Theoretically best plays and bets, where can I find that?
There is a theoratically, though no human can do it, a perfect way to play, that is tracking each card separatly and betting and playing accordingly. How can I find information on that? For the no hole card game and hole card game.
It simply does not exist....in the real human/animal world.
It does exist in the digital world.
Or at least to be skilled enough to come reasonably close.
Poker games don't use six deck shuffling machines but poker players are developing skills that could probably come close to the optimal skill level.
Quote: kimuraHi everyone,
There is a theoratically, though no human can do it, a perfect way to play, that is tracking each card separatly and betting and playing accordingly. How can I find information on that? For the no hole card game and hole card game.
Tarzan. I believe he posts here (rarely). He posts on Norm's website, blackjacktheforum.com. I don't know quite how it works, but it's something like this:
He tracks groups of cards, 2-4, 5-6, 7-8, 9, T, A [all separately]. Not sure if those are the exact groups, but it's something like that. Last I heard, he was working on figuring out a way to compile all his information into one "system" and write a book or publish a paper on it. It isn't some easy system where you bet according to the true or running count, or a true-count adjusted by excess (or lack of) Aces. It seems like it's constant calculations.
If you've ever played poker and needed to calculate the number of "outs" to win a hand....I imagine his system is very similar. Not necessary that he counts the # of outs, but calculates the EOR's (Effect of Removal) of each group for each hand-match up.
A simple one be 14vT. The more 7's that are remaining, the better it is going to be to hit. More 7's means the player will be more likely to get a 21....and that also means the dealer is more likely to have a 7 in the hole (in which case, drawing a 3 would push, and drawing a 4-7 would win). A similar calculation would be done for excess or deficiencies of other cards or groups of cards.
Most counts don't have this type of precision because they group many cards together and cards cancel each other out in a way that isn't optimal for playing decision while a good chunk of cards aren't even counted (7-9).
Quote: RSTarzan. I believe he posts here (rarely). He posts on Norm's website, blackjacktheforum.com. I don't know quite how it works, but it's something like this:
He tracks groups of cards, 2-4, 5-6, 7-8, 9, T, A [all separately]. Not sure if those are the exact groups, but it's something like that. Last I heard, he was working on figuring out a way to compile all his information into one "system" and write a book or publish a paper on it. It isn't some easy system where you bet according to the true or running count, or a true-count adjusted by excess (or lack of) Aces. It seems like it's constant calculations.
If you've ever played poker and needed to calculate the number of "outs" to win a hand....I imagine his system is very similar. Not necessary that he counts the # of outs, but calculates the EOR's (Effect of Removal) of each group for each hand-match up.
A simple one be 14vT. The more 7's that are remaining, the better it is going to be to hit. More 7's means the player will be more likely to get a 21....and that also means the dealer is more likely to have a 7 in the hole (in which case, drawing a 3 would push, and drawing a 4-7 would win). A similar calculation would be done for excess or deficiencies of other cards or groups of cards.
Most counts don't have this type of precision because they group many cards together and cards cancel each other out in a way that isn't optimal for playing decision while a good chunk of cards aren't even counted (7-9).
This is of course more sophisticated than but a little bit close to the classical counting systems who track in general 3 groups of cards(high, neutral and low). I am speaking about tracking each card separatly.
To implement, you'd keep 13 running counts. I imagine the baseline would be composition dependent basic strategy, with about 2000 indices.
Unless you can do some crazy end-decking, it's still not a guaranteed win.
edit: 10 running counts should do, instead of 13.
