Any baccarat system is based upon the probability to get A or B and if you have read my posts you know there are several ways to classify a given event as A or B.
Generally speaking, an A/B word is composed by deviations and equilibriums. The last situation could be classified into two different categories: real equilibriums and "false" equilibriums.
False equilibriums must be intended as those situations where the gap between two opposite events tend to "stall", meaning they are not gaining a sensible deviation toward one side or another despite the A/B gap is different to zero (ok, mathematically it's 1).
The deviation side is easier to evaluate: anytime A or B chance will deviate, we just have to take into account HOW MUCH or, preferably, HOW it deviates.
UNLESS THE GAME IS MATHEMATICALLY OR BY OTHER REASONS SHIFTED TOWARD ONE SIDE, the flow and the gap of these three different features itlr will tend to be equal to zero.
Notice that itlr we are certain that EC will be more or less deviated toward one side as the real equilibrium will be a kind of utopistic goal.
Nonetheless, regularly betting toward deviations without a sensible reason must be considered as a sort of "pushing the luck" strategy as such deviations (in a direction or another) are just the by product of the natural flow of the outcomes, meaning that the opposite false or real equilibrium situations are due.
Actually a world mainly composed by real equilibriums and/or limited deviations will be a perfect world to set up a simple progression.
On the other part, regularly betting toward deviations doesn't get the job as we cannot know at which degree and at which frequency (unless from a theorical point of view) such deviations will take place.
It's true that regularly betting toward real or false equilibriums without a reason doesn't make the job either as we could easily get multiple deviations intended as real deviations (going farther from a zero gap) or false deviations (partial RTM effect from a perfect zero gap) even more difficult to properly assess.
Therefore a so called "perfect" strategy cannot solely rely upon deviations or equilibriums without knowing how many times and how much a given opposite statistical situation had occurred.
Instead, we should place our confidence to get more reliable results about the EMPIRICAL probability to get favourable outcomes after having known that some expected results are due. As they must due by mathematics and common probability.
Talking about practice, let's say we want to set up a slow multilayered strategy based on the increased probability to get more P 4s than P 4+s (more Player 4 streaks than Player superior streaks).
Good, we know that itlr we'll get by 1 trillion certainty more P 4s than P 4+s, so we cannot be wrong.
Not a vulgar finding as roulette can't give us such certainty.
True, the vig itlr will cancel the validity of such finding, still we know that in this scenario A>B.
Technically and mathematically we know we will be more right than wrong whenever after a P 4 streak a Banker hand will show up by an expected or better probability to show an asymmetrical hand, a hand where B side is mathematically favorite.
Whenever after a P 4 streak the next hand will be symmetrical (50/50) placed, we know our confidence about getting a B hand will go to the toilet.
Since we know that the mathematical probability to get an asym hand after a P 4 streak is around 1/11, we know we'll be wrong 10 times over 11. No matter how we think to be genius to guess what will be the next hand, especially if we won the hand on a symmetrical spot (it's just luck!).
Nevertheless itlr we'll get the same amount of deviations, real equilibriums and false equilibriums occurring between P 4s and P 4+s. Still we have the luxury to know that real equilibriums are just utopistic or short term findings as there are no REAL equilibriums on those opposite events itlr.
That doesn't mean our strategy should be focused to stubbornly hoping for more expected outcomes, just to properly assess the "deviations-real equilibriums-false equilibriums" situations happened in the past as at baccarat they will have a slight different degree of showing up.
The more one or more features had deviated from the norm in regard of their expected probability, the better will be our results. Especially taking into account several multiple AB situations, knowing that at baccarat when certain A events will be heavily present some different and related counter B situations will be less proportionally present.