Elite Dangerous compared to No Man's Sky

LowEarthOrbit

Orbinautnautnaut
Joined
Jul 21, 2020
Messages
11
Reaction score
1
Points
3
Location
Nish
O-F Staff Note: post moved here from the No Man's Sky release thread so that discussion can continue.

No Man's Sky is a good concept but it needs to be improved upon IMO. Something like a mix of Elite Dangerous's procedural generation and No Man's Sky's vegetation and sentient life would really blow my mind.
Actually, I'm fairly certain Frontier is working on expanding on the procedural generation for Elite Dangerous as well as adding in the possibility for EVA.
It's a pretty cool thing they're working on. If you're interested, here's the link:
https://www.elitedangerous.com/odyssey/ :)
 
Last edited by a moderator:

dbeachy1

O-F Administrator
Administrator
Orbiter Contributor
Addon Developer
Donator
Beta Tester
Joined
Jan 14, 2008
Messages
9,214
Reaction score
1,560
Points
203
Location
VA
Website
alteaaerospace.com
Preferred Pronouns
he/him
Elite Dangerous was great in VR, and I had lot of fun for the first hundred hours or so (particularly in VR using Voice Attack!), but what finally killed it for me was the lack of mission selection at any given station. You only get to choose from ~20 missions total, and the missions that are available don't scale to your ship and/or aren't the type of mission you want to fly at the moment. This post sums up the mission board issues. The game has a lot of potential, though, and I'm hoping that Frontier can redesign the mission boards.

The other major thing that put me off was the endless waiting in supercruise to get to a station, particularly ones that were far away from their star. But it's a good gameplay mechanic for pilots who like to watch Netflix while they game (and some do, I read in the Elite Dangerous forums). :lol:

Speaking personally, I would love Elite again if the devs could make some gameplay changes:

  1. Rewrite the mission system to allow many more missions to choose from, instead of limiting missions to ~20.
  2. Update the supercruise mechanic so that a skillful pilot could fly faster; e.g., maybe using some less dramatic/chaotic variation of the current interdiction mechanic -- if you stay "in the blue zone", your ship flies faster than normal. That would make supercruise fun instead of just a waiting game for 5-15 minutes (or even more, for stations farther out) for each station you fly to.
  3. Add a way to view commodity prices in neighboring systems (even if that's only after you visit a given station first) so that you can find profitable trade routes without having to resort to third-party tools outside of the game.
  4. Allow pilots in private mode the gameplay option to rebuy their ship after it is destroyed without paying the 5% rebuy fee. This gets very, very expensive with larger ships, and it just adds to the frustration when you already lose all your cargo and mission progress and you are respawned back at the last port you visited (which is really, really painful when a pilot was 10,000 light years away on a 500-jump trip) when you lose your ship, even if was due to a game bug. (There was a bug where I jumped into a system inside a planet and lost my ship. That was not a fun day. :lol:)
  5. Last, but not least, add an option to private mode (i.e., singleplayer mode) where you can choose the skill level of the enemy A.I. ships. Currently, the enemy A.I. scales up based on your pilot rank, and the A.I. gets very, very good. So much so that it becomes difficult if not impossible to fulfill "go destroy 100 pirates in X hours" missions. And more than that, some days I just wanted to go blow up a bunch of ships without spending 10 minutes in a dogfight with each one.

I could go also go into the whole grind for the "Engineers" upgrade path, but doing that grind wouldn't be as mandatory if pilots could choose the A.I. pilot skill level in private game mode, at least.

Personally, I love No Man's Sky way more than Elite Dangerous, and VR in NMS is quite polished now, at least in my opinion. The exploration part and building bases w/teleporters ("stargates") between bases is my favorite thing about it. :thumbup:

I did have a ton of fun in Elite Dangerous, though. It's been said that Elite is "a mile wide but an inch deep", which sums up my feelings about it when I stopped playing it a few years ago. But it does have a lot of potential, and I'm glad the studio is still working on it. :thumbup:

Everyone has their take, though, and this is just my perspective. The world is big enough for both games. :tiphat:
 

jedidia

shoemaker without legs
Addon Developer
Joined
Mar 19, 2008
Messages
10,842
Reaction score
2,105
Points
203
Location
between the planets
NMS and ED are two very different games that are almost impossible to compare, though interestingly enough they suffer some of the same problems.

What really can't be compared between the two games is the actual gameplay: One is a no-:censored:-given for realism space exploration game, the other is a "realistic-enogh-for make-believe" MMO (though in my experience, realism and MMOs shoot each other right in the foot).

They may be compared from some technical aspects of procgen, which seems to be the major emphasis of the OP, though even in that aspect they can only be compared to show why you can get the one or the other, but combining them both is a momentuous task.

E:D does a fairly good job with its ProcGen where realism is concerned, and it spends a lot of resources on graphical representation. The one thing you need to realise with ProcGen is that the cost of any ProcGen system rises exponentially with the graphic fidelity required of it. THat's what the guys from Bioware didn't get when they happily decided "let's make a mass-effect game, but it's more like NMS", which was an actual design paradigm behind ME:Andromeda. We all know where that ended up: With most of the work getting thrown out of the window with 18 months left to actually make a game.

In any case, E:D does a great job at this, but on a very limited scope. Even 6 years after launch there still are only the most basic types of planets (and let's not forget, there were *none* at launch). And while I don't doubt that Frontiers are experimenting with systems to expand that and intend to do so, I seriously doubt that they are committing major resources to the task.

