My soreness about No Man's Sky has led me down an interesting path. I was (and still am) incredibly disappointed in the game because I wanted it to be the successor to Freelancer, which it really wasn't. NMS wasn't really much of anything. I wouldn't even call it a space game, since there's literally nothing to do in space, it's just hopping from planet to planet. Anyway, my frustration with the game has led me to thinking about how I would build a proper space game, and I've been writing down a lot of ideas since my previous post. Then I learned more about Star Citizen and realized that it will have about 90% of all the ideas I've written down. I'm bloody amazed by that. Star Citizen is going to be epic.
The reason I think Star Citizen is going to be amazing is something I realized as I was preparing to build my own space game. After having written down all my ideas I realized that I am pretty much capable of implementing all of them. "All I need to do is pick a suitable 3d engine and learn how to use it. Easy peasy". Of course, before I even got to that stage I realized that the basic setup I was envisioning would take me years to develop by myself, even if I quit my full-time job. But I can see quite clearly the reasoning that goes into building a space game: it would take me crazy amounts of time to do this all by myself, so that is a problem that needs to be solved.
Procedural generation is one of the ways to solve that. Even as a one-man team you'll be able to generate crazy amounts of 'content'. The problem with that is, as No Man's Sky quite sadly demonstrated, that the content will become utterly boring and repetitive, no matter how good your algorithms are. One of the key features I wanted in my own space game is that there are factions and factions within factions, and standings between factions, and a galaxy whose factions keep changing. NMS has none of that: it's just one bland galactic plane of a little bit of everything everywhere but not anything specific anywhere specific. Better procedural algorithms can fix this. They can add more structure, more variety, more realism. But in the end, once you 'get' the algorithm, the game is over for you. You will find nothing new.
Star Citizen does not have this issue, because the team of Star Citizen has lots and lots of money. They get to hire lots of people and they get to handcraft their content. Each of their star systems is meticulously designed; its history written and rewritten until it is perfect. There is a storyline, there is custom, unique content. It's the exact opposite direction of No Man's Sky. It's quality versus quantity.
The quality-vs-quantity thing has always been an issue with games. I remember the Grand Prix series, of which 2 was brilliant, 3 kinda showed Crammond's inability to keep up, and 4 came out in a time where the competition had larger teams and managed to release a much more polished product. A more recent example is Kunos and his netKar series, followed up by Assetto Corsa. netKar was a kickass sim in its day, but you could clearly see its limitations for having been built by one man. netKar Pro started out with the engine sounds synthesised rather than sampled, which is surely a faster way to do it, but nowhere near as immersive as having the actual engine sound as recorded, which is what they did for Assetto Corsa. The cars in Assetto Corsa are incredibly well made, with as much detail crammed in as they could. A manual process that could only be done by a larger team.
In retrospect, No Man's Sky focused on all the wrong things. It focused on procedural generation as its main gimmick, which it should never be for any game. It should be used to assist in making the game more immersive, but if you take it away the game still needs to be a game. No Man's Sky is just nothing without it. NMS also focused on being able to take off from a planet and flying into space. It's a cool gimmick, but it was implemented poorly and, to be honest, I couldn't care less about it. It's the liveliness of space and each of the planets that immerses me. With no backstory it's just not interesting.
I'm still kind of interested in building my own space game. I suspect that when Star Citizen comes out it'll satisfy most of my space needs, but there's still things that I'd like that Star Citizen doesn't have, or do differently. The ship monetization in particular is a thing I definitely do not like about Star Citizen. Also, in SC the factions seem pretty unchanging and static. I wonder if there'll be player actions that can influence the balance of the galaxy, other than predefined events by the developers. When it comes to the perfect space game we're still not quite there yet.