I wanted to love Starfield

Or, how to break your game loop

I really wanted to love Starfield, but in the end I only mostly liked it. In general, there were a lot of places where the idea was solid but the execution was lacking, and a lot of minor flaws that detracted from the whole experience. With that said, there’s still a lot that was done right, I still enjoyed my run, and I’d encourage anyone who’s even remotely interested in the game to do a blind run through the main quest because it really is best experienced without spoilers.

I could talk at length about Starfield, its ups and downs, its triumphs and flaws, but this is (ostensibly) a game dev blog so I’m going to focus in on what I think is the biggest problem with Starfield and why that is.

The Bethesda RPG game loop

How do you define a Bethesda RPG?

One way is by its game loop of explore->fight->loot->repeat. You head out into an open world towards some destination, whether it’s for a main quest or a side quest or just a neat looking tower in the distance. You may or may not make it there; you might come across an inn or a town or an abandoned mine and decide to check that out on the way. You run across some bandits, some mutants, some monsters, fight them, and then loot their corpses and the immediate area for anything valuable. Then, probably close to your carry weight limit, you return to somewhere you call home, dump your loot and get ready to repeat with your next destination in mind. All this while, you’re engaging with the narrative of the world and gaining experience points and better gear to make your character more powerful.

This structure is common to every Bethesda RPG across The Elder Scrolls, Fallout and Starfield at least back to Morrowind. It’s not exclusive to Bethesda’s RPGs, of course, in fact it’s arguably an RPG staple present in every game with an open world and loot. Nor is it the only defining aspect of Bethesda’s RPGs; you could easily point to the level of interactivity, the world design, or the conversation structure as defining traits. However, this particular structure is clearly present in every contemporary Bethesda RPG, and if nothing else it provides a useful lens to examine them through.

The first Bethesda RPG I played was Morrowind, and I’ve played every Bethesda RPG since- though not a ton of Morrowind or Oblivion. I suspect you can trace the evolution of these ideas even further back to Arena and Daggerfall, but I’m not that familiar with them and I don’t want to make that unfounded claim. I haven’t played Fallout 76, I have played Elder Scrolls Online, I feel both are different enough in intent to not be directly in the line of evolution so I won’t consider those here either.

Bethesda has evolved every part of the game loop from game to game, but of particular note is the evolution of the repeat step, what you do at the end of each cycle. You’ve always been able to return to a city and sell your loot, but the idea of having a home base to call your own has grown from Morrowind to Starfield. Player homes got more numerous, more accessible, and more functional with each game. Crafting was dramatically expanded in Skyrim, and its Hearthfire DLC toyed with the idea of buildable player homes. Fallout 4 brought in fully customizable settlements with management aspects and its DLCs introduced automated production.

These additions work with the game loop, albeit with the caveat that the player had to be interested in them, putting in the time and skill points to engage with mechanics. Crafting and settlements add a reason to go out in search of resources and something to use that loot for either in terms of making your character more powerful or just for fun, keeping the loop going. Starfield adds customizable ships that are player homes, fast travel methods, and foci of combat gameplay all in one. Unfortunately, I don’t feel they work well with the loop or in general, but the intent is there, and I’ll dig into why things don’t work this time.

A (non-contiguous) new world

There’s no point sugarcoating it or beating around the bush: Starfield completely breaks that game loop because of how its world is put together. Or, perhaps more accurately, because of how it isn’t.

Every previous Bethesda RPG* has had one contiguous, traversable worldspace. While there are exceptions for DLC and certain quests, you can generally walk from one end of the world to the other. It’s a big world full of buildings, enemies, and interactions that you can see and walk up to. Some have been sparser than others, but in general in every game if you move in a direction for a few minutes you’ll encounter something handcrafted and at least mildly interesting.

Starfield is built very differently. The world isn’t contiguous; it’s broken up into somewhere between a few hundred and over a thousand planets depending on who you ask and what you count. Travelling between planets is accomplished through what’s effectively a diegetic fast travel system. It’s theoretically possible to fly between planets within a system manually, but the timescale involved is impractical and you can’t travel between systems that way. In effect, exploration becomes clicking through screens, and it becomes much more deliberate because there’s no in between anymore. While there’s a bit of room to explore on a given planet, you’re never going to see most of the universe at all unless you deliberately set out to explore it.

This is new territory for a Bethesda RPG outside of DLCs, but it’s not unknown for other RPGs that have made this work. The Outer Worlds is built this way, every Mass Effect game is built this way. Where Starfield distinguishes itself is in its sheer scale. Not only are there orders of magnitude more planets to explore, but the explorable area of each one is also far, far larger.

