When such thing happens we should think which patterns are going to balance those losses.
Probability says that more often than not losing streaks are not promptly followed by winning streaks.
If we have lost 6 hands in a row we know we have just one pattern out of 64 to fully recoup and forcing the probabilities by the use of a progression makes things worse.
At baccarat and besides the old B less disadvantage, there's no way we could expect to get more +1 than -1, everything will balance out by 1 trillion accuracy.
Of course this happens whether we bet the same amount as no one losing bac player in the world will lose just the vig. He/she loses a lot more.
Since any single shoe is a finite card dependent proposition, what we can do is trying to get rid of those shoes who appear not to fit our plan and instead to try to get a kind of advantage over those shoes featuring the patterns we're looking for.
Up to a cutoff level on both W and L scenarios.
If we consider multiple complex situations we'll see that the "balancement world" won't happen per every single shoe as the number of hands is finite and card dependent.
Examples are when a P7 two-card point will be continuously busted by a B natural or a B 8-9 third card drawing hand or when an asymmetrical hand keep giving a wonderful third card to the player.
That doesn't mean we should start to bet respectively Banker or Player, just that our more likely plan will be disrespected in that given shoe.
Imo it's not the "what" but the "how" that matters.