Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Flying Dice]

And I mean yeah, you probably won't find people in random unexplored systems with nothing in them, but folks will figure out where stations and valuable resources are, so populations will naturally develop in areas like that.

[Putnam]

Eh. I dont really see how. Most of the time that's just how procedural generation is. When the generation of creatures follows a certain set of rules, it's not like you're going to see something that's going to knock your socks off every single time.

that doesnt excuse it being bad

procedural generation is a tool, not a goal.

I'm not so sure about that. Look at the title of this thread: "18,446,744,073,709,551,616 planets to explore". Why generate 18 quintillion planets? Why hype up that absurd number specifically? What's the use case of this, knowing that the end user is not actually going to see all 18 quintillion planets? It seems that procedural generation for No Man's Sky is the whole point, the raison d'être of its existence.

It's not an absurd number, it's a ridiculously typical one, being the size of a 64-bit integer.

[Neonivek]

Eventually I just want to make a rule of thumb

Which is the more variations there can be in a world the equally less chance there is anything interesting to see after the first.

[Putnam]

I don't think that works, since the number of variations these days is more often than not going to be exactly 18,446,744,073,709,551,616. That's how many minecraft worlds there are, for example.

[Neonivek]

I don't think that works, since the number of variations these days is more often than not going to be exactly 18,446,744,073,709,551,616. That's how many minecraft worlds there are, for example.

The rule of thumb is more that the focus on randomization is far inferior to the concept of making things interesting.

[Malus]

I don't think procedural generation is, by itself, particularly interesting. It's only when you weave in elements of simulation, like in Dwarf Fortress and Ultima Ratio Regum, that it becomes really worth exploring. (DF generates some pretty neat histories, it's just too bad the Legends interface makes them hard to find.) Without any kind of simulation behind the scenes, things lack context. Why does this long-necked herbivore exist on a planet with nothing but shrubs? Because the random number generator made it so.

Games are pretty terrible at this in general so I'm not gonna fault NMS for it. Stellaris opted for procedurally generated alien empires and the result is that they're all just very same-y. Once you've met one <personality type>, you've met them all, but hey, these ones have scales!!

Ultimately, what determines whether NMS is going to be good or not, is the gameplay and mechanics, which are entirely hand-authored. The planets might be beautiful, but if that's all it has going for it, then like greatorder said, you could just boot up Space Engine. _{Or Shores of Hazeron.}

[LoSboccacc]

DF generates some pretty neat histories, it's just too bad the Legends interface makes them hard to find.

even then for one cacame millions uninteresting wolrds are generated.

[Neonivek]

DF generates some pretty neat histories, it's just too bad the Legends interface makes them hard to find.

even then for one cacame millions uninteresting wolrds are generated.

Even Cacame's world was uninteresting for the most part outside himself.

But Dwarf Fortress' generation has a point beyond for generation sake. As its histories and landscape are ultimately changed by your actions and the world thusly reacts to it.

It isn't having a million variations of a world for the sake of having a million variations of a world.

[Sirian]

I don't think procedural generation is, by itself, particularly interesting. It's only when you weave in elements of simulation, like in Dwarf Fortress and Ultima Ratio Regum, that it becomes really worth exploring. (DF generates some pretty neat histories, it's just too bad the Legends interface makes them hard to find.) Without any kind of simulation behind the scenes, things lack context. Why does this long-necked herbivore exist on a planet with nothing but shrubs? Because the random number generator made it so.

Games are pretty terrible at this in general so I'm not gonna fault NMS for it. Stellaris opted for procedurally generated alien empires and the result is that they're all just very same-y. Once you've met one <personality type>, you've met them all, but hey, these ones have scales!!

Ultimately, what determines whether NMS is going to be good or not, is the gameplay and mechanics, which are entirely hand-authored. The planets might be beautiful, but if that's all it has going for it, then like greatorder said, you could just boot up Space Engine. _{Or Shores of Hazeron.}

Yea I think that procedural generation is useful in some games but it shouldn't be understood as an infinite content generator. In a roguelike game it is useful because you don't want your players to learn the optimal route to solving the dungeon, therefore you use procedural generation. Anytime you want people to explore the game world without prior knowledge of the map, it can be useful.

But the created content will be, by nature, generic. If you want to generate a whole universe, you can use procedural generation, but you better not make that the focus of the game, instead you could try to have some sort of emergent gameplay from player interaction or something similar.

In Minecraft for instance, what is interesting isn't the infinite worlds but rather what the users themselves create, be it in-game constructions or mods, which are the true content of the game, (user generated content > procedural content).

Edit : basically, procedural generation doesn't generate content, it only produces variations of the same content. It can sometimes produce something interesting through sheer luck, but this will be a fleeting thing, soon forgotten and incomparable to handcrafted content made to be interesting as often as possible.

[LoSboccacc]

yeah I loved to build underground lofts in the generated caves at their deepest, with lava and crystal walls, precisely because every time it was a different challenge

[miauw62]

Eh. I dont really see how. Most of the time that's just how procedural generation is. When the generation of creatures follows a certain set of rules, it's not like you're going to see something that's going to knock your socks off every single time.

that doesnt excuse it being bad

procedural generation is a tool, not a goal.

