Let me try to understand this. If the card counting system is using true count for betting than increase the spread and reduce the divisor by the fraction.Doing that the higher BC would beat the count with lower BC given that they are both the same level. But Uston SS and BRH I could both be use with running count and true count. What about in running count? How would you be able to see the difference there? Just increase the spread?Originally Posted by Gramazeka
I attached a PDF with CC of each of the important playing strategies for both HL w 7m9c and HL w 7SC which you obviously did not look at.
Yes IC of HL w 7SC is 78.4% of a 2.4% increase of HL IC of 76.0%
But BE for DAS, S17, LS of HL w 7m9s is 97.6% or 0.6% less than BE of HL w 7m9c which is 98.2%. HL has BE of 96.5% for DAS, LS, S17.
I will attach the one page PDF again comparing on an case by case basic CC of HL w 7SC and HL w 7m9c.
When all CC are considered for all situations (some HL w 7m9c beats HL w 7SC and some HL w 7SC beats HL w 7m9c) the count system are approximately equal.
I personally like plus/minus side counts, not individual card side counts. plus/minus side counts are EXACT as they do not involved any estimate of decks played and the 7m9c fluctuations around its mean of zero and is not continually increase like keeping track of sevens played and comparison to decks played. So MY preference is plus/minus side counts and my analysis was HL w 7m9c. I did not analyze HL w 7SC except to do a side by side comparison of CC of each which I gave you before and which I will attach again since it is a one page PDF.
If you want help with insurance, LL, Super 4, or hit/stand on hard 16 v 7, 8, 9, T then you need to keep Am6c or ASC with HL. Am6c helps more with hit/stand hard 16 v 7, 8, 9, T than ASC does. ASC helps more with insurance and Super 4 and LL. If Super 4 or LL is offered, keep ASC. If no side bets both Am6 and ASC perform approximately the same. In the case of no side counts, I would there fore recommend Am6c because of my preference of simple plus/minus side counts over counts of individual ranks.
HL w 7m9c vs HL w 7def.pdf
I do like EBJ2/2 which is HL + (1/2)*(7m9c).
So EBJ2 / 2 is HL w 7m9c = HL + k*(7m9c) with k = (1/2) for betting all playing strategy decisions.
By breaking EBJ2 /2 into its component parts, HL and 7m9c, and then using HL + k*(7m9c) you have more flexiibliy in using the count that is best for each situation and so as I showed before HL w 7m9c is slightly better than EBJ2 / 2.
Remember, my goal was to devise a simple side count for the HL player would never swtich from the HL count to any other count and especialy not a level 2 count such as EBJ2 /2. Thus my suggestion was to add 7m9c as a side count to the HL.
So I never considered the primary count as anything other htan the level one HL.
But you keep EBJ2 / 2 which is fine and you seem you can do it fine. This is a personal preference at this point. Does the player want to switch from the HL to a level 2 count or would the player prefer to keep the HL level one count and simply add a level one side count.
From all the counters I meet, they refused to switch from the level one HL so they would choose a 7m9c side count to add to the HL before gonig to a level 2 count and HL w 7m9c is slightly more powerful than EBJ2 /2.
But again, if you prefer keeping the single level 2 EBJ2 / 2 instead of the HL w 7m9c then that is fine. It is just slightly less powerful than HL w 7m9c.
I will attach the side by side CC for HL w 7m9c top 6 against EBJ2 /2 that I gave you before for your review since it is one page PDF.
I used a straight average of the CC because I was lazy in slapping this together. You should really use a weighed average where more important strategy changes, such as insurance are given more weight. But you can still get a good idea from the straight average CC comparisons of comparative strengths of each system.
EBJ2 CC.pdf
Last edited by bjanalyst; 01-18-2020 at 10:46 PM.
The IC for Hi-lo with side count of sevens is higher than 78.4. I did look at you junky PDF file. It is theory after theory after theory. Where in the your PDF did you show at what count to take insurance, K Select? It is in decimals you think a normal human being count in decimals?
The correct terms is BC Betting Correlation NOT BE. There is a difference as it is define in "The Theory of Blackjack".
