Blackjack in Access

[TravisJRyan]

Hey,

I have another "im bored" project and was curious if anyone has ever found blackjack for access anywhere on the net or if they have created one. I wanted to find one already made so that I can create a programmable system through modules for A.I. in the game. I have a bet going with a friend here at work on who's A.I. will win. We just have to build the game first. . I personally wanted to just find one to modify to cut down on the time. Anyone?

 

[WayneRyan]

Travis,

A.I. is good stuff. Did a LISP Othello game some years back.

As for BlackJack:

Easy enough to get the cards.

Easy enough to deal the cards.

The overwhelming amount of work is to design your tables for "remembering"
each game and what you will store to make the game learn. Unlike most A.I.
apps, I don't think that BlackJack will have to do a multi-level lookahead
like Chess, Othello, etc. No recursion ... that's a plus!

Maybe the best approach would be to start with the typical "cheat" card
and as the games are played, increment counters to validate the card's
reccommendations.

Sorry, I don't have any building blocks for you though.

Just some thoughts,
Wayne

 

[TravisJRyan]

Well, dependent upon how many decks are used i have a strategy that I would like to test against the normal 17 stand dealer technique used in most casinos. Its technically a card counting technique so that is why I am challenging my friend to the A.I. contest. . .if you come across anything let me know. .Eventually I will get impatient and just build it myself. haha

 

[tkpstock]

I actually built a BlackJack game in Javascript, and another in Excel VBA a while back - I think I no longer have the VBA one. It was built in XL 95.

Why is AI involved? Either the dealer has to hit, or he has to stand based on the rules you choose. What learning could be involved? Or are you trying to develop a game that will not use the standard casino rules, rather it would attempt to adapt based on the previous cards played?

The two games I developed were simply based on Vegas rules - dealer must hit on a hard 16 and stay on a hard 17.

 

[WayneRyan]

Tom,

I think the player is the computer (not the dealer) ... right Travis?

Wayne

 

[The_Doc_Man]

While there are thousands of ideas you can use for "counting schemes" and strategies, the house rules were computed based on a really simple fact. By putting the "hit on 16, stand on 17" rule (plus rules on ties and rules on when to reload the "shoe"), the house makes a profit in the long run.

Looking for a winning short-term strategy violates one or more of the laws of large numbers. Looking for a different long-term strategy (i.e. not the same rule as the house uses) is already known to not work because of the secondary rules on ties.

If you want to do an "I'm bored" project, try doing BRIDGE, CANASTA, POKER, RUMMY, WHIST, HEARTS, SPADES, CRIBBAGE, or PINOCHLE, any of which have more strategy involving either hand evaluation or selection of the next play or for betting games, how much to bet. Kenny Rogers had the right viewpoint in The Gambler (song, not movie). If you "know when to hold 'em and when to fold 'em" then you've got the game licked.