Step #1.
To be ahead of the game we have to get more wins than losses. Easier to say than done. But any bac player should remember this.
We do not need to be rocket scientists to do that: itlr betting Banker and its related distributions will provide more W than L. Period.
The obvious problem is that any W won't be counterbalanced by the L weight but only FROM AN ECONOMICAL POINT OF VIEW.
Therefore we must operate either in form to raise the general W probability or, more likely, to restrict at most the W appearance within the shortest periods of betting.
(Correct IMO. However, most players fall prey to what so many attempt to sell or teach on the internet and over simplify the game with almost no teaching or detailing out the psych and decision making process so vague to so many. As well, 'false positives', meaning, the bad effect winning has on even intelligent players causing them to forget everything and only focus on the larger wins that will make up for all their losses, etc.) (Shorter and quicker is better, but has to be with meaning and substance with complete focus and extreme frame of mind regarding your risk capital, buy in and the purpose of that buy in)
Step #2
Math gurus and fkn bac pseudo experts abounding on internet will teach us that every bac bet is EV-
This is a total bullsh.it statement.
It could be true whether any resolved hand is totally independent from the previous one and, of course, if the game isn't finite (that is determined by a finite deck).
For example, we know that sooner or later a 4 Player streak formed by no asymmetrical hands is a perfect EV=0 disposition (actually it could have a EV+ due to card distribution). We know that such P streak isn't the product of 0.4932 x 0.4932 x 0.4932 x 0.4932 probability.
The same is about a Banker 4 streak not forming one or more symmetrical hands in between.
In this case the probability is just 0.5 x 0.5 x 0.5 x 0.5.
Various streaks at either side will come out at different times but itlr the number of different mix of as/s hands tend to be correspondent to the expected values.
(There are times when a player can clearly see the negativity as well as the positive sections within the game if he focuses the correct way with complete neutrality, which is extremely difficult to do on a continual playing basis. Very hard. But possible. Over confident and counting those chickens before they are hatched, is the worst thing you can ever do at baccarat.)
Step #3
The most important thing in order to take advantage of what I've written so far is restricting at most our probability of success per each series of bets we wish to place.
In a given infinite succession of pseudo coin flips, trying to win one unit in two consecutive attempts by a kind of progressive wagering must invert to our favor the general 75% probability.
At baccarat this 75% W probability may come out or not, it's our duty to select the spots where such probability could be higher or lower.
Thus it's not about how we'll raise our bets.
Why 2 consecutive wagers are better than 2+ wagers or a succession of single bets followed by a kind of progression on next bets?
Easy answer.
Variance acts by steps. Our bets must be winning on the very first bet or the second one. The rest will sink into the uncontrollable variance ocean. And such ocean may present several consecutive deep holes we can easily fall into.
Restarting to bet will be covered later about step #5.
(Everyone eventually develops a different playing style and belief of the game. Yours will win and mine will lose, The other guys will always lose. Yet another player might always win for a period of time using yet a 4th one. Nothing can hold up on a continual basis. Better than anything that even will score a 60% favorable chance to win, would be complete neutrality of the player and catch that 5% or 10% or 15% of the times that seem impossible to predict, guess ot get on the winning side. I hope you understood that statement. IMO, most of us tend to want to be in the over 50% winning section and avoid the losing wagers. How about, attempting to see those sections that wipe everyone out or 'cut', etc.? )
Step #4
416 cards can arrange in numerous ways, yet it's more likely that what happened first won't be perfectly balanced by the next outcomes of the same shoe as the previous card distribution must affect subsequent card dispositions (not results).
(There are many answers in the details I outlined about 'Sections and Turning' Points, etc. That will always apply within every shoe.)
Step #5
Besides strict statistical issues, everything comes in handy to ascertain what's the best course to take: if a given negative pattern had come out, odds are it will represent again on the same shoe.
Predominance, cards issues, overall other players' outcomes, turning points, Alrelax wrote a lot of interesting and valuable topics about this.
A final note, a kind of gift: study the shoes where the very first pattern is a 3+ streak on any side (especially on B side). You'll find valuable spots to bet into the subsequent hands by a 99.9% accuracy.
(There is a very unique pattern/trend that usually does present itself after the first strong repeating section, no matter if that was B or P, and as well, no matter if it was 3 or 4 or 5 or 6 repeaters in a row after say 10 or 15 hands of 1s, 2s and 3s or any combination of those, such as 1s and 3s, then a 5 or 6 or 7 streak or more and than, at that point thereafter for a whole section is usually the hottest and best an experienced player can ever look at, at least the majority of times, all IMO and experience.)