Some of the features of the program I am presently working on are: Computes 1-8 decks Computes total dependent expected values Optionally computes expected values for normal play Optionally computes expected values given dealer's hole card is known Optionally computes expected values given dealer preferentially selects A or T for up card Computes pre-deal overall EV for any shoe state using basic strategy User has choice of computing individual hand expected values assuming either dealer has checked for blackjack and doesn't have it or dealer blackjack results in a player loss (unless player has blackjack) Marks the correct basic strategy play An asterisk is appended to marked BS play if a better play is available User can enter as many counting systems as he wishes to be referenced by the program Up to 2 counting systems at a time can be displayed in the main program User can true count any counting system (default) User can set initial running count for any counting system Displayed counting systems can be changed at any time Rules can be changed at any time (mostly) Has 2 modes - both display card images 1) User controls input Computes fully dealt or partially dealt hand expected values User can manually override basic strategy if he chooses A new hand may be entered just removing current player and dealer cards, so it is convenient for online play Shoe can be reset at any time; again convenient for online play 2) Continuous random deal User sets penetration level User can choose to compute an overall EV before each deal User controls amount of bet Expected values are displayed for each player decision (basic strat) Alternative TD expected values can be calculated by selecting manual (In some cases, will exceed BS EVs) User makes all strategy decisions (stand, hit, double, split) Up to 2 counting systems can be displayed Keyboard shotcuts make it easy to play
As someone with a low bankroll, I became interested in ways to see exactly what was happening with regard to a player who relies heavily on basic strategy while counting cards, spreading his bets, and making some index plays. By exactly, I don't mean statistically exact as in the results obtained by a sim, but exact on a case by case basis. There is nothing wrong with a statistical analysis by sim, but that's not so much what I wanted to see. In the future I am considering the feasibility of entering the results of all randomly dealt hands played by the user in a database. If that could be successfully done, the user could query the database with regard to a given count system. The query criteria could be false positives or false negatives, for instance or maybe a given total versus a given up card at a given pen. There would be many possibilities and all data would come from exact calculations. Even without database capabilities, however, I still think the program has a lot of merit.