No it is not! We are talking about applying this in the casino, not theory. In practice you didn't include insurance and in the PDF you compared insurance. So you are comparing something you didn't apply.
Side counting individual card could be EXACT, also. It doesn't have to involve any deck estimate either.
See a different way to say you are adding more components to your count. I heard this six different ways.
More theory!!! Again, how do you implement your HL w 7m9c and Am6c?
You would NOT get the same main count counting HL w ASC for HL w 7m9c. There is no way in the world you would come up with the same count counting HL w ASC for HL w 7m9c. They are two different counts even though, in your case their CC is the same, which comparing CC is unless.
Last edited by seriousplayer; 01-18-2020 at 11:00 PM.
I will attach a PDF that shows the details of insurance calculation of HL w 7SC. You will see k = 0.8 in HL + k*(7SC) which gives IC = 78.4% and infinite deck index = 3.326.
I did not show the indices for HL w 7SC because I was not analyzing HL w 7SC. I analyzed HL w 7m9c.
When comparing different systems, you just need to compare CC for each strategy change and CC for betting which is what I showed.
For betting there is only one strategy change and so Betting Correlation is equal to Betting Efficiency.
For playing strategy changes, there are many different CC for various playing situations. Griffin said CC are directly related to efficiency. So for playing strategy I just calculated an average CC for each system I am comparisons. And you can assign weights to each situation but use the same weights for all systems you are comparing.
Insurance w 7SC.pdf
Ok, you didn't show the insurance index for HL w 7m9c.
Citation needed. Where is it written that you compare CC for betting?
Betting Correlation and Betting Efficiency are not the same.
Not for Hi-Lo with Side Count Seven. There is only one index deviation. I know because simulation and studies have already been done for Hi-Lo with Side Count Seven.
Ok, you didn't show the insurance index for HL w 7m9c.
For all plays other than the top 6, HL w 7m9c is the HL. So the insurance index for HL w 7m9c is the HL insurance index.
Citation needed. Where is it written that you compare CC for betting?
Betting Correlation and Betting Efficiency are not the same.
See attached pages from TOB. Note that Griffin says "efficiency is directly related to and sometimes equal to correlation coefficient between the point value of the card counting system and the single card payoffs approximating the blackjack situation considered". When there is only one situation being considered separately, such as betting or insurance, the betting efficiency or insurance efficiency is the betting CC or insurance CC. If you are trying to make a grouping of various playing strategies and get a ingle efficiency for those playing start then there is no single CC to apply to these and efficiency is calculated by some other means that I am not sure of. I just used an average or a weighed average CC and according to Griffin efficiency is directly related to these CC.
Also every single simulation that Gronbog did with HL w AA78mTc and then HL w AA78mTc & 5m6c and KO with AA89mTc and 5m7c, and now HL w 7m9c when I added additional indices and situations based on CC, the SCORE increased every single time. And this included even some very marginal changes with soft double indices added to HL w AA78mTc where the SCORE had a very small increase, but it still increased!
So CC comparisons are a legitimate technique to measure the strength of various counts. And when all CC for all situations of HL w 7m9c and HL w 7SC are considered, they are overall approximately equal. I do not like single card side counts, I like plus/minus side counts since they are EXACT and easier, at least for me, to keep. Thus I analyzed HL w 7m9c.
Not for Hi-Lo with Side Count Seven. There is only one index deviation. I know because simulation and studies have already been done for Hi-Lo with Side Count Seven.
I did not list indices, just CC, which is all you need to compare strengths of various counts - compare CC for betting and for playing strategies.
Griffin on CC.pdf
Last edited by bjanalyst; 01-19-2020 at 05:59 AM.
Please list the page number that you got that information from. When Griffin say "efficiency is directly related to and sometimes equal to correlation coefficient between the point value of the card counting system and the single card payoffs approximating the blackjack situation considered" he doesn't only mean "Betting Correlation". Betting Correlation as he said in the next section is just one example of efficiency. I think Griffin is defining "efficiency" for a card counting system to mean the measure of all three parameters BC, PE and IC determines the efficiency of the system. He didn't say explicitly that "Betting Efficiency" is the same as "Betting Correlation".
