thicc_girls_are_teh_best
Gold Member
So in the wake of the Starfield Direct, we've seen a good number of journalists and reports, not to mention fans, big-up the game and shower it with glowing praise, calling it a magnum opus, potential "Game of the Generation", and the most ambitious game to release yet for the new consoles.
Now, personally, I liked the Starfield Direct. I wish they had actual uninterrupted gameplay showing a whole gameplay loop, but I did seem some improvements over the 2022 footage namely in certain character outfit textures and framerate stability/framepacing. Some of the expanded looks into features like shipbuilding were awesome, showing off a lot of variety in how you can build a ship and how that build affects its performance as well as interior makeup. The space battles looked fun, and some of the planets with alien life on them look like they would be fun to explore and observe in a jungle safari kind of way.
But I didn't see anything that made me think "Oh wow! This is easily the most complex and ambitious game ever made!", and I saw nothing in terms of systems or complexity, nor in terms of visuals, that made me think the game NEEDED to be locked to only a 30 FPS Quality Mode option.
Let me clarify what I mean when I say complexity: it is NOT in terms of content scope AKA total amount of explorable content. And the reason why is simple: that's the easiest way to try defining game complexity and has been that way since the 1990s. If you make a 2D Mario game with 5,000 levels, is that suddenly more "complex" and ambitious than SMB3 simply because it has a crapton of more levels, but otherwise the actual game mechanics and level of simultaneously engaged systems remain the same? No! It just means you have a larger SMB3 with 5,000 levels. You aren't playing all 5,000 levels at once, your actions in one level aren't cascading in real-time onto the other 4,999 levels (or even maybe 3 other levels). You just have a larger playground to play in.
This is basically how I see Starfield. It is a larger Fallout set in space, with a lot more "levels", and maybe a few extra systems (such as the aforementioned shipbuilding), but otherwise has the same level and degree of simultaneously engaged systems & mechanics as previous titles like Fallout 4 or Skyrim. Your "playground" is much larger but your bucket of toys to play with in that playground is more or less the same in number of toys, and their complexity.
Now this isn't me saying that there aren't things of higher complexity in Starfield vs. Fallout 4 or Skyrim. They have a gravity system in place for your ship that runs concurrent to the gravity of outdoor areas, for example. There are more particle effects from gunfire here vs. say Fallout 4, meaning more particle calculations, etc. But when we get to other things such as number of onscreen enemies, scale of onscreen enemies, number of onscreen NPCs, NPC idle & walking/item interaction routines, animation systems, object physics, atmospheric weather systems etc., there is virtually nothing in the Starfield footage that looks leaps and bounds over other high-budget open-world AAA games already on the market. Because in some of those games, there is at least comparable or in some cases, more concurrently happening, in terms of the immediate locality and space of the player, on-screen both in terms of visibly observant and background systems.
Part of my conclusion here could be based on the fact the Starfield footage was not uninterrupted gameplay, but rather snippets of certain segments of the game with developer staff acting as transitions to them. Indeed if that is the case, then the blame wouldn't fall on me, rather on the way Bethesda chose to show off the gameplay with segmented cuts rather than as an uninterrupted gameplay sequence through a story mission or so. Maybe they couldn't show off a mission with all the mechanics revealed actually being a part of said mission, but that would also signal (to me) that the typical gameplay loop is probably less reflective of the impression of a gameplay loop the Starfield Direct presented. Regardless, it doesn't change what I've observed.
And let me also clarify: this is NOT the same thing as me saying Starfield is not complex or not ambitious. Several people, when I've mentioned this opinion in previous topics, interpreted it that way when I in fact said no such thing. Starfield is clearly an ambitious game, large in scope and with a great deal of complexity. However, I feel it is fairly easy to challenge the notion that it's "magnitudes" beyond some other open-world games on the market in that sense because when you look beyond simple amount of content, there is no evidence proving this magnitudes of increase is actually present.
