Some people are going to find what I'm talking about in this post so obvious they won't understand why I feel the need to mention it. To those people, I offer my apologies, but I have realised from discussion with a broad spectrum of roleplayers over the years that it is not so obvious as it seems.
The core activity of roleplaying games is making decisions. Every other activity feeds back into this core in one way or another. Rules are there to control and shape the field of possible decisions, determine the likelihood of particular outcomes of those decisions, and to structure who makes which decision. Characterisation, party dynamics and the world exist to shape the style of decision-making and the values and tools brought to it; to establish decision procedures; and to provide stakes and consequences for decisions.
I would contrast this understanding with the one that superficially claims that a roleplaying game is like a television show, novel or movie that is focused on telling a captivating narrative for an audience. This kind of understanding, taken to its extreme limit, leads to railroading, which I understand as the negation of player characters' ability to make meaningful decisions, combined with the players' knowledge that this is the case.
I think a decision-based understanding of gameplay is superior to this narrative understanding for two reasons. The first reason is paedogogical, and the second is schematic.
When explaining to a new player or potential player what the game is and how it is played, people often repeat some line like "You're like the star of a TV show going on an adventure" or "You're a character in a fictional world who lives out a story". This implies that the most important activity or skill is acting, and the structure of play will follow the scene-framing and development of other narratives. It is therefore, a rude shock when they start playing a game and they are not very good at acting out their character and the flow of a session does not follow the narrative progression of other media.
This shock is now sometimes called the "Matt Mercer effect" after players who got into D&D from watching Critical Role, but who are displeased to find out that they can't plunge immediately into a complicated and immersive narrative. I don't blame these people for this shock. They were led to believe one thing, and rudely surprised when it turns out that belief was just propaganda. On top of that, they don't necessarily know how to improve the situation they find themselves in and move to the sort of immersive game they were lured in with - witness the endless threads on Reddit and rpg.net and other forums on this very topic.
Paedogogically, I think focusing on decision-making instead is probably more useful for that new player. The new player begins by making and justifying decisions with the character as a pawn (on the character's behalf), then learns how to make decisions based on the character's subjectivity (as the character). Similarly, they learn to justify their decisions to the referee or other PCs on behalf of their character (or out of character), and then how to justify them as the character to the other character in world.
They're not rigidly locked into any one of these four options, of course, and further learning would be how and when to deploy each. This helps new players to feel like they're increasing their agency over time, rather than constantly failing. In my experience, this is both an easier course of development for new players to grasp, as well as being more likely to retain them as players than frustration and failure are.
Schematically, I've written about this before in my long post on "anti-narrativism" in constructing D&D stories. A distillation of the basic point is: Rather than structuring the progression of a campaign as a series of dramatic scenes with the PCs flowing from one to the other, I think games should be structured as a set of situations where PCs must make decisions. The occasional montage or descriptive flourish isn't awful, but thinking of the development of the story in terms of decisions actually provides a clearer and more effective plan than trying to think of the next "chapter".
As a bonus, the superficial sensual effect of a decision-based structure is that it actually seems more like a traditional narrative than trying to plot out a story as if it were a novel, television show, or movie, because making decisions and dealing with the consequences of them is what happens to characters in a well-written story.
I think many people will claim that they already accommodate the idea of decision-based adventure gaming in their planning, and I'm sure many do, but I am 100% certain that there are many more people who could improve their games by shifting their mindset over to this, or increasing its prominence and salience within the mix of ideas they have about planning. I especially encourage new players and new referees who are having trouble creating the kind of richly immersive world and story they admire in other people's games to try adopting this mindset.