO2 w ASC.

    Tarzan Count
    4 levels of complexity

    Gordon Count with side counts
    Level one primary count but have to keep multiple single card side counts which is exactly opposite of what you said when you said you needed only one side count.

    BRH II combined with a secondary count to get BRH I for betting.
    Level 2 primary count with side count

    USTON APC w ASC.
    Level 3 primary count with side count of Aces

    VICTOR APC w ASC

    Level 3 primary count with side count of Aces

    So every one of your suggested counts that beats the HO2 w ASC are either a level 2 or higher primacy count with a side count or the level one Gordon count with multiple side counts and they are all more complex than KO w 5m7c and AA89mTc and even your very best count, the Tarzan count, does not beat KO w 5m7c and AA89mTc for the backcounted scenarios.


    I already have an idea of a level 1 count that could perform on the level of HO2 w ASC that doesn't require 3 or 4 side counts. Why should I share it with you so you can copy it. I am not going to give you the details for it. Yes, most of the count that was listed was level 2 or higher counts because those are the counts used at the professional level.

    If you think that saying any count on the list that beat Hi-OPT II w ASC is misleading you are mistaking.

    Quote Originally Posted by bjanalyst View Post
    If 45m79c were used instead of 5m7c with KO and AA89mTc giving you KO w 45m79c and AA89mTc then the advantage over the very best count you have, the extremely complex Tarzan count which you are totally OK with and have no complaints about its complexity, would be even greater with three levels of complexity as compared to Tarzan's four levels of complexity.

    A level one primary count, like the KO, and plus/minus side counts, are the easiest to keep. I do not like level 2 or 3 primary counts and I do not like side counts of individual cards like an Ace side count for example.

    Than in the same sense we should start adding more components to Hi-OPT II w ASC. What are you going to do if more components are added to Hi-OPT w ASC and it beats your crappy KO and AA89mTc??? Then you don't know what to do but to add more components to your KO and AA89mTc.
    Last edited by seriousplayer; 01-21-2020 at 09:32 PM.

  13. #26


    Did you find this post helpful? Yes | No
    Quote Originally Posted by bjanalyst View Post
    So why did forum readers complain about how complex KO w 5m7c and AA89mTc was with three levels of complexity but everyone is fine and no complaints about Tarzan's count with four levels of complexity,
    Bjanalyst, since you've asked the question twice and not received an answer, I'll try. The reason is that Tarzan hasn't tried to get others to use his system. Also, I doubt that very many people on this forum knew the intricacies of Tarzan's system prior to Gronbog's work.
    Last edited by 21frogman; 01-21-2020 at 02:07 PM.

Page 2 of 4 FirstFirst 1234 LastLast

Similar Threads

  1. Roulette/Tarzan
    By Bricklayer in forum General Blackjack Forum
    Replies: 51
    Last Post: 08-01-2017, 09:34 AM
  2. Any news on the Tarzan count?
    By Goodboy in forum General Blackjack Forum
    Replies: 1
    Last Post: 12-12-2016, 06:58 AM
  3. KJ, T3, and Tarzan
    By Exoter175 in forum General Blackjack Forum
    Replies: 22
    Last Post: 08-09-2014, 02:05 PM

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.