I have to admit - initially, I considered it. It became clear though that it wasn’t you. Perhaps you guys share common dna allelesOriginally Posted by Three
I do not have simulation software. I had suggested in a previous thread that someone who has a canned HL simluation software make only six changes.
I am including at the end of this post weighted average CC that I used to compare the strength of various counts. Note that the Weighted Average CC agrees with simulation results for strength of counts. So the weighted average CC is an excellent tool to determine the ranking of the strength of various counts.
Use HL for betting and for all strategy changes except use Tc = pseudo Ten count = HL + AA78mTc for: (dr = decks remaining)
(1) insure if Tc >= 4*dr
(2) stand on hard 12 v 2 if Tc >= 4*dr
(3) stand on hard 12 v 3 if Tc >= 2*dr
(4) stand on hard 12 v 4 if Tc >= 0
(5) stand on hard 12 v 5 if Tc >= (-2)*dr
(6) stand on hard 12 v 6 if Tc >= (-1)*dr
So just have the original HL simulation and another copy with these six changes. Run them both and compare as the only difference between the two is using the Tc instead of he HL for these six situations I mentioned above.
And I prefer CC to simulations as the answers are immediate and exact and do not depend on betting patters or other variables. You are isolating PE (playing efficiency) with weighed average CC. I also calculated weighted average CC for various count systems and sorted the counts by weighted CC and the lowest weighted CC (for playing strategy) was HL and then KO and then Hi Opt 1 and then Hi Opt 2 and the strongest Hi Opt 2 with side count of Aces. So weighted CC agrees with simulations results.
Also you do not need simulations for common sense. A Ten count is what should be used for insurance. I do not need a sim to show that. Also a Ten count should be used for hit/stand decisions on hard 12 v 2, 3, 4, 5 and 6. This is not difficult to figure out.
But if someone has a canned HL simulation that they can modify and run with just the six changes I listed above, I would be interested in the results. No one who replied to my posts said they had such a canned HL simulation program. If you did have such a program and ran it, I predict that your results would show HL with AA78mTc is just as powerful as HI Opt 2 with side count of Aces.
Actually I chose only the six most powerful changes to HL using AA78mTc. There are more changes but I did not include them in my requested simulations because I wanted to keep the changes to the HL program to a minimum to reduce the chances of errors when changing the program. And the six chosen are the most important and will prove my point.
Because I did not include all of the AA78mTc changes to the HL perhaps the results with just the six changes will come somewhat less that HI Opt 2 with side count of Ace but my prediction is that it will be very close.
CC Comparsions.jpg
Last edited by bjanalyst; 12-23-2018 at 11:30 AM.
That's your job. Literally. If you have a system, you need to provided some context as to what the reader is to expect when using said system. All other information is useless to us.
Repeating the data as nauseam does nothing to convince anyone.Use HL for betting and for all strategy changes except use Tc = pseudo Ten count = HL + AA78mTc for: (dr = decks remaining)
(1) insure if Tc >= 4*dr
(2) stand on hard 12 v 2 if Tc >= 4*dr
(3) stand on hard 12 v 3 if Tc >= 2*dr
(4) stand on hard 12 v 4 if Tc >= 0
(5) stand on hard 12 v 5 if Tc >= (-2)*dr
(6) stand on hard 12 v 6 if Tc >= (-1)*dr
So just have the original HL simulation and another copy with these six changes. Run them both and compare as the only difference between the two is using the Tc instead of he HL for these six situations I mentioned above.
Also, please read Griffin chapeter 3. He provides a chart to show where most of your gains will be. Notice the hard totals from 12-16 for dealer up-card T plus 13 vs dealer up-card 2-6. That is where most of your playing gains will be.
Okay. However, simply using CC does nothing from a win-rate perspective. What is the average expectation for each deviation as well as their respected probabilites? What can we expect with hitting/standing/doubling hard 16 vs T? CC gives us no indication as to what our respected return is. Simulations can.And I prefer CC to simulations as the answers are immediate and exact and do not depend on betting patters or other variables. You are isolating PE (playing efficiency) with weighed average CC. I also calculated weighted average CC for various count systems and sorted the counts by weighted CC and the lowest weighted CC (for playing strategy) was HL and then KO and then Hi Opt 1 and then Hi Opt 2 and the strongest Hi Opt 2 with side count of Aces. So weighted CC agrees with simulations results.
