procedural generation

No Man’s Sky: My Concerns

by

(This column is posted at www.StevenSavage.com, www.SeventhSanctum.com, and Steve’s Tumblr)

With No Man’s Sky (NMS), the giant procedural space game coming out, I am gladly analyzing as A) I game, and B) I love procedural generation. So let’s turn back all my speculation on what could be and focus on what could go wrong.

As much as I am enthused about it, I can see areas where the game could have problems. I’m going to explore these areas, so we can review how right/wrong I was – which should be useful to measure both my predictive abilities and how the NMS team works!

Now to make this more useful, I’m going to rank the chances these things could go wrong as Red (at least 50% chance), Yellow (50-25%) and Green (under 25%).  These are not necessary interest-killers or will make it a bad game – but it would be a problem for enjoying it and experiencing the game.

Now let me get predicting:

High-pressure Survival Grind (Red): NMS is a survival game, but my concern is that the game is going to mix high-pressure survival with tedious grind – you’ll be on the edge of your seat all the time, but the edge is going to feel the same and never end. That’ll get both stressful and boring, and that would be an interest-killer.

Hopscotch (Red): Planets may be procedural and bursting with detail, but I’m also concerned that planets could be clusters of neat stuff separated by not so neat. This means hat exploring a planet is really a game of scanning-and-flying hopscotch that will also turn into a kind of grind. My concern is that this would not be optional but required to really experience the game.

Pacing (Red): You start out with little equipment on a distant world, have to survive, and eventually build your technology and resources.  Sounds standard, but unless the game is carefully designed, you could experience highly erratic pacing – most likely a slow start but a surprisingly fast end if you max out equipment (see below).  I also see potential pacing issues in different worlds and goals making it extremely hard to predict what one has to do to achieve a goal – because of the procedural generation.

Every Planet Different – And The Same (Yellow): I’m pretty confident the planets themselves will vary interesting, but not quite confident every planet will be different enough to warrant interest in exploring it a lot. I could be wrong (which I why this is yellow), and the NMS team seems to want to avoid this, but I can’t shake the concern. It seems like there’s a lot of impressive math, but what I’ve seen suggests some relatively standardized environments and all planets are single-environment. That can get boring – it’ll be new then quickly seem the same.

Stretches Of Boredom (Yellow): I don’t mind a bit of boredom or peace. But one of my concerns about NMS is that it’ll have uncontrollable stretches of boredom, stuck on dull worlds and sectors of space.  Good visuals and environments will alleviate or eliminate this (yes, you spent 30 minutes looking for a mineral but it looks awesome).

Topped Off (Yellow): There’s supposed to be all sorts of ships and blueprints to find, but I’d be concerned the game could have some people max out their equipment and the like too early – loosing challenge and initiative. It’s procedural, so it may be hard to put pacing into the game.  This is part of my larger concern about Pacing (above).

The Hunt (Yellow): Certain items, equipment, minerals may be vital for parts of the game, for equipment – but for some players they may be out of reach (again, due to procedural generation). If it’s not something people can find/buy/substitute for in a reasonable amount of time the game may be frustrating.

Same Old Equipment (Yellow): We get various ships, suits, and Omnitools, but from what I see they’re mostly about premade traits and various plugin spaces. Not sure they’re going to be that interesting after awhile.  Are you going to go that far to get an Omnitool that moves a plugin space to one grid cell further rightward?

It Doesn’t Hold Together (Green): Though I trust Hello Games on the Lore, I’m concerned that it won’t be experienced enough, in enough context, to keep interest. The game may not need a story, but it’s sense of experience requires Lore.  The whole thing could not cohere, have no sense of “there.”

Different, But Not Different Enough (Green): I’m mostly confident Hello Games can deliver varied worlds – but not entirely convinced it’ll be different enough for a whole game.  I’m concerned that past a certain point – say about 70% of the way – things will start looking too much alike.  I’m aware we’ve only seen a limited subset of worlds, but I’m not totally convinced.  Yzheleuz and Phlek do give me some hope.  This is one of my lesser concerns, but if planets aren’t different enough from each other and individual planets are large stretch of “same” (above) it’d get boring fast.

So there’s my concerns, roughly boiling down to:

  • Most concerned about serious pacing/challenge issues
  • Mildly concerned about equipment, resources, and variance on a planet.
  • Slightly concerned about lore coherence and worlds varied enough.

What concerns do you have?

– Steve

No Man’s Sky: The Need For Procedural History

by

(This column is posted at www.StevenSavage.com, www.SeventhSanctum.com, and Steve’s Tumblr)

I’m hyped for No Man’s Sky, the space exploration game that uses math to give us a procedural universe – since it’s all constructed from equations, the game has quintillions of possible world to explore.  On the rare time two people find the same world, it would be the same for both due to – math.