There’s a catch, of course, and that’s that Starfield leans heavily on procedural generation. Of those hundreds of planets, there are maybe a dozen or two with specific handcrafted content, and only in certain areas. The rest of the game is vast expanses of flora, fauna, and resources with a handful of points of interest within walking distance. Those points of interest might be ruins, or a building, or a ship landing, but they’re effectively built from templates and start to feel the same very quickly.

Frankly, I don’t think the scale Bethesda bragged about is a good thing at all. To echo an increasingly common sentiment, Starfield is a mile wide but an inch deep. I just don’t see the point in clicking through random planets and bunny-hopping across the barren surface over and over again to experience drab, repetitive content. This is just my opinion, though.

I’ve heard a lot of comparisons to No Man’s Sky, a game which I have never personally played. I was immediately reminded of the first Mass Effect. Much as I love that game, I was not a fan of the barren planets with one copy/pasted building then or now. At least you had a rover to drive around in, which honestly I wish Starfield had too. Incidentally or perhaps not so incidentally, Starfield is very reminiscent of the original concept for Mass Effect Andromeda that we never got. A concept that Bioware dismissed as not only technically infeasible, but also fundamentally unenjoyable.

Fundamentally, it’s the breaking up of the worldspace that breaks the explore part of the game loop, because the organic discovery of new locations and new encounters isn’t there anymore. If you want to look for something new, you have to specifically look for something new. The heavy use of procedural generation isn’t itself a deal-breaker, but it puts a damper on exploration because of the limited amount of meaningful variety.

*In the sense discussed in the previous section, so ignoring the very early Arena, oddities like Battlespire, and Online.

Minor changes, for the worse

We’ve established that the explore part of the loop is broken. There are two views you could take here. One is that with that broken, the whole loop falls apart. The other is that the rest of the loop still works well enough to drive through the broken explore step. There’s still a meta around gathering resources, building up your character, building settlements, building wealth.

Unfortunately, the repeat part of the loop, where you take your loot and turn it into something that represents useful progression, is also worse. It’s very similar to Fallout 4 to the point where it almost feels like a mod at times, but it’s tweaked in ways that make it more tedious and less rewarding.

I could go on about the changes at length, but I don’t want to dwell on it too much. One issue is that mass limits are smaller and both your ship and settlement containers have them, meaning you have to be more judicious about scavenging, but there are no new tools to track what you still need, only the option to mark resources required for a project as needed (crucially, not how many). Another is that crafting has another step- research- between taking a crafting perk and being able to craft mods. There are fewer mods in general, and items come in various grades you can’t upgrade between yourself, so crafting feels less rewarding. Setting up settlements requires hauling in significant quantities of resources, and while it’s possible to send resources between settlements they can’t share a crafting inventory with the simple assignment of a provisioner like in Fallout 4.

None of these are game breaking, although the changes to settlements probably could warrant an article of their own. Unfortunately, with a broken explore step and combat that’s about as satisfying as its immediate predecessor but not significantly better, Starfield really needed a strong repeat step to drive the loop and it just doesn’t have that. The game pushes you to engage with the crafting system and the outpost mechanics, but I didn’t find them rewarding enough, and I just wasn’t interested in micromanaging. I wanted to roleplay and shoot things, not play a management game with hostile UX. I think this is going to be subjective to an extent, but I really feel if Bethesda wants to lean into micromanagement, they need to build the game for it, and if there’s more complexity, there need to be more tools to manage that complexity.

Trying to fix it

There are ways to be directed toward new quests and new locations, of course.

There are a few systems where you receive a radio transmission (conversation) when you jump in. This is how I was given one of my favourite sidequests in the game, Groundpounder, which had some great environment, narrative, and encounter design. It’s also how I was given my least favourite sidequest in the game (First Contact).

The thing is, it’s a band-aid fix. It still requires you to jump into the system for one reason or another in the first place, and it only really works once per system. I’ve seen a few variations on the theme, like having a quest notification pop up or a distinctive icon implying a point of interest, but it’s the same thing: a workaround that helps but is ultimately just that, a workaround.

A lot of the scripted, handbuilt content in Starfield is in the faction quests. Putting a heavier focus on major quests and relying less on organic discovery, even though it seems to run counter to the implied focus on exploration, is a good idea given all the issues mentioned above.