Invocation of 'common sense' as a premise for your argument is weak. Why not use a 5 vs all count for hard 16 vs T? Why not use an Ace vs Face count for hard 10/11 doubles? Those are also 'common sense' approaches. And, if it is 'common sense', then it should be easy for you to sim your ideas. So, yes, a sim is needed. What is your aversion to running a Monte Carlo? Can't you do that in Excel?Also you do not need simulations for common sense. A Ten count is what should be used for insurance. I do not need a sim to show that. Also a Ten count should be used for hit/stand decisions on hard 12 v 2, 3, 4, 5 and 6. This is not difficult to figure out.
Because, as I have said *multiple time already*, the burden of proof falls upon you! No one here thinks your system is the next greatest thing since sliced bread. They are comfortable with what they have as of now. If YOU sim your system and prove beyond a shadow of a reasonable doubt that is outperforms all other systems under the given win rate (NOT CC!), then we may be more invested in your idea. Until then, run a sim for your system.But if someone has a canned HL simulation that they can modify and run with just the six changes I listed above, I would be interested in the results. No one who replied to my posts said they had such a canned HL simulation program. If you did have such a program and ran it, I predict that your results would show HL with AA78mTc is just as powerful as HI Opt 2 with side count of Aces.
And how do you know there are the "most powerful" out of all other indices? And please don't invoke the CC nonsense. We want to know win-rate/expectation.Actually I chose on the six most powerful changes to HL using AA78mTc. There are more changes but I did not include them in my requested simulations because I wanted to keep the changes to the HL program to a minimum to reduce the changes of errors when changing the program. And the six chose are the most important and will prove my point.
I just presented my findings. If you do not believe me then stick with what you have. I am not going to try to convert anyone. You make your own decisions.
I have done everyhting short of DIRECT simulations but I sort of did INDIRECT simulations in that my program I used to calculated indices I used for HL indices and every HL index that I tested, the resulting indices from my program using EoR, agreed with the published HL from simulations.
Also I showed you weighted average CC for various counts and you can see that if you arrange the counts by weighed average CC that the counts with the highest weighted Average CC also are the counts that have been determined to have the strongest playing strategy according to simulations.
So I am not just using common sense. I used Effects of Removal from Don Schlesinger's Blackjack Attack, 3rd edition, that were calculated to six significant figures and Don explains in his book how important EoR are and how to use them.
So I have tested my results through indirect simulation of calculated HL indices using EoR agreeing with published indices for the HL and that the order of strength of various counts using the Weighted Average CC agrees with simulation order of strength for the various counts and I used Schlesinger's proven EoR for my calculations.
So I have done everything short of direct simulations of my results. And I also showed the value of camouflage play using AA78mTc with HL.
That is all I have to say. If you don't believe me then keep on using what you have. I am not twisting anyone's arm to believe me and I do not have simulation software to do a direct simulation to prove my results.
I do keep an open mind on many topics. I also lack patience in the nanner of certain presented scenarios. That being said, this is the type of theoretical discussion that I would prefer over a long afternoon, with a notepad and pastry.
Just because I’m not particularly interested at this time, does not mean that I won’t be at a future time, or that I don’t see potential merit to your ideas.
Hence, why the relevance of actually providing a monte carlo sim data-sheet would be beneficial!
I would *highly* recommend that the OP (bjanalyst) obtain the CV suite sim from our site host. Could help you expand into other ideas as well. No need to program. Just plug and chug.
I do play with both cvdata and cvcx. Even our resident master, Dog Hand, would not be able to sim my exact game. Shortcuts and assumptions would need to be made, which would throw off the actuals. That being said, I’m happy with my results and the avoidance of ridiculous debate, such as this, that would ensue if I were to promulgate my theories.
A well respected poster on this board did suggest to me a method by which I could sim. His commentary was based on tracking performance in a team environment. My preference is to do my own thing as a solo artist, and thus, I did not incorporate the suggestions.
Posting Permissions
Bookmarks