Systems Archive 4
[ Read Responses | Return to the Index ]
[ Previous | Previous in Thread | Next in Thread | Next ]
Re: Victor what have you... The perfect strategy
Posted by Victor on 20 February 2002, at 8:00 p.m., in response to Victor what have you decided to do? (nt), posted by T-Hopper on 20 February 2002, at 11:17 a.m.
I was not able yet to get a number (like the win rate) of an strategy that keep separated account of all cards.
I got in communication with the manufacturer of a blackjack machine, and he told me that with 8 decks, there was little difference with a good counting system, yet, he was not able to give me a number, so I can finally compare and decide.
So this is what I’m doing right now:
I have a software, that allow you “evaluate” a particular hand (let’s say 13 against a 6), according with a specific true count status. So, for example, you define a counting system, like Hi-Opt II, have the soft to play, with the basic strategy, and after millions of hands, you go to the evaluation chart, and the soft shows you how many hands the machine won by STANDING and at a particular true count status. For example, it says:
“with true count –10 it won 2343223 times and lost 2132134 times” and then the machine tells you what would have happened if you would HIT (always with the true count –10) and goes “321132423 won and 2324324 lost” and then it says the same data for true count –9 and so on. You define what range of true count you want to exanimate. And then,, based on the numbers, the machine tells you what is the best thing to do at every true count status.
I put basic strategy running like 50.000.000 hands with Hi-OptII counting system, I evaluated all hands, and guess what!, when checking the charts, the whole Hi-Opt II strategy was there. After millions of hands, all the values the machine found to be right at specific true counts, were the same that the booklet I bought of Hi-Opt II said. For example, in a 15 vs 3 situation, Hi-Opt II says that you stand if true count is –10 or higher. In the evaluation section of the simulation, after 50.000.000 hands, you can see that with a true count of –11 the machine wins more hands by HITTING, the same happens for true count of –12, -13 and so on. But, when true count is –10, the machine starts winning more hands by STANDING. The nice thing of this situation, is that the simulation is being run with 8 decks, and all the rules I told you.
So, once I saw that every counting system can be developed in this way (of course testing it like 100.000.000 hands, so as to be sure of the numbers), I developed a special account system for every type of hand… for example, when you have a 15, you loose with 10,9,8 and 7 and can hit with all the others. Now, suppose, that you have a special counting system where 10,9,8 and 7 are –1 and all the rest +1 (for example). This counting system would be useful for this type of hand but not for the other type of hands. But, if I will be able to count all the cards separately, I will have this information for every hand. So I run the simulation, and evaluate only the hands that you have 15 and what happens according with the different count status. At a certain point, there will be a true count value where it is good to HIT and besides this value, it is good to STAND. You know that value after millions of hands with perfect accuracy. In this way, I will have the strategy developed for all “15” situations. Then, I will make a counting system for “14” situations and will run a huge simulation again. Finally, I will have a strategy developed as if you had counted every card separately (i.e. you can tell how many 2, 3,6 etc.. has been dealt). Once I will have all these numbers, I will need to find a soft that let me run a simulation keeping count of all cards separatedly (which looks difficult to find) and if I cannot find such a soft, I will need to evaluate the winning rate of every type of hand, and then to compare it with Hi-OopII, to see what extra advantage I got, and of it worth the effort of building a whole machine or if with Hi-Opt II and side count, is the same thing.
Then, I will study the betting strategies you suggested me to read, in order to see what is the best way to go (I have just read somewhere that in order to play against 8 decks, bet spread should be 1-32, for example).
Then, I will go into the subject of shuffling, and will send you an e-mail with the way they do shuffling here, and we will see what piece of extra advantage we can get.
My main concern right now is to find a powerful soft that allows me to keep separated accounts and to define the strategy according to these counts. Otherwise, I need to do all the above calculations. Even if it looks a bit difficult, it can be done. If you know such a soft, please let me know.
Of course that if some mathematician who likes BJ, already tried this, it would be really helpful, but who knows where it can be.
What do you think about all this?
Thanks for keeping in communication with me.
The Systems Archive 4 is maintained with WebBBS 2.24.060398.