Improbable Icon


Game World Sizes



We had an interesting query from @zachc over on Twitter. Thought I’d answer here as well for visibility!

In terms of games currently being developed on SpatialOS, the Lazarus map is 400km by 400km for a total game world of 160,000km^2.

As far as I know, the largest contiguous land mass in a game is The Elder Scrolls II: Daggerfall at 161,400km^2. Now, things become tricky when you start looking at games like Eve Online and how to measure geographic distance there. That being said, you could well create a physical map larger than that of Lazarus on SpatialOS. The question then would be whether it’s worth doing so and how it benefits players.

Cool things happening right now with SpatialOS

For Elrakis I am using a larger map where each unit represents 1 AU and I currently have an area of approx 500.000 by 500.000 AU, My aim is to create a seamless experience in space and indeed: it has it’s own special challenges


Which games have you most enjoyed where the scale of the world itself is a core feature of the fun, and how has that changed gameplay?

(An interesting thing here is that many games appear to promote huge scale, but the area players experience and affect is actually pretty small?)


Speaking for myself; I find that it breaks the immersion of a game when it contains unbelievable experiences. A game set in space, for example, loses part of its charm when the science is way off. An example is that in No Man’s Sky the scale of a solar system is disturbingly off. All planets are immense and really close together. For me it breaks part of the feeling.

In addition, I consider games revolving around exploration and discovery tend to benefit from more room. I do however agree that void is meaningless unless it contains significant content. Large sizes alone mean little, what you do with that is what counts.


The other thing that interests me is how the nature of questing changes if the world is truly open, persistent and massive - as opposed to quests being sideshows, where each player can complete the identical fork, in isolation. Are you planning to do anything on this in your game @draconigra?


For me large open worlds are all about isolation.
I think I’m yet to find an experience where the loneliness and isolation of being in a truly open environment felt real enough to be enjoyable. In most games its obviously faked: and like @draconigra said it can be pretty jarring.
I’m unsure that having to walk for an hour across a desert on your own would be a fun experience in real life; but I can totally see people enjoying the sense of being truly alone.
“Journey” (PS3 title)was pretty good at making you feel isolated, although it was set in a small finite amount of space.

I think some people play single player games to be on their own and escape…but imagine if the world was big enough that you could do both! Carve out your own private corner of the world and keep it secret: drifting in to society only when you need. Maybe that would appeal even more to a loner like myself?!


@callumb In terms exploration Metroid Prime was a great example of using isolation as a game element. However, whilst it took place in an intricately connected game world, it wasn’t exactly an open world either. Uncharted 3 had a go with their brief foray into the Rub’ Al Khali which was a wonderful experience although criticised by players for being overly long, in spite of a brief 15 minute runtime.

I think that consistency is key as @draconigra noted. The fact that the planets in No Man’s Sky do not actually rotate accurately and instead have the sun rotate them, like some sort of 16th century Catholic astrological treatise, was what broke it for me.

Then again, I guess what we’re really discussing here is whether or not the sparsity that comes with such gigantic land-sizes are a boon or bust to players. Whilst @callumb appreciates solo exploration, I also eventually got sick of the endless wastes of Mad Max and the pointlessness of driving around to grab collectibles and repeat cookie-cutter side content. Clearly there’s always a balance.


I agree with all the points above; as an example: for Elrakis the scale is large but the main events take place in ‘zones’. Which are mainly centered around star systems. As such the void between systems is an atmospheric element (even worse: I am considering ‘folding’ the void so that the space in Spatial is less than what the player perceives).

@mark regarding quests and story line: my biggest beef with MMOs is grinding and the linear character of the quests. Although it is premature to comment to much on what I have planned I can tell you what I would like to build: that players are offered a storyline where their choices offer alternate paths through the storyline, comparable to a ‘Choose your own adventure’


@draconigra I’m currently writing something regarding modular design within traditional narrative and quest frameworks. Would be great to get your feedback once its posted! Keep an eye out for the first week of the New Year :slight_smile:


On the subject of exploration, in my opinion, it is not meaningful on it’s own. People explore to find something not for exploration sake.

Also, I think that big game world with thousand km have been made already. What about a game world that is deep instead.
Instead of expanding the world outward why not inward?

Great discussion BTW !


A good point @SionoiS, about meaning - what would be interesting is how to code games where what happens in different zones/playing areas actually affects others, even if the absolute scale of the world is condensed/divided in various ways. (Or movement/travel is shortened, even if the world itself is one massive persistent space, to avoid the vapid emptiness or pointless 21km commute that @callumb describes!)


I would love to give that a good read through! I’m a fan of good storytelling in new media and I hope that being a D&D Dungeon Master for almost 2 decades can also bring some interesting perspectives :slight_smile:


@mark what you describe here is actually one of my pet peeves with MMOs. I get the feeling that it is not too impossible to have a Narrator AI (as I dubbed it) that tracks a larger story in a game world and create interactions in locations where players regularly visit.

