October 21st, 2010 at 11:13:48 AM
permalink

Me being interested in the "why" behind everything, my question is.. how do you calculate, for instance, P(17|2) for the dealer, assuming 6 deck S17? I tried Bayes, but it doesnt really help me.

Could anyone shed some light on this?

Could anyone shed some light on this?

October 21st, 2010 at 11:24:46 AM
permalink

I'd think you'd need to create a tree of all the possible hole cards and hit cards, along with their probabilities. Then add up all the branches of the tree that lead to a 17. I'd think that a spreadsheet would be especially useful for doing this.

October 23rd, 2010 at 3:34:10 AM
permalink

Meh, that is a whole lotta work. Guess I'll have to do that then.

October 23rd, 2010 at 6:44:44 AM
permalink

Quote:PapaChubbyI'd think you'd need to create a tree of all the possible hole cards and hit cards, along with their probabilities. Then add up all the branches of the tree that lead to a 17. I'd think that a spreadsheet would be especially useful for doing this.

Yep. If you can live with an infinite deck assumption it can be done in Excel pretty easily.

It's not whether you win or lose; it's whether or not you had a good bet.

October 25th, 2010 at 3:54:47 AM
permalink

And if I wanted to do it for a 6-deck? Would simulation be the best option?

October 25th, 2010 at 6:29:26 AM
permalink

Quote:JannerAnd if I wanted to do it for a 6-deck? Would simulation be the best option?

It is an option, but I don't think it is required because of the 6-deck restriction. Let's say you're building the tree, and you're already on a branch which includes a two and jack for a dealer total of 12. You're building the branch which adds another two to the hand to make 14. For infinite decks, the probability assigned to this branch is 1/13. For six decks, the probability is 23/310. It'll take a little extra effort to build the tree for six decks vs. infinite decks, but its not prohibitive.