Simulation showed that counting performance is poor, how come?

[raymondc]

I just downloaded a Blackjack simulation program and assigned different betting strategies for 3 players on the same table playing 8-deck game. All 3 simulation players use the same Basic Strategy rule table.

At most times, the player which bet according to a true count bet size table performed no better than the other one which did flat betting.

The final player which used a progressive betting strategy (1 unit start; each consecutive win bet one more unit to a max of 5 units; back to 1 unit when lose 1 game) outperformed the other 2 players.

During the simulation the true count seldom went higher than 3. I think this might be a problem of uneven random-number generation in the program.

But could it somehow also indicate that the Hi-Lo Count method is not much effective in 8-deck games?

[KenSmith]

While an 8-deck game requires a large bet spread to be profitable, the Hi-Lo will still provide an advantage. The concept is well-established and not in doubt. If you're seeing results that dispute that, there's a problem with the simulation.

It takes a lot of hands to eliminate the luck factor and insure that you're seeing valid numbers. Perhaps you simmed too few hands to accomplish that, or perhaps there is a flaw in the software. What software is this?

[raymondc]

It's called "Simon Says Blackjack" by Accidental Software.
I found it by googling "blackjack simulation" last week.

It a game plus simulation. Perhaps I just misused the software. The software only allows free trial 6 times so I can't test it anymore.