Simulation $$?
  • I have looked at few BJ books that use simulations to show how much $$ you can win by counting. In thinking about it, any time you can get an edge, no matter how small, (even .01 or less), and bet big with a few million hands the computer will show you will make money. That’s just basic arithmetic; any edge, no matter how small an edge will show a profit if you run enough hands. So to get a bigger $$ winning, just run more hands, anyone for a billion??
    So the better question becomes how real is the edge (player advantage number)?
    1.Where (how) is the number of +.14% (.0014) player advantage, when the player can Double after a Split come from?
    2.Where did resplit Aces at .07% (.0007) come from?
    3. etc,etc. on when the count goes up?
    In short, how were the various % advantages developed?
  • sage01 said:In thinking about it, any time you can get an edge, no matter how small, (even .01 or less), and bet big with a few million hands the computer will show you will make money.

    But that assumes that the player will play a few million hands without going broke! If you check the RoR calculations for that sim you will see that the player has a very large chance of going broke in relation to his EV. Since going broke early means having an EV of $0, that style of betting is very dangerous and not optimal.

    The fastest way to increase your bankroll is to bet proportionally to your advantage (with an adjustment for the variance). That will give you a much safer RoR, faster bankroll growth and get you to the long run sooner.

    sage01 said:In short, how were the various % advantages developed?

    The advantages you mentioned are basic strategy advantages. They show how much each rule will change the house edge for the basic strategy player. They are not always accurate for card counters. Don Schlesinger’s “Blackjack Attack” has a chapter about this.

    -Sonny-
  • I agree betting big with a very small advantage is not correct.
    But I was wondering about the advantage numbers.
    Where do they come from and why are they correct,
    Just saying the "computer sayes so" is a little weak!
    How do "they" decide that RSA are worth .07% advantage?
  • The computer will take identical games except in one you can resplit aces and in the other you can not. Then sim it for perhaps a million or two hands and compare the results. Then be able to replicate it by simming it again several times.
    Often they have what they call a standard game which has been throughly simmed with certain rules. Then take your rule changes and run the new game and compare the differences.
    So say you have a standard game of 6 deck, dealer stays on soft17, DAS, resplt to 4, DA and LSR, a .33% house edge. Now you add resplit aces, simulate it for enough hands and show a house edge of .26%, the rule change adding .07% for the player.

    I think a dead website for cardcounter.com had a sim computer on it that would show your ups and downs on a long sim. Often when simming a combination of rules, pen, spread and min bet, you might find yourself in the negative or in other words, have lost your whole bankroll.

    ihate17
  • sage01 said:
    I have looked at few BJ books that use simulations to show how much $$ you can win by counting. In thinking about it, any time you can get an edge, no matter how small, (even .01 or less), and bet big with a few million hands the computer will show you will make money. That’s just basic arithmetic; any edge, no matter how small an edge will show a profit if you run enough hands. So to get a bigger $$ winning, just run more hands, anyone for a billion??
    So the better question becomes how real is the edge (player advantage number)?
    1.Where (how) is the number of +.14% (.0014) player advantage, when the player can Double after a Split come from?
    2.Where did resplit Aces at .07% (.0007) come from?
    3. etc,etc. on when the count goes up?
    In short, how were the various % advantages developed?



    They are _not_ showing your gross wins. They are showing your EV (expected value). That's a different thing. The more hands you run, the more reliable the estimate of your EV. When you play, the more hands you play, the more you win because your expected win (or loss) is EV * #hands * average_bet

    But note that EV is not related to the number of hands. So a billion rounds in a sim doesn't make your EV bigger, it makes it more accurate.
  • Essentially sage, all of these good plays were deduced by allowing/not allowing them and seeing the effect.

    I am reminded of a question i once put forth to Qfit... If Aces have a negative running count when removed, isn't that a balance between the Aces that count as 1 in the hand, and those that count as 11 in the hand? Should we count them differently depending on their worth (one or eleven)?

    Of course we know that removing them all has a 'net' effect thats bad for the player. But some of them Aces ain't worth Jack.

    N&B
  • sage01- The expected returns on each hand at the wizardofodds.com, will help to understand the value of sims. 16 vs 10 is a good example because the ER for standing is -.5404 vs -.5398 for hitting. A difference of -.0006.
    That ain't much and to prove that hitting is the best option one would need a bunch of hands. Sims eliminate luck and as a result give a true picture of the facts. Now suppose that our 16 amounted to 88 vs 10. As you likely know spitting is the better play in this situation. And yes, proving that will take a lot of hands as compared to hitting or standing.

    Another value of sims is to indicate how things are likely to turn-out in the short-term. If everything else is correct in relation to your advantage (BR & bets), you should win more money than you lose over time. We play the short-term based on long-term results because it is the best we have. Only card counting can alter that.

Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!