The long term money you'll win is a by product of the money you haven't lost

As banal as this statement could appear actually is one of the best recipes to consistently collect profits.

Suppose we have two systems to adopt:

a) The preferred system, that is one we particularly like for whatever reasons;

b) The perfect opposite system.

We know that besides some very rare circumstances (of course privileging the (a) method, I guess), both systems are losers itlr for the HE.

But after having tested hundreds and hundreds of shoes shuffled in the same fashion, we'll see that at both systems W and L streaks will show up or stop by a better than expected probability in relationship of how things went so far in the actual shoe.

Obviously it seems better to bet toward positive a) system streaks, yet those streaks are (almost) equally likely to show up while using the b) system so becoming losing streaks.

Cutting a long story short, in order to get long term wins we have to fictionally cut off an equal or almost equal number of losses (that is wins on the other system's side), but this procedure must be partially insensitive of the previous shoe(s) distribution.
The "ploy" not to expect something more due than average without assessing current patterns was set up by not ruling out the possibility (albeit very rare) that the same or correspondent card distribution will be dealt twice or more.
An heaven when things went good at our favor but a terrible and worse nightmare in the opposite scenario.

Therefore and simplyfing a lot, back to back a) or b) system "long" streaks are not coming out around the corner and even less likely are the situations where a kind of overalternating mood stands for "long".
Nothing new: for example run 6 coin flip hands and symmetrical quality patterns are just 4 possibilites (ABABAB, BABABA, AAABBB, BBBAAA) out of 64 possible results.
So 60 patterns are making asymmetry either from a quality (e.g. AABABB) or quantity (e.g. AAAABB) point of view.
 
In real baccarat terms, it's the distribution of back-to-back patterns that counts, in the sense that once a symmetrical pattern had shown up we have to wait for a pattern breaking the previous symmetry then assessing whether the symmetry will stand another time for the new different pattern or conceding room to a more probable asymmetrical one.

Utilizing a 0.75% probability oriented to get an asym pattern (and providing the action of proper random walks), you'll see that first and second winning attempts will come out clustered at least one time (so negating long isolated successions) by percentages getting astounding levels of probability (so profitable edge to exploit) as, again, asymmetry is the norm at baccarat.

An opposite less desirable system to utilize (so privileging the less likely symmetry) should be employed with caution once asymmetrical patterns seem to be silent in relationship of the number of hands dealt, especially when we suspect that the production wasn't randomly distributed.

as.