So, imo, a long term winning system should be able to totally erase and invert the negative expectation on some selected P bets.
P bets cannot benefit of any positive mathematical factor like a part of B wagers, on the contrary they must bear it along the way.
We want to transform a 50.68/49.32 game into a perfect 50/50 game, hopefully deviated to the right term of the ratio.
Imo, there are many reasons to pick up the P side bets to assess whether our system is really working or not.
The most important reason is that P bets are payed even money and itlr there are no tricks to alter the registration of W and L distributions knowing that one side is underdog.
Unlike roulette, where a binomial proposition (R/B) could give the lucky player long term positive outcomes by SD issues (unfortunately destroyed by the zero/zeroes appearance), at baccarat any hand will provide an outcome placed on one chance or another (ties excluded, but they are neutral and not negative results).
Therefore, at baccarat a neutral or winning long term P bet placement is, imo, the best tool we could have to ascertain if our tests were the subproduct of luck or something else.
Admitting a no fixed game, we are 100% sure that in the long run the expected gap between B and P hands will approach with more and more precision the 50.68/49.32 expected ratio.
Hence, after 10.000 placed bets on P side we expect to lose an average of 136 bets (124 if we consider the "resolved bets", but I take the first value for simplicity).
After 100.000 bets, our P wagering will show an average of 1360 loss and after 1 million of P placed bets, we'll lose an average of 13.600 units.
Now I dare to state that if after several hundreds of thousands of P side placed bets a player is having a neutral result or a small profit, well, it means that he/she was able to utilize a good bet selection.
Anyway, moving such knowledge into the practical environment is a difficult task, despite of the appearances.
The B side is either more enticing as it's less unfavorite and more silly as we have to pay a 5% vig on many B winning bets not showing a given mathematical advtantage.
For example, 20 B winning bets not contemplating any asymmetrical hand will produce 1 sure unit loss.
Of course mathematics will tell us that itlr the best move to take is betting B as it's a 0.24% better move (meaning we'll lose 0.24% less than on P bets).
Good, as baccarat is a mathematically unbeatable game.
And mathematicians keep stating that any hand ON AVERAGE will be always 50.68/49.32 placed.
That's true, on average.
For example and giving a card composition topic (anyway not working at a substantial degree), a terminal deck particularly rich of 7s, 8s and 9s will provide a huge amount of symmetrical hands.
In that instance only a fool would bet the B side as the AS/S hands ratio will be much lower than expected.
Luckily, we don't need to counting cards because itlr the mere distribution of hands will help us.
It'll be a very difficult and very diluted task but we can make it.