PDA

View Full Version : Blackjack


Felipe
07-21-2005, 12:39 PM
What is the probability that I will win the first black jack hand I play at?

LetYouDown
07-21-2005, 01:03 PM
Depends how you play. Assuming you play optimatally...somewhere around 49%.

BettyBoopAA
07-21-2005, 01:10 PM
it's less than that, or you would have an advantage. Ignoring ties (which is 9 %) of the hands, I think the answer is closer to low 40 to mid 40's

LetYouDown
07-21-2005, 01:15 PM
Semantics I suppose...but correct. I consider any tie a win when I play blackjack. Assuming I'm not counting cards.

IIAce
07-21-2005, 01:48 PM
I don't wanna hijack the thread but any brief explanation will do. I am a newbie to blackjack and just don't understand how the casino can get an edge in a kind of game like that where even the players know how the dealer will play his hand? Dealers hit if below 17 and stand with more than that (I think that's how it works). Wouldn't it just be 50/50 if the player played exactly like the dealer?

BettyBoopAA
07-21-2005, 01:58 PM
the house edge in BJ comes from:
when you both bust the house wins

SheetWise
07-21-2005, 02:17 PM
A tie is just masturbating with the dealer and changing the deck composition -- just don't count them.

Zero memory expectation with good rules +.0009, number of hands won vs. lost --- never thought about it. It is interesting to note though, that under no situations, even including splits and optimal play, will the player ever win more hands than the dealer -- even in extreme plus count situations. I've always found that interesting. The players + expectation is always considerably less than the advantage gained by doubling.

Just that I'd toss that in.

07-21-2005, 02:49 PM
A related question. I know a substantial portion of the player's equity comes from being able to split, double and surrender. Assuming I can't surrender and I have only one unit to bet with (so no splits/doubles) what is the house edge?

SheetWise
07-21-2005, 04:50 PM
I decided to run a simulation. I created a strategy chart with no doubling or splitting. I then ran 1 million deals with 4 players making flat bets of $1. on a four deck game (no insurance, since there is no count). There were 78,449 shuffles, representing 12.7 deals per shoe -- 9,630 hrs. of play each for 4 players.

It isn't pretty. The charted table results show a very straight line form 0 to -$97,338.50. Handle of $4,000,000, Loss rate -2.43%

LetYouDown
07-21-2005, 05:15 PM
What kind of application do you use to run these simulations or do you just quickly write your own?

SheetWise
07-21-2005, 05:25 PM
I've been writing gaming software since '74. I have some excellent 21 software in my library that's been tested over 25 years. The question made me think, because I didn't recall ever collecting similar metrics -- usually pretty focused on W/L. It was easy to run since over the years I've written GUIs for all the libraries.