Back to The Blackjack Page
The Blackjack Page Archive
It's All in the Percentages

If you're counting cards at a Blackjack game, there are some small, incremental gains to be made by varying from Basic Strategy when to count gets high or low. For example, Basic Strategy says to always hit 15 against a dealer's up card of 10, but if the count is high (which indicates that there's a higher proportion of 10s than 'little' cards remaining to be played), then it makes more sense to stand. The point at which that occurs is called an 'index' and I've covered some of the most important in a prior lesson. (See "Basic Strategy Variations" in the archive at the bottom of this page.)

I recently downloaded the shareware version of a program which can help you understand why these changes are made as the count fluctuates. It's called Blackjack Plus Version 1.4 and I got it at www.zdnet.com/anchordesk/killerdownloads/index.shtml/. It's only $21 to register and, if the author works out some apparent kinks in the program, this could become an important tool. First, the good part: This program allows you to replicate most casino-playing conditions like double after split, dealer hits A-6, the number of decks used, etc. and it keeps an accurate Hi/Lo count which is converted to true count as the game progresses. When you are dealt a hand, a dialog box shows what your expectation is for playing the hand in different ways: hitting, standing, splitting, or doubling, as applicable. But that brings up the bad part. For example, in a 6-deck game where the dealer hits A-6, a hand of 9-4 versus a dealer's 6 shows that standing has an expectation of -.198; hitting -.234 and doubling -.469, when the true count is at 1. In Stanford Wong's book, "Professional Blackjack, he shows the expectation of all possible hands, but those are calculated at a count of zero, so there's bound to be some differences. For the above hand, Wong shows that standing has an expectation of -.140; hitting -.239 and doubling -.618. You can see the problem here; nobody in their right mind will ever double a 13 against a 6, but the calculation for one of these is off and I don't think it's Wong's. I suspect that the author of Blackjack Plus didn't take into account the amount of $$$ which are put on the table when one doubles. It seems that he just figured it on one bet, instead of two, but don't know for sure. Whatever the case, the basic concept behind this program is excellent, but it's not of much help if the numbers are incorrect except in the broadest terms. But, just as a tool for playing, this is a fine program with good graphics and smooth action.

If I knew how to create a program like this (and believe me when I tell you that I have the highest regard for those that do), I'd add a feature where a user could set a running count and the true count and then analyze a particular hand to see what the expectations are for the various options. Further, the program only allows bets of 5, 10, 25 and 100 units, but a more realistic spread of 5-60 would be a considerable improvement.

However, with all its flaws, this is a good start to what could be a great Blackjack tutor and, at a registration price of $21, is probably worth getting. At any rate, it won't cost you anything to give it a try; the shareware version is fully functional, but can be used only 10 times before you have to pay the asking price.

See you here next time.

RETURN TO THE CURRENT BLACKJACK ARTICLES