In last week’s column, I explained how in order to be considered a valid strategy, it needs to be validated mathematically. I can make up a strategy, walk into a casino and win 9 times out of 10 and this doesn’t mean that there is any validity to it.
On the other side of things, it may not be possible for a human to play a strategy devised by the computer or a math model. Sometimes, the strategy is so complex that it is not realistic to expect a human to play at the same level. That said, however, this does not mean that the strategy is not valid, nor should it be discarded.
A few years ago, when I was working on the game of Ultimate Texas Hold’em (UTH) for Shuffle Master, I devised a strategy for the game. Because of the nature of UTH, it is not possible to look at every possible hand. With a total of 9 cards being dealt, in the fashion that they are, there are more than 27.8 TRILLION possible hand combinations.
I had to use a variety of short cuts and had to summarize hands in a way that would be useful for a player. This was made very clear in my math reports and I acknowledged that the actual payback of UTH might be a smidge higher than I reported, but that it would be very difficult for a human to actually play at this level – perhaps even impossible.
My report contained the strategy that was required in order to obtain the reported payback, and I have gone over this strategy in this column. It didn’t take very long before somebody else claimed to have analyzed every possible hand and stated that my payback was about .25% below the perfect theoretical payback. However, this person admitted that he had no useful strategy to show from his work except for the strategy associated with when to wager after the pocket cards are dealt. His strategy matched mine.
I don’t want anyone to think that I’m criticizing this gentleman for his work. Quite the opposite. I’m commending him for incredible tenacity in accomplishing it, and it indirectly confirmed my own work.
My point is that if you determine that a game’s payback is 98.5% but can’t tell anyone how to achieve that payback, it has limited value. For a strategy to really be useful it has to be one that players can actually use.
So, while it may be important to know that a game has a computer perfect theoretical payback of 98.5%, if realistically a player can only achieve 98.35% using a complex but useable strategy, then I’d rather have the knowledge of that 98.35% WITH the useable strategy instead of the 98.5% with nothing else I can use.
Of course, there is frequently no absolute right or wrong in this regard. There are many people who probably consider the basic strategy for blackjack to be too complex and attempt to play an easier one. Then there are the diehards and card counters who probably try and play the game taking into consideration the specific hand make up (i.e. a 16 consisting of 10-6 vs. a 9-7).
In similar fashion, in video poker, we know that one prominent player promotes very heavily the incorporation of ‘penalty cards’ in his strategy. My father, when first developing his Expert Strategy for Video Poker, chose to ignore these, feeling that this would be too complex for the masses.
If someone is able to master a more complex strategy and eke out a better return, more power to him. At the same time, you have to be careful not to try and bite off more than you can chew. If you attempt to memorize a more complex strategy and it results in more errors, you may find yourself taking one step forward and two steps backwards.
Our Expert Strategy was never promised to be ‘perfect’ strategy for every game. In many cases, it is. In almost all cases, even where it is not ‘perfect’ strategy, it provides a simpler strategy at very small cost.
In ALL cases, it has been developed using sound mathematical principles and/or proper computer models. If your strategy can’t guarantee this last one, it isn’t worth the paper it is written on!
You can try out your strategy by playing our video poker game.
Question? Comment? E-mail me at: Elliot Frome