Also, we don't care about HL w AA78mTc & 5m6c and KO with AA89mTc and 5m7c I don't know why you keep bring that up. This thread is about Hi-lo + 7m9c. CC is might be a legitimate way to determine count strength but it is not the method used to calculate and determine SCORE.
If you didn't list any indices how does anyone know how to apply and implement your system?
Last edited by seriousplayer; 01-19-2020 at 03:36 PM.
Please list the page number that you got that information from. When Griffin say "efficiency is directly related to and sometimes equal to correlation coefficient between the point value of the card counting system and the single card payoffs approximating the blackjack situation considered" he doesn't only mean "Betting Correlation". Betting Correlation as he said in the next section is just one example of efficiency. I think Griffin is defining "efficiency" for a card counting system to mean the measure of all three parameters BC, PE and IC determines the efficiency of the system. He didn't say explicitly that "Betting Efficiency" is the same as "Betting Correlation".
If you look closely at the PDF I sent to you, it lists the pages numbers right under the quote. The first is page 43 and the 2nd is pages 52 and 53. And I have seen many publication listing BE and IE and he values that they were was the BCC and ICC. If you are considering only one strategy change then the CC and Efficiency is the same. When you talk about PE you are grouping many strategy changes and so I am not sure how efficiency is calculated in those situations. What I did was take a (weighted) average CC. As Griffin said, CC are directly related to efficiency so as CC increases so does efficeincy.
Also, we don't care about HL w AA78mTc & 5m6c and KO with AA89mTc and 5m7c I don't know why you keep bring that up. This thread is about Hi-lo + 7m9c. CC is might be a legitimate way to determine count strength but it is not the method used to calculate and determine SCORE.
The reason I mentioned HL w AA78mTc & 6m6c and KO w AA89mTc and 5m7c sims is that the CC increased when I gave those to Gronbog to simulate and the results were that the SCORE increases. So larger CC means the SCORE increases as shown by every simluation I had Gronbog do for me.
If you didn't list any indices how does anyone know how to apply and implement your system?
I listed indices for HL w 7m9c in my posts. Remember, it is the HL w 7m9c for the top 6 plays and betting and HL for everythng else. I did not list HL w 7SC indices because I did not analyze HL w 7SC. I only wanted to compare CC of HL w 7SC and HL w 7m9c to show that the two systems were approxmiatley equal in power. I do not recomemnt HL w 7SC and did not analyze HL w 7SC thus I did not prodcue any indices for HL w 7SC because I do not like side counts of individual cards, I like plus/minus side counts.
Top HL w 7m9c
brc = betting running count = HL + ½*(7m9c)
Stand hard 14 v T if HL + 3*(7m9c) >= 10*dr
Surrender 8,8 v T DAS if HL + 2*(7m9c) >= 2*dr
Surrender hard 14 v 9 if HL + 2*(7m9c) >= 6*dr
Surrender hard 14 v T if HL + 2*(7m9c) >= 3*dr
Surrender hard 14 v A if HL + 2*(7m9c) >= 6*dr
Surrender hard 13 v T if HL + 2*(7m9c) >= 8*dr
All other situations use the stand along HL count and indices
Pretty simple for the HL player would wants to keep the HL count and indices with just a simple side count and a few strategy changes to increase HL efficiency.
Last edited by bjanalyst; 01-19-2020 at 07:05 PM.
You mention many publication listing BE and IE. What are the names of those publications?
Griffin said, in some case efficiency is directly related to correlation between the point values of the card counting system and the single card payoff approximating the blackjack situation considered. He didn't say it is alway the case in page 43 paragraph 1 fifth edition of The Theory of Blackjack. Efficiency is the ratio of the actual profit in card counting system to the total potential gain from information and interpretation of the unplayed cards. Betting correlation tells you how closely your card counting system estimates the advantage in betting.
Griffin said that ? = correlation coefficient it didn't say when ? increases so does the SCORE. I looked at page 52-53 and it doesn't say anything about SCORE. You are making that up. In what paragraph does Griffin say as ? increase so does SCORE? The correlation coefficient is the point value of the card counting system and the payoff of the game itself. I shows how to calculate correlation coefficient but never mention that as the correlation coefficient increase it would increase SCORE.
