If you’re writing LitRPG, one of the things you have to figure out is why this video games even matters. Other genres don’t have this problem, because they take place in some kind of “real world” (whether or not it bears any resemblance to the actual real world) and the stakes are pretty much automatic so long as anything interesting is happening. The villagers are people, not NPCs, the dark lord is an actual villain and not just a very elaborate toy soldier, and so on.
So what premises can solve the problem of stakes?
THERE ARE NO STAKES: There’s no fundamental stakes at all. Character arcs revolve around regular, real world problems, and the friendships and rivalries that develop over the course of the game. The major problem for this is that the game becomes a backdrop, and actually beating the game becomes pretty unimportant. Indeed, the most obvious character arc would be for the protagonist to initially be over-invested in beating the game because the rest of his life is a shambles on account of some terrible inciting incident (lost his job, girlfriend broke up with him, girlfriend broke up with him because he lost his job, that sort of thing), and that in the end he forfeits victory because the real treasure was the friends he made along the way. But, y’know. People read LitRPG to vicariously experience video games that don’t exist (or which they aren’t good at), so they’ll want a plot where actually beating the game is important.
Kayaba Strikes Again: The Sword Art Online solution. Someone has, intentionally or accidentally, prevented people from logging off, and also optionally causes anyone who dies in the game to die in real life. The only way to get out is to beat the game. The major drawback here is that it’s almost impossible to really justify. Sword Art Online gave the infamously unsatisfying answer of “even the guy responsible doesn’t know why he did it,” and Sword Art Online Abridged gave the answer that it was a series of dumb mistakes made due to sleep deprivation from launch crunch, which is funny, but only really works as a parody.
There’s also the problem that it makes little sense for players to oppose each other. Sure, a lot of people who wanted to play dwarf-aligned might grumble when the all-players omni-clan picks elves, but unlike in regular gameplay, no significant number of them will pick dwarves anyway and actually murder other players over it. Not enough of them to serve as an ongoing problem and not an early arc that gets resolved by the establishment of a single mega-clan, at least.
There’s another problem, although one that doesn’t matter as much to most LitRPG stories, which is that it makes no sense to have new players still coming in after the plot begins. On the other hand, if you can still connect to the server, just can’t log out, you could have an interesting sub-plot that everyone who joins after the death trap goes public is someone intentionally entering the murder server either to take advantage of the situation or as a rescue mission. If you see someone has a join date after the day the murder game was revealed, you know they came here voluntarily, which instantly makes them dangerous. Unlike everyone else, they came here fully expecting that they might get murdered, and probably expecting to do some murders themselves.
Isekai: Protagonists aren’t playing virtual reality. They have discovered a portal, been abducted, or reincarnated to a world that runs on video game logic. Their lives are on the line because they actually live here, and NPCs matter because they’re actually alive. “Beating the game” is desirable because “the game” is just an actual state of war or tyranny in the world. The main issue here is that part of the gameplay experience is that NPCs don’t matter. They’re expendable in a way that players are not, and that works only because players alone are real people rather than polyps of an AI. A lot of LitRPG like to ignore this and make a point of how their protagonists care about NPCs, which definitely works in isekai in a way that it doesn’t in other setups. For some reason, most stories that use the “hero is nice to NPCs” conceit don’t actually use the isekai premise, so the NPCs actually are expendable polyps of an AI, and the story just hopes we don’t notice this.
Willy Wonka’s Video Game Factory: The first person/group to beat the game, perhaps within some specific constraints, wins some fabulous prize. In extreme examples, ownership over the game’s company as part of the eccentric CEO’s search for a successor. In more realistic examples, a million bucks set aside by marketing to fuel interest. It’s easy enough to attach a Charlie Bucket protagonist who badly needs the money, and rich antagonists who don’t even need the money, competing just because they want to win.
We Live Here: The Threadbare option, wherein protagonists just live in a world that runs on game logic. There may or may not be any explanation as to why. LitRPG is still in the phase that fantasy went through just about a hundred years ago, where it’s still expected that there be some kind of tie to the real world, some sort of explanation as to why this universe runs on video game logic, and only a handful of stories are just having the world run on video game logic without any further explanation besides “that’s the premise.”
One irritating issue is that there’s one premise here that’s been driven into the ground (Isekai), and two other premises that are significantly more compelling (Kayaba Strikes Again and Willy Wonka’s Video Game Factory) but strongly associated with specific works (Sword Art Online and Ready Player One, respectively). There’s an immediate risk of standing in the shadow of an existing, well-known story if you use either one of them. I think this risk looms much larger in the minds of writers than readers. Writers, especially amateur writers, tend to value novelty far higher than audiences do. At the same time, Sword Art Online and Ready Player One had their respective moments, and those moments are over. Is there still an audience out there for whom “that sounds kinda like Sword Art Online” is a selling point rather than a condemnation?