It's the discussion of the summer: games are too big, both in the sense that producing them costs obscene amounts of money and in the sense that they outstay their welcome, leaving players feeling fatigued with a treadmill of repetitive content.
There are reasons for this. Reasons that are not excuses but that nevertheless exert a powerful force on the many thousands of talented, creative developers currently making video games and on their visions, which often do end up compromised in myriad ways by the material conditions under which their creators toil. I won't get into them here. No one is truly interested in how the sausage is made after all; they're interested in how it tastes.
I do want to talk about a key design element that often gets overlooked when discussing open world games, and why it might make the difference between an open world game that feels too big, and an open world game whose vastness serves it: the story engine.
A story engine is the setup that allows your game to generate new stories. It's the reason why there are so many cop shows on TV. Detectives always have a reason to solve a new crime every week; it's their job. It's also the reason why so many people suspect Murder, She Wrote's Jessica Fletcher of being a secret serial killer; that one woman would end up in the right place at the right time to investigate so many murders strains credibility.
A good story engine generates new stories intuitively, without the need to establish afresh certain expositional needs: why a character is invested, why they're the only one who can do a certain thing and not someone else.
The problem with many open world games is that they are overly focused on telling one story and not focused enough on creating the engine that would allow the player to stitch together different stories within their worlds.
This distinction is often the difference between a linear game that takes place in a wide open space and a game that I would describe as a true open world: one in which the player can truly pick a direction and go in it and not find that the interesting things to do are all locked away behind quest progression.
Stories are at their best when they are sequential, even if they are sometimes presented out of sequence, there is usually a linear ordering to events. Open worlds are just the opposite and are therefore better suited to a more modular mode of storytelling--that is to say, for your open world narrative you need not one big story but dozens of smaller stories which can be encountered in any order. And you need a story engine that can explain why the player would, at any given moment, get involved in any or all of them.
The other week I read an opinion piece on a major games website (I think it was PC Gamer) regarding Starfield. The writer concluded that Starfield might just be a good open world game as long as they can build enough compelling content to fill it. As someone who creates quests for video games for a living, this struck me as a mad proposal. Like digging through a brick wall with a toothpick, I can't imagine how Bethesda will fill Starfield's thousand planets with enough bespoke content to keep the game interesting. I don't even understand why they should.
I have played Assassin's Creed Origins, Odyssey, and Valhalla. I have played Horizon Zero Dawn and Forbidden West. I have played Fallout 4 and Skyrim, and countless other open world games. The problem with these games was certainly not a lack of bespoke content. In all but one of them (the comparatively modestly sized Horizeon Zero Dawn), I stopped playing the game long before I'd run out of handcrafted side quests to complete. I only completed the entire main storyline of three of them. There was plenty of story left to discover.
The problem with all of these games, saving maybe Skyrim, is that they give you an inert open world and no real reason to explore it. Apart from farming materials for crafting (yawn), the entire purpose of the game is funelled into linear quests which fight against the promised freedom of the open world. The problem is that my role and motivations as a character disappear the moment I don't have instructions from an NPC to fulfill.
Who is Aloy without the quest givers that tell her what to do to save the world? If I'm a parent desperate to save their child in Fallout 4, why am I building settlements? What connection does one action have to the other? I cannot be any kind of viking in Assassin's Creed Valhalla, I have to be Eivor, who for whatever reason is dead set on uniting England. Taking instructions from NPCs grows tiring after a while. They obviously care about these tasks, but do I?
Skyrim, at least, gave me freedom to decide what kind of character I wanted to be. The mix of the procedural (oh, no, a dragon!) and the bespoke (what is this intrigue I've stumbled into in this town?) complemented each other. I stopped the game when I achieved my character's main goal: a set of armor forged from the hearts of daedra. Saving the world from dragons might have been some other character's goal, but it wasn't mine, and the game was, for the most part, not too pushy about it. Morrowind was even better about this; the world was weird, it was yours to discover, and you didn't have to be the chosen one at all.
What is the difference between these games and the others? Assassin's Creed, Horizon, Ghost of Tsushima, Red Dead Redemption 2, and to a lesser extent, Fallout 4 are all story-driven open worlds. Though they are filled with side quests and distractions, they all exist to drive you through a crafted story about a character with defined, personal motivations that override and overshadow any side role you might fulfill. You cannot choose to be anyone else in these worlds, and you will quickly run up against the limitations of what you can do and how deep you can go if you depart from the story. Take the most simulation-focused of these examples, Red Dead Redemption 2: you can't even get your fishing rod until you've progressed to a certain quest in the storyline.
Morrowind, and to a great extent Skyrim, are story engine driven worlds. The narrative in them exists as a pretense for letting the player loose to engage in the world how they see fit. The moment-to-moment of playing these games is filled with exploration, wonder, and discovery precisely because they invite the player to find their place as a stranger in a strange world. Maybe they are drawn to the magic and politicking of the mages' guild; maybe they feel a pull toward Daedric quests; maybe they just want to explore, experiment with the game's systems and forge their own path. There's no sense that any way is more valid than any other, and the player's defined motivations (or lack of them) enable this free-form play.
Yes, if you want to defeat the big bad, you have to pursue the main storyline quests, but in Skyrim you don't even have to learn Dragon Shouts if you don't want to. You can shrug off your destiny and go do what you want and there's still plenty of game there to engage with.
That's not to say that the main quest needs to be optional for a story engine driven game to work, but it does require that the open world have a lot more focus and generally be a lot smaller. Think Spider-Man, an open world game that is refreshingly small and light on distractions, lazer-focused on realizing the fantasy of playing the webslinging superhero; or Breath of the Wild, a game in which everything you do is forwarding Link's ultimate goal of defeating Ganon. There's barely a pre-scripted quest in sight, just a dangerous world to explore and a set of interesting tools with which to do it.
The promise of an open world game is its size and the freedom granted within it. It is the feeling that there's always more to do, more to see, and that therefore my experience is going to be different from anyone else's experience. It's a world where you can wander off in any direction and find something interesting to do, a mystery to unravel, a secret that intrigues you. Pre-scripted content has an important role to play in shaping that open world experience; memorable scripted sequences can connect seamlessly to player-shaped narratives, as when a drinking contest in one town causes you to black out and wake up halfway across the map, altering your plans for how you intended to explore and thrusting you into a new, unfamiliar place.
In a great open world, setting off in any direction should not feel like you're rebelling against the game's intentions, but rather playing in the space it's created for you to play in. A good story engine gives the player an idea of what that space is: it's a combination of player fantasy, rules of the world, and premise.
Starfield will succeed not by having more bespoke quests than any other open world, but by providing a story engine that encourages the player to explore its vast galaxy and define their role within it, and having bespoke content that helps reinforce the player's chosen role, and occasionally forces them to change how they play. The player is the driver; the story is merely the engine.