If you’re sensing a “but”…

The problem, at least for me, is that there isn’t any narrative reason to engage with the faction quests. The more I think about it the more I’m baffled by this one, because historically it’s something Bethesda has gotten right despite their other narrative foibles. Skyrim’s civil war was visible everywhere, and you were forced to (temporarily) pick a side at the very beginning of the game. Fallout 4 had you engage with every major faction during the main quest, and you had to pick one to finish the game with. In Starfield, the factions are just kinda there, you’re pointed toward the initial quest givers but never given a compelling reason (at least, not compelling to me or my character) to start them. It seems that Bethesda understood on some level that players couldn’t or wouldn’t be able to discover content organically in the same way and tried to work around it. However, at least for me it just wasn’t as fun an experience as the exploration and discovery in previous games.

Point and counterpoint

I’ll be the first to admit that there’s a subjective element here. While it’s a fact that the game world of Starfield is fundamentally built differently than previous Bethesda RPGs and that significantly changes how the explore step works, whether that’s broken or merely different is subjective. While it’s a fact that there have been changes to crafting and settlements, whether that’s worse or better or merely different is subjective. I think I’ve made pretty strong arguments here, and even if you disagree with me (which is totally fine!) I think I’ve laid my reasoning out well enough that you can understand where I’m coming from.

As I said earlier, despite its flaws I enjoyed my time with Starfield. And though I feel the changes from previous games are a negative overall, some might not mind them, or even prefer them. The breadth of the world is impressive, and though I got bored of what I felt was repetitive content quickly, some might enjoy wandering through this vast universe. Some might find the combat and the progression compelling enough to drive the loop. Some might enjoy the complexity of the crafting and outpost settlements and not mind the lack of tools, finding the need to figure out their own way to keep track of things a plus rather than a minus.

I would still argue that Starfield is flawed in design. It’s possible to go hours without interacting with the handcrafted, narrative parts of the game, and the main quest, critical path story content falls by the wayside. Rather than synergizing as they have in previous games, the open exploration and closed narrative portions of the game are separate, even in opposition to each other. This is very much into opinion, though, and touches on some broader conversations about narrative and roleplaying in role playing games, artistic intent versus played experience, and cost/benefit tradeoffs in development which are all well beyond the scope of this piece.

Synergy and discord

As I wrote this piece, I struggled to figure out what note to end it on. I think I’ve made a strong argument that Bethesda broke their game loop in Starfield- and if you disagree, repeating that the in the conclusion probably isn’t going to convince you- but what’s the takeaway here? What can we learn from this? This is meant to be a piece about game design, not a critical review of Starfield.

My first thought was something along the lines of seemingly unrelated design decisions can have major effects, but that felt like stating the obvious. The next was something about answering the biggest questions first and then building the game around those instead of twisting an existing formula blindly, but everything I’ve said so far has talked about the results of the design, not the design process.

The best I’ve been able to come up with is along similar lines, but distinct. Making what seem like small changes to an existing formula can fundamentally alter the way the game works.

From one view, Starfield isn’t fundamentally different from its predecessors. If you sanded the logos off, most players would still recognize it as something from Bethesda Game Studios. If you look at small pieces of it, it looks and feels a lot like its predecessors. It can be described evolutionarily, as a series of small changes. Start with Fallout 4. Add space travel. Add procedural generation. Tweak progression. Tweak crafting. Add production management to settlements. You have Starfield.

From another view, though, Starfield is fundamentally different than its predecessors. The big, densely packed traversable world has been a staple of Bethesda RPGs since Morrowind, and it’s not here. Exploration takes a much different form with a lot more clicking through menus, the game loop doesn’t drive organic discovery the same way, and the game just doesn’t flow the same way. In every previous game you can walk from one end of the world to the other, stopping for tea, a fistfight, and storytime along the way without ever leaving the viewpoint of your character. You can’t do that in Starfield.

Games are complex systems of systems that interact with each other in messy ways. Even when we’re not explicitly acknowledging it, that reality is baked into the language we use to talk about games. How often have you heard phrasing like “everything meshes so well” or “this addition really makes it come together”? A game consists of parts, and depending on how well those parts work together, it may become greater or lesser than the sum of those parts.

Starfield is an incredibly ambitious game. A galaxy with thousands of worlds to explore, blending handcrafted and procedurally generated environments, is an incredibly achievement. It’s a selling point for good reason, and for some it’ll absolutely make the game. But everything has tradeoffs, and for me, it wasn’t worth the cost.