math excess
OK, Ken, you got me!
Player 4 won't, or at least shouldn't, stand if both players 1 & 2 double.
Of course you are right - no matter what outcome the doubles have.
The other problem is that you unfortunately can't multiply out the probabilities like this: "So you advance with a tie only in 56.8*62.8*70=25%."
The 56.8, 62.8 and 70 numbers are not independent events. They're all correlated because of the common dealer that all these players are competing against.
True again, I admit.
For example, your percentage of success for player 2's double (43.2%) and player 5's double (30%) are mostly overlapping when the dealer busts.
Even better - when the dealer busts, the whole calculation is obsolete. We are considering the case that Player 1 has a standoff!
So is goes this way:
Player 3 has me inevitably. I need to be ahead of the all doubling Players 2, 4, and 5.
Dealer 21 - Player 1 21:
Only Player 4 can beat me and only with doubling to 21 (92.0% advance chance).
Dealer 20 - Player 1 20:
If at least 1 out of the 3 players 2, 4, and 5 double to 21, I do not advance. Also if player 4 doubles to 20.
Player 2 has 8, needn't be considered.
Player 4 20 or 21: 15.9% -> player 4 below 20 or busted: 84.1%
Player 5: 21: 8.0% -> player 5 below 21 or busted: 92%
-> Advance chance ((player 4 below 20 or busted) and (player 5 below 21 or busted)): 77.4% (These events /are/ independent, aren't they?)
Dealer 19 - Player 1 19:
Almost same as above.
Player 4 19 to 21: 23.6% -> player 4 below 19 or busted: 76.4%
Player 5 20 to 21: 15.9% -> player 5 below 20 or busted: 84.1%
-> Advance chance ((player 4 below 19 or busted) and (player 5 below 20 or busted)): 64.3%
Dealer 18 - Player 1 18:
Almost same figure, but player 2 can also beat me with an ace...
Player 2 19: 7.6% -> player 2 below 19: 92.4%
Player 4 18 to 21: 31.6% -> player 4 below 18 or busted: 68.4%
Player 5 19 to 21: 23.6% -> player 5 below 19 or busted: 76.4%
-> Advance chance ((player 2 below 19) and (player 4 below 18 or busted) and (player 5 below 19 or busted)): 48.3%
Dealer 17 - Player 1 17:
Player 2 18 or 19: 38.2% -> player 2 below 18: 61.8%
Player 4 17 to 21: 39.2% -> player 4 below 17 or busted: 60.8%
Player 5 18 to 21: 31.6% -> player 5 below 18 or busted: 68.4%
-> Advance chance ((player 2 below 18) and (player 4 stiff or busted) and (player 5 below 18 or busted)): 25.7%
Now multiplied with my formerly calculated respective standoff probabilities:
0.0357 * 0.920 +
0.0097 * 0.774 +
0.0099 * 0.643 +
0.0105 * 0.483 +
0.0102 * 0.257 = 5.2%
Overall advance probabilityfor player 1 when doubling (Ken, please correct me): 59.0% + 5.2% =
64.2%.
Interesting. Seems like you shoulda doubled, Rick.
Tournament probabilities get unwieldy in a hurry.
Too true! Some surgeon with us? I would very much appreciate a cyborg-style coprocessor for my next tournament...