NMS on the other hand skipped all the work E:D did with its magnificent galaxy generation, it skipped all the work of coming up with believable planetary formation, all that jazz was just glossed over. The generator for the overlaying structure of the galaxy was probably written in a week or so. Together with all that, they abandoned the major paradigm that is a bare necessity for systems that want to do planets at a realistic scale: The 2.5D approach that everybody, from Spaceway over Spaceengine to ED is using. You generate the sides of a cube as noise and project it onto a sphere, then generate microterrain by octaving. That system has one major disadvantage that Hello Games managed to bypass: It's really only 2-dimensional. You don't get caves, you don't get overhangs, you don't get most things that actually make things look interesting when viewed from *really* close up. You get a 2-dimensional blanket that has ripples in it, but no actual dept, so to speak

That system is still used in NMS, but only as the first LOD. After that, the game starts to build an actual 3-dimensional terrain on top of it using voxels, which is much more similar to minecraft than to E:D. But it pays a price: The planets are smaller than they should be, though just big enough not to matter once you're standing on them. The highest mountains in NMS could be considered hills at best, and the deepest ocean feels more like a lake. The reason for this is memory, of course, since any given stretch of surface in NMS requires hugely more data than a comparable area in E:D, and the thicker the layer of 3-dimensional terrain on the blanket gets, the more that increases.

When flying over them in a spaceship, NMS planets often reveal their themeparky nature in a painful way. It's really quite boring. Compare to ED, where you can find height differences between valey bottoms and mountain peaks of 10'000+ literal meters, it's almost laughable.
But then you land, and get out of your ship, and all the impressive terrain of E:D becomes bland, samey and boring while in NMS a charming, intimate landscape starts unfolding.
The two focused on entirely different things in their procgen, and they took two completely different technologies to get there. The big fallacy is thinking that these approaches could actually be combined to get the best of both worlds. Technically they can, but not as long as either game is bound to consoles. And without console sales, neither game would be feasible financially. So practically, I'm afraid, it's impossible. And I'm not even mad at that, because it means that both games can still run on my laptop...

I mentioned above, though, that both games share some weaknesses. And I'm quite frankly flabbergasted at the lack of will by either game to resolve those, although there are reasons.

The biggest weaknes of both games are a lack of emergent gameplay. In both games, there's just a huge lack of things that *could* happen. There's almost nothing surprising or unpredictable to either of them. ED attempts to compensate for this in two ways as far as I can tell: It's MMO mechanics as well as hand-placed easter eggs and narrative hooks. Completely forgetting that either of those are almost irrelevant if your possibility space encompasses a literal galaxy. You can't be eve when you have billions of planets people can disperse to, and even the populated bubble dwarfs anything any MMO has ever done. You'd literally need millions of players to make the MMO parts actually mater for players that don't actively seek them out, and those players limit their possibility space severely, to a point where the entire generated galaxy becomes irrelevant. E:D has created a huge possibility space, and is now at a loss for how to effectively exploit it.

HG on the other hand keeps heaving new content into the game to fill that space, almost every single piece of which repeats the same problem in its microcosm: That nothing really unpredictable ever happens. A new update comes out, you get surprised by a few things, but after another two to five hours you know how everything works and you're back at the old problem that nothing unpredictable ever happens. Don't get me wrong, I enjoy many of those additions a damn lot. But things don't connect. If there was an overarching system that made all those things connect in ways that would be harder to predict, we might have something, but so far no attempt has been done to really tackle that.

One reason for this is because it is notoriously difficult to make procedural systems interact. It's anathema to their encapsulation, which is the entire point of procgen. There's no fixing this, but I do believe that there would be approaches that could at least alleviate it. Somebody just has to have a really bright idea first. Until then, I still think a lot could be done by second and third-order generators that do not just create things from lookuptables, but generate diverse lookup tables for the generators that create things from those lookup tables, and that generate lookup tables to generate lookup tables to generate things. It's hard, it's a lot of work, and nobody can tell whether it would actually be worth it. I think that's the main barrier here.

As for E:D, I think the best thing they could do to break the mould would be such a second-order generator for missions, that could string them up in a not entirely predictable manner and put some light narrative glue between them. The game would require more interaction points between its procedural world and the player to make it really worthwhile, though. But ED decided to be an MMO despite having a world not very suited for it, so that's what we have...
 
Last edited:

Linguofreak

Well-known member
Joined
May 10, 2008
Messages
5,017
Reaction score
1,253
Points
188
Location
Dallas, TX
Compare to ED, where you can find height differences between valey bottoms and mountain peaks of 10'000+ literal miles, it's almost loughable.

Wait... 10,000 *miles*? That blows realism away with out-of-scaleness in the other direction: you could hide frickin' *Earth* in a valley of that scale, and Neptune would be an unusually prominent mountain, but not overwhelmingly so.
 

jedidia

shoemaker without legs
Addon Developer
Joined
Mar 19, 2008
Messages
10,842
Reaction score
2,105
Points
203
Location
between the planets
Whoops, sorry about that... :lol: I meant meters, no idea why I wrote miles. Possibly because they both start with an m? No idea, it's late...
 

Linguofreak

Well-known member
Joined
May 10, 2008
Messages
5,017
Reaction score
1,253
Points
188
Location
Dallas, TX
Whoops, sorry about that... :lol: I meant meters, no idea why I wrote miles. Possibly because they both start with an m? No idea, it's late...

Mille passus formicae?

Your punishment is to move to a country where all the roads have speed limits of 100 mph. *Meters* per hour.
 
Top