Gazz wrote:While I think that it would be cool to have some 3D in that map, I wonder if it would add any useful gameplay.


If it's 2D you have a clear and obvious relation between jumpgate travel (follow the road...) and jumpdrive travel (go off-road).
You can judge distances by eye without having to rotate the map every which way.

I feel that all that 3D would add here is a chore. In 2D you have all the relevant information at once. I'd like that.

Apologies to all who are in the pure-3D camp, but this describes perfectly how I feel. As a person who is strongly driven by the spatial relations in a game, and liking to feel that the game space has some interesting structure, I find it overwhelming to be in a 3D connected structure. Hard to explain why, but I know that I wouldn't like the feeling of trying to understand a 3D connectivity structure. Another one of those strange personal preferences, I'm afraid.... With an almost-planar system layout and an almost-planar galaxy layout, I feel that I can really come to a strong understanding of the structure of a region, and this gives me a really nice sense of "personality" in the areas - something that would be lost on me if it were more complicated. It's the same reason I don't want dynamic celestial bodies. That sense of structure is just too important to me

That being said, the final star map will be 3D, and there will be some slight variations in the depth offset of certain systems, but not enough to prevent you from being able to think of it as a 2D map in your head.

JoshParnell wrote:there will be some slight variations in the depth offset of certain systems

There are several ways to visualise the depth of a system.

• Stalks like on the Elite 1 RADAR display.
  You all know what I mean.
• A ring around each system.
  The "plane" is at zero depth and there are no systems below it.
  The higher a system is above the plane, the wider the ring around it.
  
  Player's current location is shown as a "large crosshair cursor". A screen filling cross on the X / Y axis like you see in modeling / graphics software.
• Size of the system circle / sun instead of a ring.
  
  Greater danger of overlap and you get a better sense of scale with rings because each circle / ring set has the circle at a constant size.

With rings you can lock the view to full top-down.
Judging stalks' height requires an isometric view, which isn't so good for judging distances as a whole. Or a rotating map which, while cool, can be confusing as hell.

I wasn't planning on saying more on this subject, but hey, if Gazz feels it's OK....

My objection to flattening universe space is summed up in these points:

1. The "what gameplay does making the universe three-dimensional add?" question is a fair one, but it's incomplete in a couple of ways. The first way is simple: what gameplay is added by restricting the universe to two dimensions? If it's fair to ask proponents of a 3D universe to try to explain how their preference demonstrably adds specific gameplay features, it's fair to ask the proponents of a 2D universe the same question.

2. The question "what gameplay does making the universe three-dimensional add?" is also incomplete in that "a game" is more than just pure mechanics -- the What Do You Do Now player actions.

A game -- and in particular, a game that creates a world, such as Limit Theory appears to be -- also needs rules that determine how the world changes, both based on player actions and on simulated elements in the gameworld. Those dynamic features make the world feel alive, and make it a more interesting place to play in. This is important for the gamers who are less interested in "beating" a game than in experiencing it as a place to hang out and enjoy poking things to see what happens.

In addition to mechanics and dynamics, a good game that is a world also needs satisfying aesthetics. It needs to look right for the stories that can be told there... and it needs to be a place that can generate stories. A world-game needs lore that explains why all the places and objects are as they are.

A 3-D universe may not directly generate mechanical gameplay any more than a restriction to two dimensions does... but a universe in three dimensions makes more aesthetic sense, and possibly more dynamic sense, than 2D universe.

3. Flattening universe space to two dimensions is inconsistent with system space being 3D. Plenty of space combat games establish a plane in system space and automatically roll your ship to that plane. Limit Theory could do this as well... but it doesn't. If gameplay in system space asks players to perceive and act on the positions -- and movements! -- of objects in a three-dimensional volume, it is no worse, and actually somewhat easier, to ask them to understand position only in universe space.

This isn't to continue to argue a decided question, but to provide the other perspective on this question. I look forward to Limit Theory in either case.

And now that I have a couple more minutes, let me also try to be constructive.

The "2D/3D universe" question is actually two questions:

1. Does the universe have a flat 2D structure, or a volumetric 3D structure?
2. If the universe is 3D, how can that information be presented to players in a way that's easy to manage?

It sounds like there will be some 3D-ness to the structure of the universe (at least in part). So this really comes down to how to let players see that structure in a way that most of them will be able to easily understand and manipulate.

