Number of Trials To Prove Accurate?

Okay, down further under the post regarding John Patrick's BS there's a link to 2 100 million hand trial showing the difference between JP's BS and what is normally considered mathematically sound BS. My question is, how many trials are needed to get to the FINAL ANSWER. I'm not questioning the results of this study, but in my other goofing around exercises I've ran for other things I ran (several) trials of a million, 10 million and I believe 50 million respectively using Simsimp with slightly different results each time, and I was just wondering how many should I be doing to get a truly accurate picture?

Thanks!

I am sorry that I missed some of the posts about JP's strategy. I don't know exactly what his strategy is. (Reading those posts was not a high priority for me.)

But given any "basic" strategy, we can get an exact calculation of what it will do by using Combinatorial Analysis. CA is really ideal for these types of problems.

By a "basic" strategy, I mean any strategy which is based purely on the players cards and the dealers up-card. For example, "don't double if the dealer is hot" would NOT fit into this category.

If JKP has such a strategy, then any of us with CA programs (Cac, Steve, Bob F(?), myself ) could answer your question.

By a "basic" strategy, I mean any strategy which is based purely on the players cards and the dealers up-card. For example, "don't double if the dealer is hot" would NOT fit into this category.

This is the trouble with analysis of Patrick's methods, as they rely heavily on this type of reasoning. For example, he actually recommends the use of a count system, but gives specific advice to "not get your brains beat out" when the dealer is winning hand after hand regardless of how high the count is. It is impossible to pin it down.

Despite what you've written, anyone can buy a Patrick book (God forbid!) and turn to his multi-deck BS chart. It exists.

Compare that chart to the correct BS and use CA to compute the difference.

Piece of cake.

Don