The control of variance is impossible

It depends on what you consider as controllable.

Baccarat is a finite dependent asymmetrical card process, it's not a coin flip endless succession, thus not every A/B fight will produce the same correspondent sd values a CF game provides. And neither a static 50.68/49.32 A/B probability model will produce what really happens per every shoe of baccarat.

The deeper you analyze different A/B situations (random walks) the better the variance will be restrained, that is the pendulum swinging range.

I could decide to let it go some shoes not adhering to my statistical conclusions right at the start as no single, no streak, no pattern is equal to another one. It appears to be equal but it isn't.

Btw, thanks for your kind words.

as.