Gazz offered some suggestions, but I wonder whether most of them actually trade one kind of complexity for another. Using additional abstract information (e.g., ring size) to stand in for height is not necessarily easier to understand than showing the actual 3D relationship of star systems to each other.

A frequent objection to presenting 3D system connections as an EVE-like web of points is that it's hard for some people to work with that structure. The combination of scaling the web (zooming in and out) and rotating it around the current zoom point, is described as a real problem for some gamers.

I'm not about to argue that it isn't. But instead of using that to argue that the universe shouldn't be 3D at all, why not focus on improving the presentation by simplifying the web model?

One improvement might be to limit the number of star systems displayed at a time. Josh has showed us a simple generation of a group of something like 20-25 star systems. Why not make that the standard? Say that groups of 10-20 (on average) star systems constitute a "sector", possibly with some characterstic in common so that sectors are easily distinguished from each other in the memory of players. This way players are only dealing with a manageably small number of star systems at a time.

(Note: With my gamer hat on, this isn't the approach I like best. Like I've said before, I much prefer a continuous universe to one broken up into discrete points. But I can compromise. In fact, I think it would be OK for this universe map to be "flat," given that it's a closer analog to the galactic plane than individual star systems.)

The "universe as a collection of connected small sectors" model suggests a second improvement over the EVE-style map, which is to not implement scaling at all. When you're trying to grok the connections among hundreds of points, it's nearly mandatory to be able to zoom in and out to see smaller parts of the web at one time. If you limit the number of points to no more than 25 or so, however, they can all be displayed at one time (with some judicious map dragging) with no zooming necessary. That simplification of controls would, I think, go a fair way toward making this kind of map more usable by people who've found other 3D maps cumbersome.

So I suggest:

1. Define the universe as a flat web of connected star sectors.
2. Define each star sector as a 3D web of 1-25 connected star systems (with 10-20 being the average).
3. Let 1-4 of the star systems in a sector have outbound links to other sectors.
4. Display the universe map as a plane of sectors, allowing zoom-in/zoom-out to areas of the map to show about 10 sectors.
5. In the universe map: left-clicking a sector brings up the sector map showing the star systems in that sector; right-clicking a sector displays information about that sector (number of star systems, total population, primary political "owner" of star systems, etc.).
6. Display the sector map as a 3D network of star systems, allowing click-and-drag to rotate the web around the (predefined) center point of that sector.
7. In the sector map: left-clicking a star system sets it as the destination for the player's currently designated ship/fleet; right-clicking a star system displays information about that system (star type, number and types of planets, system ownership status, "specials," etc.).
8. Both the universe map and each sector map will have "information" modes that allow players to a) filter for particular characteristics (location of player ships/fleets, high-population worlds, metal-rich worlds, cargo destinations, mission targets, uninhabited worlds, unknown/unexplored worlds, etc.) and b) visually highlight sectors/systems meeting filter criteria.

I've been on the wrong side of several LT design plans already, so I won't be surprised (or offended) if this suggestion is also rejected.

But I did at least want to put it out there for consideration, if for no other reason than to maybe help spark some additional thinking about how the structure of the LT universe will affect not just mechanical but dynamic and aesthetic aspects of the game. All those modes of play matter -- not in the same degree to every individual player, but in some degree to all players.

Nobody said building a universe would be easy.

My opinion can be summed up by reality... I have never seen a map (not a globe!) that puts Australia on the bottom of Europe.
It's natural to make a flat map, because that's the only way we humans can easily understand what we're seeing. A representation like a globe is nice to look at, but very impractical to plan a voyage with in great detail. Not to mention you'd need Pi a lot to convert your distances.

If "more 3D" is required then I agree that a 2D area of 3D star clusters would be the best solution.
It prevents the deep 3D stacking, that would overlay your systems of interest with a dense jungle of connection lines when you zoom out.

Katorone wrote:My opinion can be summed up by reality... I have never seen a map (not a globe!) that puts Australia on the bottom of Europe.

I, however, have not see Australia at the same time than Europe from pictures from space. Probably because it was behind... Because the universe is 3D. Maps are 2D projections, simplifications. There are several ways to realize projections, for a computer it is a trivial task to make a 2D projection of a 3D universe.

