Thanks KFB, I appreciate a lot your comment.
After all you are one the few "coin flip" scholars (a real expert, I mean) capable to know when things could get us an advantage after the inevitable HE we are destined to face.

An important quote I particularly like is one from Alrelax that more or less sounds as:

"Do not bet what you desire will happen".

Paraphrasing it, that means that long term statistical data surely help us but at baccarat we shouldn't forget that we have to deal with many "short term" distributions easily deviating from the "norm".
Thus in those natural situations we have to set up a kind of perfect compliance over the actual results (AR) in relationship of the most expected results (ER).

Sometimes AR seem to be so deviated that any devised ER plan need a lot of hands (or shoes) to get its full power.
On the other end, the ER strenght acts constant and almost always toward the house.   

More later

as.

"Do not wager for what you desire.  Wager with what is happening"

Because, what you desire probably the highest majority of the time will not occur on your wagered hand(s).  It very well might occur while you are standing, seated, waiting, etc., and not wagering on anything, but that is how most all will get sucked in. 

Short term bursts of 'whatever is being presented' is by far the most lucrative wagers to place, IMO.

The entire point of playing is to win money.  Not to prove therory, statistics or pattern/trend triggering or anything of the like. 

Thanks Al, here is the correct quote:

"Do not wager for what you desire.  Wager with what is happening"

Of course that line of reasoning is the basic approach every bac player will adopt and without the proper experience we know the results...

So a more careful strategy would be to put within ranges what is more likely to happen with what actually happens.

In our opinion cutting off many hands from the betting opportunities is the first key point to transform a sure loser into a possible advantaged player.

The second important tool is to control the betting amount: We'll never know precisely (say in the vast majority of the times, at least) when our probability to win is valuably shifted toward our favor, even after having assessed that a same situation running infinite times will get us an edge.

Third, constantly assessing the average sd values even by a general point of view (expected results) and by what is showing up at the shoe we're playing at (actual results).

Fourth, when we have reasons to think that the production isn't so random (basically the vast majority of the actual scenarios), dilute at most our betting unless a univocal line seems to weirdly predominate what we're more likely to expect on average.

Now let's consider those points from the casino's point of view.

1- More hands a player will bet, higher will be our (casino) return as the HE constantly works at our favor.
At any rate, we (casinos) can safely rule out that a selected strategy will get a player any edge: itlr every player will lose, maybe less but keep losing.

2- Variations of standard betting amount, imo, should be limited within percentages of it (say 10% or 20%).
At a general negative edge game, it's always better to consider losing situations than winning situations.
Doubling or even tripling up (or more) a standard bet needs a very long experience and of course strong data to support such move.
We think that it's way better to win less and lose less in a particular spot than losing a lot or winning a lot in the same spot.
Of course there are some (very rare) players capable to get the best of such progressive betting, mainly because they approximate at best the spots to get a kind of advantage.

Anyway any player raising his bet is loved by casinos.

3- Sd (standard deviation) values are the watchdog of randomness but most part of the actual shoes we have to face aren't random by any means.
I've already stated that we have strong scientifical reasons to think that RNG instructed productions are not random and I've provided some examples.
Therefore, consider a shuffled machine shoe as a strong "actual" production weirdly shifted toward one side or another of the AS or S world by strong clumps or deviations.
Basically and paradoxically such productions will elicit an abnormal number of S patterns intertwined by few AS situations (despite of the 0.75 AS probability), but good news is that some shoes will produce interminable sequences of AS streaks up to the point that the entire shoe is asymmetrically shaped (a thing that we haven't encounterd once at real shuffled shoes).

Seen from the casino part, S patterns could get the players a distorting perspective that symmetry tends to overcome the rest up to the point that some shoes will totally deny it by producing long asymmetrical successions that are not so easy to detect (unless one had read my pages).     

4- Shuffle machines dislike the "average" word (finite and dependent random models do like averages) and when "average" doesn't act at the shoes we're playing at it's better not to risk any money or to ride the situations when the S patterns seem to be "too much" silent.
In any instance I'd suggest to bet toward symmetry, let such symmetrical weird long patterns go without risking a dime from our part, instead waiting for the relatively rare shoes where AS strongly overcome the S counterpart.

as.