I love the Nemesis system of Shadows of Mordor for this and it would not be too impossible to set this up in a larger scale, even to the point of giving individuals in your gameworld their own personalities.


Size, or player numbers, or entity count don’t in themselves make anything fun.

Our view (at improbable) is that if we can make creating these things trivial and unimportant they can act as building blocks for developers creating new types of core experience. New genres basically.

We take for granted that games can be 3d or be multiplayer, once we take these new tools for granted and begin to shape them into the concepting phase for new projects, that is when I think we will start to see really profound innovation.


I think the problem is also in how centers of game action tend to be path-dependent when games are multiplayer. Since most multiplayer games have pretty static game worlds, there’s nothing to disrupt this.

Works like this: in a more freewheeling game like say, DayZ or the old anarchic PVP servers in Dark Age of Camelot, certain areas emerge as action centers. Then everyone goes there since they know that’s where the action is, making the area even more of a center of action.

This means that the geographic spread of “interesting experiences” is highly clustered. So even cool and well-mapped worlds like DayZ’s can be largely devoid of any interesting stuff because all the peak experiences are in 2-3 places.

In worlds like these, @callumb’s solo exploration is accommodated but kind of meaningless because you don’t end up doing anything especially interesting by going off by yourself.

What I’m most excited by is the possibility that SpatialOS will allow players to so deeply alter the world that these action centers are unstable. If you could flatten the town where everyone loves to camp, you would certainly inspire some geographic reshuffling!


I also think that, as a technology, Spatial opens up a legion of new possibilities. For example, most games ‘turn off’ areas where there are no players, meaning that those stand still in time effectively. With Spatial you can at least simulate on-going events in that world, making for a more immersive experience when a player returns to a location to find out stuff has changed or is going on.


So I wonder what features of a game world could act over time to increase the value/interest/fun of going to emptier areas, either independently (say, due to ecology/coded evolution - resources bloom in sparser regions?) or because of high impact actions taken by the fewer players who did visit?

This might mean that a designer could create a world that both encourages higher total simultaneous player numbers, and mitigates the cramped action centres point @zachc makes to deliver the increased meaning/fun and completely new class of game that @Herman describes.

(Just thinking out loud) players who choose to be explorers could take actions that actively improved the value (however that manifested itself in your game) of regions they discover for others who chose to focus on building/society creation in the later waves of player movement. If the “value” only benefited certain factions/player classes/in-game social constructs, these world-altering actions could form part of either competition or collaboration, and encourage much more complex gameplay and clan/team/alliance planning because the persistent evolution of the world would mean the changes are not stable?


I love the idea, @mark. A payoff to @callumb style explorer-pioneers definitely makes sense. The challenge seems to be that there are few non-contrived ways to keep the pie expanding so that meaningful exploration is always possible.

For example, a big region could have resource deposits open to players who discover them. Prime locations get claimed, setting off the path-dependent hot spots. Resources are depleted unless the game world reset them, renewed them, or generated new ones.

Without some sort of renewal, the world becomes zero-sum. Or perhaps if resources are not ‘removed’, but just transformed, then scavenging can replace the initial extraction from the game world like in World’s Adrift.


If you applied the rule “energy is never created or destroyed, just changed in form” to resources, or similar, then you would have a world where part of the competition/collaboration dynamic depended on where in the energy cycle you placed yourself, relative to others.

E.g. some harness heat, others light, others chemical, gravitational etc (potential energy could be a tough one!) - and energy “use” results in changing quantities of other energy types. Some you could use, and some others could.

Gameplay would then evolve to focus on controlling larger sections of the energy cycle, however it was described in your game. Disruptive interventions would change the rules of the cycle, and the relative optimisation of different player types for harnessing energy categories etc.

This sort of thing could be applied to any element of the game where players interact with the world’s parameters (ecologies, the atmosphere, terrain types etc), but could also be used for subjective population behaviours - e.g. NPCs attitudes evolve in response to player attitudes/actions (a “hostility/placatory cycle” instead of an energy cycle).

Anyway… back to work - over to the design experts!


With Elrakis I try to stick to the following mantra

Everything has a reason and everything has a consequence

For me, this means that everything a player should be able to do has some sort of influence; be it their personal situation (progression) or in the grand scheme of things (power balances, etc). The most difficult part here is when it comes down to economics since it is difficult to be able to reward players for their actions without ‘magically’ creating an amount of ‘currency’ to provide them.

A concrete example is resource harvesting; if you want your economy to be sustainable then resource nodes should deplete and not respawn, or the respawn rate should be equal to the resource destruction pace. But this can cause areas to become crowded when there are resources there worth harvesting, or areas becoming desolate when they are completely harvested.

Power balance is somewhat easier since it is non-tangible and cannot be possessed by the player. This means that whenever a faction gains power that another automatically loses it, keeping the balance.

I hope this kind of makes sense and feels applicable to the discussion; I found it hard to properly write down my thoughts on this topic.