The question is therefore not whether we may need a 2D map ( we can have it anyway) but whether a 3D universe has an added value - large enough to justify the somewhat more complex resulting 2D projections. For me it does - because of the difficulty to visualize it, it gives a better feel of scale. It makes it more difficult to cut parts of the universe in quasi self contained sections, thus being more realistic ( try to make a blockade in 3D). In a word, it more immersive, the map is more a continuation of the 3D in-system play.

But does it add features significantly different from a 2D universe? Probably not.

imo, 3D stars maps look much more beautiful than plain ones. Sure 2D ones are easier to use, but they don't look... right. I agree about Clusterf**k looking maps but I think thats not what we, 3D map lovers, are after. For example, look at this map:
THAT is what I call a Clusterf**k. Thats not the kind of 3D map im talking about.

Alright, that being said, let's take a plain and simple flat map. That looks boring right? Sure it does! But what if we tweak the z-axis of these object a little bit?
The differences are subtle, but add that and the ability to rotate the map view and you probably have the best of both worlds (as long a there is a "reset view" button of course!).

Edit: Sorry for that last pic but that's the best example I found out there.
Ze last image link: http://media.pcgamer.com/files/2012/06/10.jpg

Swordmania wrote: Alright, that being said, let's take a plain and simple flat map.

That looks boring right? Sure it does! But what if we tweak the z-axis of these object a little bit?

The reason your 2D map example looks boring is not because it's flat but because it's symmetrical...

Other 2D maps like adapted ones in Eve: http://i.imgur.com/Kj57r.png
, or like the pictures Josh posted from universe generation, does not look boring at all.

I rather liked the map from Frontier: Elite 2. It was 3d, but pretty damn flat - you could scroll on the x and y axes but only zoom on the z access, and it had stalks to show how far each star was from a flat plane. Comically low resolution picture below!

Ixos wrote:

Swordmania wrote: Alright, that being said, let's take a plain and simple flat map.

That looks boring right? Sure it does! But what if we tweak the z-axis of these object a little bit?

The reason your 2D map example looks boring is not because it's flat but because it's symmetrical...

Other 2D maps like adapted ones in Eve: http://i.imgur.com/Kj57r.png
, or like the pictures Josh posted from universe generation, does not look boring at all.

Symmetry or not, you can't deny that 3D maps look better tho. Hell look at Freelancer's Map There. No symmetry. And no matter how much I loved this game, the map lacked something special. What would be best is some sort of EVE-like map interface. If I recall correctly, you could flatten the map just by pressing a button. Boosh. Everyone's happy.

Look, I don't care THAT much in the end. I will play the game anyway. That's just a little something that , I think, would make LT stand apart even more.

Swordmania wrote:If I recall correctly, you could flatten the map just by pressing a button. Boosh. Everyone's happy.

Look, I don't care THAT much in the end. I will play the game anyway.

I agree with this. Given Josh's evident coding proficiency, it surely can't be beyond his wit to make the map 3D but then allow a 2D projection (or a camera lock, or whatever). It's the kind of extra detail that just makes a game feel premium, though in the end isn't really a deal breaker for anyone.

I also think this is a pretty deep rabbit hole that we haven't really discussed that much beyond the 2D/3D question. For instance, how many map levels are there? For instance, is it broken into three:

• System map
• Region map
• Universe map

Or is it just one, endlessly scalable, map?

It's also possible that the 2D/3D question is more than just aesthetic. It could play heavily into exploration. In my head, for instance, I can see a scenario in which the player starts in a well explored region that s/he gets a full map for. As soon as they leave the origin region, however, the first thing they may do in the new location is to dock into a station/planet/moon/etc. and buy a map of the region - and this map may not be complete. How would you choose where to explore when looking for a connection between systems that is undiscovered? A 3D map could reveal proximity that would be hidden by a 2D map (or the reverse: two systems aren't in fact close together but are considerably vertically separated ("Z minus 10000 m!").

Spores 3D maps.

(close up) (medium zoom) (zoomed all the way out)

Swordmania wrote:[
There. No symmetry. And no matter how much I loved this game, the map lacked something special.

Uhm no... that map got LOTs of symmetry (even if it's not 100% symmetric).
The entire center is symmetric around the vertical middle axis.

And it's also based on a grid with a majority of connections being straight or 45 degrees, which is a great way to make a boring map.

Something like this might be a good comprimise between a 3d and a 2d map...