Firstly, you are looking at the 5th edition of The Theory of Blackjack which was published on 1996. The term SCORE was created by Don Schlesinger in 1999. So there is no way Griffin would mention anything about SCORE. Secondly, correlation coefficient doesn't compare risk and expectation which SCORE was made to evaluate. Your analysis doesn't even take into account the number of decks, penetration and bet spread. All I see was S17, DAS, LS. Okay, S17, DAS, LS for how many decks?
It better comparing the HL w 7SC and HL w 7m9c using simulation rather than doing CC calculations.
Last edited by seriousplayer; 01-19-2020 at 08:51 PM.
You mention many publication listing BE and IE. What are the names of those publications?
I do not have the publication but I remember seeing BE and IE and the values next to them were the BCC and ICC. But I am not an expect on the definition of efficiency. All I know is what I read in TOB where you saw Griffin say CC is directly related to and sometimes equal to efficiency. But you are getting off the track here with exact definitions. If someone else knows the exact definition of efficiency and how it related to CC please post it.
It better comparing the HL w 7SC and HL w 7m9c using simulation rather than doing CC.
Of course, simulations are always best. CC and EoR assume linearity and simulations do not. Simulations show everything, including non-linearity. But unless you are at less than one deck, linearity is a very good approximation for blackjack.
But all I can say is the CC correctly predicted the results of each simulation that Gronbog did for me before he did it. Also when Gronbog was analyzing HL w 7m9c he was trouble shooting his sim program to see which of the top 6 strategy changes was the problem and reduced the SCORE because of a coding error. So Gronbog individually tested betting and each of the top 6 strategy changes separately. For each of the top six strategy changes and for betting the CC increase over the HL and when sims were run the SCORE also increase on a item and item basis when each change was separately tested.
So you should used both CC and sims as a check on each other.
If CC increases, Efficiency increases and SCORE increases.
So using CC is an excellent way to predict which counts are better than other counts. Once you have preliminary indications with CC then you can see exactly the increase in SCORE with sims.
I have yet to find a single incident where the CC increased and the SCORE decreased. It has not happened. When Gronbog initially rant his sim program he saw eh SCORE decreased when I added my top 6. I told him there must be an error in his code and he testes each top 6 separately, as mentioned above, and he found his error. Both CC, Efficiency and SCORE all go in the same direction and they all are a check on each other to make sure the analysis was done correctly.
Disadvantages of Simulations:
- Take a long time to run and are tedious to set up.
- Require the calculation of indices for each playing strategy (another potential source of error) for the counts being compared.
- There is also the problem of variance with simulations which is reduced by increasing the number of hands simulated.
Advantages of Correlation Coefficients:
- WACC and BCC can be done in a matter of minutes.
- WACC and BCC are exact with zero variance.
- WACC and BCC used to compare various counts require no indices.
Correlation Coefficients are calculated with the tag values of the count being analyzed and the EoR. EoR are LSL estimates, so they assume that blackjack is linear. Simulations have no assumptions on blackjack being linear. With less than one deck is remaining, non-linearity kicks in.
Last edited by bjanalyst; 01-19-2020 at 09:09 PM.
Than you are making stuff up. Not that I am getting off track with definitions it is because you don't know your Blackjack and you are trying to act like an expert.
It better comparing the HL w 7SC and HL w 7m9c using simulation rather than doing CC.
Nothing in Blackjack is linear beside insurance.Again, nothing in Blackjack is linear beside insurance. You forgot to list the disadvantage of Correlation Coefficient:
Disadvantage of Correlation Coefficient
1. It doesn't calculate or give information about average bet.
2. It doesn't give info regarding win rate/round and win rates/hour.
3. It doesn't give info regarding standard deviation for hours and rounds played.
4. It doesn't give info on risk of ruin.
5. It doesn't give info on performance (DI, CE, CE/WR and SCORE).
6. It doesn't give info on the N0 it require to overcome one standard deviation.
7. It doesn't give info on the expect win per hour.
The list could go on and on and on....
Posting Permissions
-*-
Bookmarks