To further illustrate this, I kick off from a point raised in Digital Foundry's Starfield Direct analysis. They spend a decent chunk of time defending Starfield having no console options beyond 30 FPS, and one of them related to object permanence. Specifically, they said that Starfield (and other Bethesda games) have to mark the state and location of every object a player can interact with, which in DF's eyes, contributes significantly to the complexity of the game and justifies locking it at 30.
In actuality, that is a false assertion.
The way they describe object permanence in Starfield's case is NOT one that increases the processing burden on the CPU/logic side for such a state of time that it would be a contributor to lower framerates. Noting the state and location of an object is as simple as getting the coordinates, a unique reference ID of the object (and any parent object(s)), the location the object was in, and some marker for its state. You can keep that data in RAM if it's going to be used again very soon, otherwise write it out to storage.
The actual math for this should not be that complicated; you can technically break down items to their base components in games like Fallout and I expect that to remain the case in Starfield, but those would be static objects meaning persistent asset instances can be created for said base components and loaded into memory, something other open-world games already do. It's not like you can actually break the objects in real-time, with per-pixel object deformation to calculate. There is no indication that's the case in Starfield and that is the only way in which DF's assertion regarding "object permanence" would actually be relevant to performance dictating a design choice for 30 FPS.
In other words, using "object permanence" as a buzz term may sound like a good enough reasoning to justify no performance options beyond 30 on console for this game, but it doesn't work as an excuse when you break down what the game is actually calculating with those object interactions. And this bleeds over to other aspects of the game, when I question the claims of "magnitudes more complexity" over other open-world games. Todd Howard says there's a planet in the sky and that you can actually travel there...cool. But, when I'm on the current planet, to me that planet you're saying I can go to might as well be a JPEG in a skybox. The vast majority of these planets are barren, so there are not a lot of systems in terms of plant photosynthesis, terrain deformation etc. to calculate. Not only is that not required, but I doubt it would be happening at all. Can I warp to that planet in the sky you say I can travel to? If I can't, then that planet might as well be a JPEG, load in the actual data from storage once I have my cutscene to travel to it. Is anything I'm doing on the current planet, having some real-time cause-and-effect on that planet in the sky I can travel to? No? OK then.
Hopefully in reading this some of you can understand what I mean in talking about scope in terms of complexity of systems and interactions between systems occurring concurrently. Systems that have immediate results on the screen a player can see and feel, as well as background systems that are informed by or help inform the ones producing those visible results. I also hope some of you understand that this is me not saying Starfield lacks such complexity in this aspect of scope; it is very much present and you would be a fool to say it isn't. However, I also feel it should not be controversial to claim that Starfield's degree of such scope in complexity is not "leaps and bounds" beyond at least a few other open-world games on the market, but rather at best on par with them.
Some of the hyperbolic takes around the gameplay footage feel a lot like fluff pieces to me, like an overcorrection in pouring praise onto the game (or like we've seen for the Xbox Showcase itself, and I say that as someone who still says it was better than the PlayStation Showcase) when they don't need to do so. Then there are the ones who want to posit that every other game on the market, or coming to the market aside from Starfield, is immediately a lesser experience. I would say, those are some wild takes fueled mainly by feelings more than anything factual, but I feel some of the misappropriated analysis from the media regarding the game's showing help fuel those wilder takes, and that's unfortunate.
I wrote this in hopes of counterbalancing that while still making sure to acknowledge the impressiveness of what has been shown of Starfield to date. Indeed, even amid the points mentioned above (and some specific visual downgrades that I avoided to bring up), Starfield is going to be a big game, and a big deal. It's the shot in the arm Microsoft needed after the disappointment of Halo Infinite, relative sameness of Forza Horizon 5, non-commercial cult favoring of Hifi Rush and disaster of RedFall. But we don't need to overinflate its importance or impact to show that excitement or appreciation.