I'm not so sure about that. Look at the title of this thread: "18,446,744,073,709,551,616 planets to explore". Why generate 18 quintillion planets? Why hype up that absurd number specifically? What's the use case of this, knowing that the end user is not actually going to see all 18 quintillion planets? It seems that procedural generation for No Man's Sky is the whole point, the raison d'être of its existence.

There doesn't seem to be any use case for the procgen in this game, other than to produce an illusion of content and depth that makes the rest of the game seem immersive. Maybe you can say that's where proc-gen being used as a tool, but no, I think it's just being used for...well, hype isn't the right word here...

An indie game developer, Bruno Dias, is currently developing a space game called Voyageur, where procedurally generated text is used to provide description for the various planets you can visit. He mentioned casually in an interview:

"In games, procgen originated as a workaround for technical limitations, allowing games like Elite to have huge galaxies that never actually had to exist in the limited memory of a 90’s computer. But it quickly became an engine of surprise and replayability; roguelikes wouldn’t be what they are if the dungeon wasn’t different each time, full of uncertainty. Voyageur represents an entry into what we could call the “third generation” of procgen in games: procedural generation as an aesthetic."

That's what procedural generation is being used for in No Man's Sky. Sheer beauty and taste, divorced from form and purpose. Let's have a vast, massive galaxy that you can explore and see content that no other player will ever see, so that players can then partake of meaning that is meaningful solely because it is specific to them and to them alone...Not out of any sense of it actually being useful, but because it's just cool.

Like most things about aesthetics, it's really a matter of taste, and while I can see many people who would love procgen for procgen's sake, I can also see why it wouldn't be appealing to many other people as well. And, of course, my mini-rant doesn't excuse bad procedural generation...

yeah, they advertise 18 quintillion "unique" planets, but all of them are going to feel the same. that number is just a very scummy marketing trick, imo. it's a really big number ment to distract from what that actually means.

the problem is that procedural generation as an aesthetic has consistently been shit. you need to design a very sophisticated procedural generator and then only use it, say, half a dozen times per playthrough. "the trees on this planet are pink" is unique, but not memorable, because you've already seen twenty other palettes for trees on twenty other palettes.

This obviously isn't a problem when procedural generation isn't literally your entire game. people don't care that their civ map is unique or not, they just want it to be unexpected .you can procedurally generate other things to keep things fresh and add replayability. but after so many disappointments, i dont believe in procedural generation for its own sake.

[spümpkin]

I honestly have never been interested in this game solely for its procedural generation. I'm not gonna try and change your opinions (if you'll refer to my personal text), but I've personally been interested in it, because I just have a really insatiable wanderlust, and want to go around exploring planets etc. And that's why I enjoyed Starbound. So, 3D Starbound where you can fly spaceships? That sound's pretty good to me. But yeah, I do agree that it could've been marketed better, focusing more on it's other, actual aspects instead of procedural generation. But when it comes out here on the 13th, I will be rearing to play it.

Also, I think it'll just make for good stories, whether or not the gameplay is 'good'. Procedural generation systems have a habit of doing so.

[Neonivek]

By the by...

I am glad YEARS later that people are finally getting the message that "and now this patch of dirt is over there!" is not where "Fruitful content" lies.

But GOODNESS did it take a while for people get this point... I had to sit there for quite some time hearing about how ALPHA TERRARIA is just soo totally drastically different every time and it is like jumping onto an entirely different planet!

It didn't help, I guess, that I am rather aggressively pessimistic so people likely compensated by overhyping procedural generation.

---

I know this is silly to say... but Spore... did planet generation better.

If only because they understood that random generation could only take things so far and so they actually programmed in variations that could occur in themes.

Not that Spore was flawless, but why is it that we need so many games to hint at these concepts before we stop making the same mistakes?

[LoSboccacc]

spore creatures were soo cool before cartonification 

[Salmeuk]

I know this a NMS thread, but Spore was the game that changed my perspective on AAA hype. As in, I no longer hype at all because if I do drum up expectations then I might purchase a game and get something entirely unlike what I imagined it to be. Everyone remembers the early demos of Spore, right? That looked amazing, in fact they still do! Just like those early demos of NMS.

I've been watching gameplay streams, and

Spoiler (click to show/hide)

it looks not that great. The combat / upgrades / spacefaring seems awfully simplistic, and even though they seem to have created a huge number of potentially interesting worlds I have no interest in exploring those worlds in the context of meaningless progression. The beauty is in the meaning, just look at DF and the importance of world history in appreciating the landforms. Proc gen, if it strives to be interesting to a human, should always be paired with a meaningful narrative. And I long for the day that devs realize gradual progression of material wealth is a terrible narrative (a narrative reserved for the disturbingly dull reality we are all forced to share).

If I wanted to just explore galaxies I would save sixty bucks and play Noctis.