Strategy for Must Win 2 Bets

Discussion in 'Blackjack Tournament Strategy' started by gronbog, Apr 10, 2017.

  1. gronbog

    gronbog Top Member

    In another thread, Hopinglarry asked the following question:

    "It might be interesting to know what the probability might be for a person to win a blind Double (for lack of a better term) where the cards have not been dealt yet."

    I started a similar thread a few years ago to discuss strategies for when you must win at least one bet

    https://www.blackjacktournaments.com/threads/strategies-for-must-win-one-bet.8201/

    and I though that this topic should have its own thread for ease of reference.

    Similar to my optimal strategy for Must Win (at least) One Bet, I have a strategy for Must Win (at least) 2 Bets. I recently re-ran the simulation in order to obtain an estimated success rate. Here is the latest result:

    http://gronbog.org/results/blackjac...one/generated/complete/2.0/p1.X/strategy.html

    Astute observers may notice one small difference from the previously posted version, which is that A,A vs A is now a double where it was a split before. This latest sim was run for much longer than the previous one and so, I believe that the latest version is more correct. The main reason for regenerating the strategy was to obtain the estimated success rate which is 32.2326%.

    This assumes that you can double down on a blackjack and on any soft 21 that you end up with after splitting. I know of at least one tournament where you can do neither, and so I ran a couple of variations:

    Blackjack Pays 2 to 1: 34.2393% (+2.0067%)
    Can't Double Blackjack: 29.7169% (-2.5157%)
    Can't Double any 21: 29.7166% (-2.5160%)

    Of these variations, only the first affects the initial hand strategy (we now don't double blackjack).

    As before, we need some reasonable simplification for playing the split hands. One reasonable simplification is to play a "no bust" strategy for these hands. Since losing a double on the first hand would very likely cripple the chances of success, I also propose that this "no bust" strategy should not double or split any hands. Since soft hands are already no-bust by definition, I propose playing these hands using basic strategy.

    Playing the initial hand as in my table and playing the split hands as described above results in a success rate of 32.0750% or 0.1576% worse than optimal. Not bad for a very simple strategy. Since none of the variations above are affected by a no-bust strategy on split hands, you can subtract this 0.1576% from each to get the updated success rate for each.

    Thoughts, ideas, errors? Discussion is welcome!
     
  2. London Colin

    London Colin Top Member

    It's quite surprising that the cost of the simplification is so low. It looks like there isn't much point in memorizing a more complicated strategy!

    Wong recommends the no-bust approach and quotes a success rate of 33%.

    Ken gives his strategy for the first split hand in this old thread - Splitting and then standing on any stiff

    It's funny, I can't remember what I had for breakfast yesterday, but I can remember a thread from ten years ago! :)
     
  3. gronbog

    gronbog Top Member

    Thanks for the link to Ken's comments. I'm interested to compare my optimal strategy to his.
     

Share This Page