The emptiness is, the non empty areas being generic was laziness. Though I suspect both are the result of Bethesda expecting modders to fill out the galaxy.
I don't think it's laziness, I think it's a tighter grip over development.
I've heard that with Skyrim's development, many of people's favourite side quests were conceived, performed, and coded in a couple nights by a couple of people just cause they thought it was a fun idea, vs. Starfield had a way more locked down and controlled development cycle with way more of a top down, prescribed, development style.
I never heard why that might be, but it does make sense in terms of if you were trying to plan content for 1000+ worlds and doing it top down, you would lean way more on procedural generation and consequently you'd want way tighter controlled areas and things that can happen to not mess with your multi planet generation engine, but then that ultimately results in them all being soullless.
That and they broke the fundamental exploration loop by making loading screens such a core and frequent requirement. In Skyrim you'd have to do a quest over there, start walking towards it, see something neat, go check it out, find someone doing something, help them kill this villain, collect a reward, climb to the top of the mountain you're partway up, look around and see some more cool stuff, remember you have a quest to do, and climb back down and towards the cave entrance that starts your quest.
With Starfield you have a quest to do over there so you fast travel to your ship, fast travel to the planet's orbit, and fast travel to the quest location.
It does make sense. I mean it's about resource management; they could add more locations, but that might mean more burden on QA, maybe some performance optimizations to support newer sections, maybe dialing back some of the art work on other areas to make it happen. Skyrim had great quests but no doubt it still performs quite poorly even on modern consoles at times, and still has more bugs given its relative size.
I got a strong sense of laziness from so many aspects of the game, but now I've had more time to think about it... it's more likely yet another case of upper management forcing the game out earlier than the devs wanted. Same thing happened with CP2077.
Upper management often doesn't give a shit about the game at all, all they're interested in is how much money it's going to make; so the longer it's in dev, the more times it gets pushed back, the harder they're chomping at the bit to just release it already.
There was probably some pressure as the game was in development for a really long time it got already delayed by almost a year, but I also feel like it has a lot of design choices which aren't great. The development started right after Fallout 4, so I feel like 8 years of development is a pretty long time and if it had to be rushed out, there had to be some management problems. Feature creep is probably what especially happened, because there are so many game mechanics and features that don't really have good impactful use cases.
Not to spoil anything, but the story and lore is pretty meh. There's really no real sense of exploration like I have felt in Elder scrolls or Fallout. There's just too many loading screens all the time (This is probably an engine limitation, but something that should have been realized in the early phase of development. Probably not really something you can change later). There's a lot of sidequests, but most are just boring without a real story behind them.
The AI is just dumb and once you play for a while, the battles just doesn't challenge you in any way even on higher difficulties.
The outpost system just doesn't make sense. It's not like there's not enough features in the outpost system, it just doesn't make sense to use it properly. The stealth system often doesn't make sense and is not useful in any way.
And to not to end it too negatively, I still liked the game. It's just frustrating because I see so many potential game mechanics around me while playing that don't really make sense to use.
The development started right after Fallout 4, so I feel like 8 years of development is a pretty long time and if it had to be rushed out, there had to be some management problems.
That's a really good point tbf. 8 years, and this is what they released!
I've completed the game and agree, while it is an engaging game in some aspects, many systems felt 'mile wide / inch deep'. Especially the AI. And I fucking hated the stealth, it just didn't seem to work (though apparently a big reason for that is, many of us had armour 'set to invisible in settlement' which meant we're tramping around with all our armour on, which makes us very visible / audible even though we can't see the armour ourselves).
The development time was longer because they finally rewrote their old engine, so several of the years were just getting the engine working. So the game itself still could have been rushed out in the last couple years.
By design was using procedural generation to build out the planets and to place points of interest into the world.
But what's wild is that they had such repetitive and uninteresting "points of interest."
They aren't interesting at all. The name is a lie.
So yeah, if you wander the procedurally generated planet you can come across a cave. But it's effectively the same cave you have gone through a half dozen times already.
My suspicion is that the creation engine isn't very well equipped for complex procedural generation, so the point of interest areas are also broken up onto reused tiles that get repetitive quickly, as opposed to more modern engines that allow for procedurally generating levels and assets with enough entropy that even if it looks familiar it isn't that it looks exactly the same.
Like, just look at the demoed features of UE5's procedural generation, and how much less repetitive it looks when filling in areas than Starfield.
I expected more from Bethesda given Todd Howard has been focused on procedural generation for decades now, but I think they really shot themselves in the foot with not sending creation engine to a farm upstate.
Yes, they were able to reuse a lot from past development, but the core of what they were trying to deliver completely fell apart ultimately and his decades old dream project is going to be forgotten within a few years.
It's too bad procedural generation gets away as such a blanket term. What Starfield's non quest planets really have is procedural generation of the geography of the surface, and then procedural placement of hand designed plant, animal, and mineral assets, then procedural placement of hand designed points of interest and buildings. The factory dungeon you just went through was by no means procedurally generated.
