Only until they learn to take proactive measures before anyone gets dropped to 0 to prevent that occurrence: avoiding combat, using healing abilities while still well above 0, changing tactics to favour ranged attacks and eschew melee when possible, hiring more muscle (e.g. henches etc.) for the front line, etc. etc.; and most of all being willing to bail out when things go sideways.
Just focusing on
hiring more muscle - what makes that better
game play than an expectation that the PCs can succeed in their endeavours via their own abilities?
I mean, Conan doesn't generally hire more muscle - and even if he starts out with friends or a warband, by the end of the story he's generally fighting solo.
The Fellowship of the Ring doens't hire more muscle - and those points when the protagonists become warband leaders (eg when Aragorn enters the Paths of the Dead; or when Merry and Pippin take command of the resistance in The Shire) are all about the development of character and theme, not a study in tactics and expedient generalship.
I'm not saying that a game which foregrounds "hiring more muscle" is inherently bad. But nor is it inherently good. It does suggest a genre quite different from what seems to be the norm for contemporary D&D play, and for at least much of the inspirational fiction.
For some reason this seems to be the hardest lesson for some players
Mostly because it's not fun for some players. Same with avoiding combat or relying on ranged weapons.
Some people paly this fantasy roleplaying game for the fantasy.
This. Off the top of my head I can't think of a Conan story where he runs away. Ged does run away from the Gebbeth in A Wizard of Earthsea, but not because he is losing a fight with it.
There is nothing per se
better about a fiction that involves fleeing, than one that doesn't. It depends on what sort of fiction we want.
There are other concerns with running away. Turn order.
In order to run away you need to get away from your opponents, and as JG says above, that might not be easy to do with the way movement works. To make matters worse, initiative also becomes a problem. It's very easy for someone to be caught in a bad spot because of initiative and turn order.
The dynamics of tactical combat discourage running away because running away itself becomes a tactical challenge. Can you afford to spend a turn not attacking your opponents if you can't guarantee that your enemies will not start attacking your teammates?
There's also the fact that unless the DM specifically ends combat and either A) invokes the chase rules or B) decides the bad guys have better things to do*, enemies are just as fast if not faster than PC's and are perfectly capable of pursuing them and continuing the assault.
This too. How is the action declaration "We flee!" going to be resolved. The resolution in 13th Age is clear (the PCs escape, but suffer a "story" setback as decided by the GM). The resolution in Torchbearer 2e is clear (the PCs initiate a Flee/Pursue conflict, which gets resolved using the conflict rules). The resolution is relatively clear in classic D&D (the evasion/pursuit rules are activated).
In Torchbearer and class D&D the resolution of the plural declaration "we flee" is facilitated by the fact that there is generally no "turn order" across the PCs.
A system which separates the PCs in turn order and hence action economy, which resolves fleeing simply by movement of squares on a board, etc - ie that has the general properties of contemporary D&D - is already not well-positioned to support the resolution of a group decision to flee. 13th Age gets around this by using a "meta"-style of resolution for the declaration. I'm sure other technical solutions are possible too. But it's hopeless to approach this as if it's some sort of moral or intellectual or emotional failing on the part of players, as opposed to a question about both the technical and the aesthetic aspects of game design.