But as I’ve read and watched the news on NMS, there’s also talk of the lore of the game.  The story, the meaning.  The developer Hello Games has been very cagey on it, for obvious reasons – they don’t want to spoil the “story” in the game.

This lore, however, is already designed as far as we know.  That brings up something I think it a potential disadvantage in NMS – and in many procedural/random games.  A lot of the “story” is disconnected from the way the setting is made.  The lore is set, and at best sets the stage for the generation of the world – or at worst isn’t just connected anyway.

This means in many cases the randomness of the world is sort of meaningless even if there’s some meaning in the components.  There’s no history, just algorithms.  Why is the dungeon built the way it is?  Why are these artifacts on this world?  I see little to no attention paid to not just generating a setting but the meaning behind it – the history – in many a game.

Like it or not, a lot of these procedural games are about making something that seems “right” but doesn’t have much real history.  Now I love procedural games, I can get into them, but I admit this flaw, and I think the art is limited by this disconnection.  There’s no “real history,” just a shadow play of numbers.

But this also gives us an insight into what future procedural games could be.

What if large chunks of their history, their backstory, are generated?  What if, in turn this history affects the generated environments.  What if this history is part of the lore characters find, from the names of places to the powers of procedurally generated items?  Perhaps the characters themselves are connected to some procedurally generated lore.

Some examples.

  • Imagine an NMS-like-game where the basic expansion patterns/conflicts of various species are procedurally generated – and in turn the effects on certain worlds and areas is created.  Places between two peaceful species have great trade.  Worlds right at areas of conflicts may have graveyards of crashed spaceships.  Bits of history can be worked in, again procedural – you don’t just salvage equipment from a downed ship, but find out when and why it fell.
  • A procedural dungeon crawler could have history generated depending on what the origin is.  If there’s a gate from the netherworld burrowing up from underground, later levels would be older and more hellish.  Perhaps earlier heroes went in to battle and fell, so each treasure has meaning.
  • A game of global domination (or galactic domination) could start not just with the usual empty planet/galaxy trope but one filled with existing politics and peoples – with histories (and relations).  The games become not just standard 4X experiences, but ones of discovering – and manipulating, a rich history.
  • Such games would be not just fun like any good random/procedural game, but also far more compelling.  Rich, unique lore exists -perhaps even if only until roguelike permadeath means you start a new dungeon.  That lore in turn is meaningful because it explains and tells you something about the world.  The tale comes to life because the history has a living quality, not one made static or one bolted onto a randomizer.

Procedural history is procedural meaning, and that brings the game further to life.

Maybe NMS will inspire enough people to do even more procedural work, some will look at procedural history for their games.

– Steve

Make It So: Procedural Jogging

by

The exercise treadmills at my apartment complex are pretty neat – beyond a variety of options (including a game of Solitaire), they have an option to show a video that simulates walking at a location. Basically it gives you something to look at beside your heart rate, has interesting historical notes, and provides an interesting experience. Running indoors can be a might boring, and it’s an appreciated addition I take advantage of because, hey, when am I going to get to jog through scenic parts of Germany.

But it’s the same few tracks. So one day while using the treadmills I began thinking about how it could be more interesting. This conflated with my love of procedurally generated environments, including “walking simulator” Proteus and the amazing environments of the upcoming “No Man’s Sky.”

Then I asked why can’t someone create a Procedural Environment generator for these Treadmills?

Imagine this. You can select from some common setups, or tweak them from the start, and then have the display screen walk you through a generated environment. Maybe you want a desert or a forest, a sci-fi landscape or a fantasy land – set it up and explore as you exercise.

There you go. Every Treadmill trip is a new experience. One day you may walk across a ruin-filled desert, the other up a green mountain path. Maybe you’ll take a run through the Dwarf Kingdom or across an unknown world. Every time it’s different even if you choose the same basic settings.  It’d add a lot to the experience.

Now imagine taking this farther.

Perhaps you could have generated narratives, notes, or events. A bit of history-that-never was pops up here and there. A little note about the properties of an imaginary plant appears. A starship streaks across the sky and crashes nearby. The trip can feel like you’re in a real, living place.

On top of this, perhaps add a bit of gamification. Every now and then you may find an option to climb a stairway or open a door or turn left or right. The trip becomes even more engaging as you have power to affect what you see next – and you never know when the next option will pop up.

Finally, perhaps you can save your experience to share with others, even if it’s just writing down a code. Other people can see the same sites you did – and if there’s options, you can re-explore an old new world differently.

Go on people, Make It So.

  • Steve