Wasn't this by design?
The emptiness is, the non empty areas being generic was laziness. Though I suspect both are the result of Bethesda expecting modders to fill out the galaxy.
I don't think it's laziness, I think it's a tighter grip over development.
I've heard that with Skyrim's development, many of people's favourite side quests were conceived, performed, and coded in a couple nights by a couple of people just cause they thought it was a fun idea, vs. Starfield had a way more locked down and controlled development cycle with way more of a top down, prescribed, development style.
I never heard why that might be, but it does make sense in terms of if you were trying to plan content for 1000+ worlds and doing it top down, you would lean way more on procedural generation and consequently you'd want way tighter controlled areas and things that can happen to not mess with your multi planet generation engine, but then that ultimately results in them all being soullless.
That and they broke the fundamental exploration loop by making loading screens such a core and frequent requirement. In Skyrim you'd have to do a quest over there, start walking towards it, see something neat, go check it out, find someone doing something, help them kill this villain, collect a reward, climb to the top of the mountain you're partway up, look around and see some more cool stuff, remember you have a quest to do, and climb back down and towards the cave entrance that starts your quest.
With Starfield you have a quest to do over there so you fast travel to your ship, fast travel to the planet's orbit, and fast travel to the quest location.
It does make sense. I mean it's about resource management; they could add more locations, but that might mean more burden on QA, maybe some performance optimizations to support newer sections, maybe dialing back some of the art work on other areas to make it happen. Skyrim had great quests but no doubt it still performs quite poorly even on modern consoles at times, and still has more bugs given its relative size.
I got a strong sense of laziness from so many aspects of the game, but now I've had more time to think about it... it's more likely yet another case of upper management forcing the game out earlier than the devs wanted. Same thing happened with CP2077.
Upper management often doesn't give a shit about the game at all, all they're interested in is how much money it's going to make; so the longer it's in dev, the more times it gets pushed back, the harder they're chomping at the bit to just release it already.
There was probably some pressure as the game was in development for a really long time it got already delayed by almost a year, but I also feel like it has a lot of design choices which aren't great. The development started right after Fallout 4, so I feel like 8 years of development is a pretty long time and if it had to be rushed out, there had to be some management problems. Feature creep is probably what especially happened, because there are so many game mechanics and features that don't really have good impactful use cases.
Not to spoil anything, but the story and lore is pretty meh. There's really no real sense of exploration like I have felt in Elder scrolls or Fallout. There's just too many loading screens all the time (This is probably an engine limitation, but something that should have been realized in the early phase of development. Probably not really something you can change later). There's a lot of sidequests, but most are just boring without a real story behind them.
The AI is just dumb and once you play for a while, the battles just doesn't challenge you in any way even on higher difficulties. The outpost system just doesn't make sense. It's not like there's not enough features in the outpost system, it just doesn't make sense to use it properly. The stealth system often doesn't make sense and is not useful in any way.
And to not to end it too negatively, I still liked the game. It's just frustrating because I see so many potential game mechanics around me while playing that don't really make sense to use.
That's a really good point tbf. 8 years, and this is what they released!
I've completed the game and agree, while it is an engaging game in some aspects, many systems felt 'mile wide / inch deep'. Especially the AI. And I fucking hated the stealth, it just didn't seem to work (though apparently a big reason for that is, many of us had armour 'set to invisible in settlement' which meant we're tramping around with all our armour on, which makes us very visible / audible even though we can't see the armour ourselves).
The development time was longer because they finally rewrote their old engine, so several of the years were just getting the engine working. So the game itself still could have been rushed out in the last couple years.
Not really.
By design was using procedural generation to build out the planets and to place points of interest into the world.
But what's wild is that they had such repetitive and uninteresting "points of interest."
They aren't interesting at all. The name is a lie.
So yeah, if you wander the procedurally generated planet you can come across a cave. But it's effectively the same cave you have gone through a half dozen times already.
My suspicion is that the creation engine isn't very well equipped for complex procedural generation, so the point of interest areas are also broken up onto reused tiles that get repetitive quickly, as opposed to more modern engines that allow for procedurally generating levels and assets with enough entropy that even if it looks familiar it isn't that it looks exactly the same.
Like, just look at the demoed features of UE5's procedural generation, and how much less repetitive it looks when filling in areas than Starfield.
I expected more from Bethesda given Todd Howard has been focused on procedural generation for decades now, but I think they really shot themselves in the foot with not sending creation engine to a farm upstate.
Yes, they were able to reuse a lot from past development, but the core of what they were trying to deliver completely fell apart ultimately and his decades old dream project is going to be forgotten within a few years.
It's too bad procedural generation gets away as such a blanket term. What Starfield's non quest planets really have is procedural generation of the geography of the surface, and then procedural placement of hand designed plant, animal, and mineral assets, then procedural placement of hand designed points of interest and buildings. The factory dungeon you just went through was by no means procedurally generated.