Inspecting level design through 'Loops'
Howser
UK Join Date: 2010-02-08 Member: 70488Members, NS2 Playtester, Squad Five Blue, Reinforced - Shadow, WC 2013 - Shadow, Subnautica Playtester, Retired Community Developer
UK Join Date: 2010-02-08 Member: 70488Members, NS2 Playtester, Squad Five Blue, Reinforced - Shadow, WC 2013 - Shadow, Subnautica Playtester, Retired Community Developer
Howdy old mapping chums... This probably will not be helpful for the experience NS2 mappers, but I thought I would share it for my fellow noobies. Not sure how helpful this will be but its good to share!
For the past few months I've been trying to learn how to design a map for NS2. I like to consider myself a hobbyist environment artist and as such I only really wanted to just assist mappers with random assets they need. However, the more I played ns2 maps - both custom and official, the more I became interested in designing a map of my own. I'm not a level designer, and haven't really got much experience in this area so I think I've hit every pitfall along the way (and I'm still learning).
When I started drawing and planning my map the first few floor plans were crazy, far too complex, you would probably struggle to find your teammates let alone be able to defend all your bases from sneaky bastards rushing down some obscure vent or corridor. I had to go back and study the layouts of maps that worked. Breaking down the maps into Resource nodes, tech points and their connections (in the standard way) didn't really help me- it just confused me more.
Maybe I'm a bit slow, but breaking the map down into rings or loops really helped me understand how maps should flow, and how complex the connections and routes should be.
This is a very crude, oversimplified way of inspecting the design of a map, but for me, it really illustrated how complex a map should be in order to have the right 'flow'. Here's some examples:
Summit:
Veil:
Tram:
Docking:
Descent:
Caged:
Mineshaft:
Refinery:
The simplicity of Tram and caged (my personal favourites) is very interesting and something I'm going to be applying to my own designs!
NS2 must be one of the most difficult games to design levels for! in my opinion a map should provide a space for both interesting tactical choices for teams and varied, fun combat spaces where players will be involved in regular bouts of combat. Achieving this in a balanced and fun way seems to be down more to getting the connectivity and the scale of the map right than anything else. I'm stating the obvious I know, but I play a lot of custom maps and over/under sized and over/under complex maps (or sections of a map) are most frequently what kills a map. While checking a design with this method won't help with the scale, perhaps it'll highlight potential problems with the routes and complexity of the map.
Has anyone used this method before or have some alternative methods of dissecting and inspecting existing maps?
Thanks
EDIT: Added the other official maps.
For the past few months I've been trying to learn how to design a map for NS2. I like to consider myself a hobbyist environment artist and as such I only really wanted to just assist mappers with random assets they need. However, the more I played ns2 maps - both custom and official, the more I became interested in designing a map of my own. I'm not a level designer, and haven't really got much experience in this area so I think I've hit every pitfall along the way (and I'm still learning).
When I started drawing and planning my map the first few floor plans were crazy, far too complex, you would probably struggle to find your teammates let alone be able to defend all your bases from sneaky bastards rushing down some obscure vent or corridor. I had to go back and study the layouts of maps that worked. Breaking down the maps into Resource nodes, tech points and their connections (in the standard way) didn't really help me- it just confused me more.
Maybe I'm a bit slow, but breaking the map down into rings or loops really helped me understand how maps should flow, and how complex the connections and routes should be.
This is a very crude, oversimplified way of inspecting the design of a map, but for me, it really illustrated how complex a map should be in order to have the right 'flow'. Here's some examples:
Summit:
Veil:
Tram:
Docking:
Descent:
Caged:
Mineshaft:
Refinery:
The simplicity of Tram and caged (my personal favourites) is very interesting and something I'm going to be applying to my own designs!
NS2 must be one of the most difficult games to design levels for! in my opinion a map should provide a space for both interesting tactical choices for teams and varied, fun combat spaces where players will be involved in regular bouts of combat. Achieving this in a balanced and fun way seems to be down more to getting the connectivity and the scale of the map right than anything else. I'm stating the obvious I know, but I play a lot of custom maps and over/under sized and over/under complex maps (or sections of a map) are most frequently what kills a map. While checking a design with this method won't help with the scale, perhaps it'll highlight potential problems with the routes and complexity of the map.
Has anyone used this method before or have some alternative methods of dissecting and inspecting existing maps?
Thanks
EDIT: Added the other official maps.
Comments
https://developer.valvesoftware.com/wiki/Loops_(level_design)
The reason for loops in multiplayer games is also simple. It is the best way to get mobility in a map. How does on get around a checkpoint? Go around in some way. The only way that works, loops in some manner. It also gives stealth options and the like. It also keeps the map size small and compact. Even in push maps, like 5 control point maps in TF2, one needs loops to get around people.
Loops is all apart of mobility.
I'm curious what did looking for loops tell you about these maps? Any particular observations/insights revealed?
"Verily, upon Satan's butthole" rears its head again. That will become its official name at any rate.
At first glance I didnt understood the complexity, but now I am starting to see the pattern. Did you concluded any universal observation from this metod usefull for future designs?
Seems, that most balanced maps like Tram and Caged are fairly simple, consisting mostly from primary loops. On contrary, Veil looks like one huge loop-mess. How is Mineshaft standing?
EDIT: Another observation, it seems that managing the loops as close to circle as possible is beneficial for balanced gameplay. Also loops have to intersect at reasonable spots to allow players to change loops while navigating the level.
I haven't played with it myself, but it could definitely be used for analyzing and designing NS2 maps.
Here's his Gamasutra article on it:
The Metrics of Space: Molecule Design
Rudy nailed it on the head there but i'll add this:
What surprised me is how simple some seemingly complex maps are. (PERSONAL OPINION TIME) In caged you never feel like you're tactically limited- there's always ways to sneak off or flank enemies, but at the same time it offers quite fasted paced and regular combat. Same can be said of Tram. Also I think it's a nice way to identify points of contrast throughout the map, which might be a good thing but I feel areas with small dense loops don't work so well - IE center area of descent, maintenance access in summit or the pipe hive in veil.
Adding mineshaft (old layout) I noticed it actually becomes quite hard to apply this theory, however with a rough go it does seem to suggest why mineshaft is a complete pain for holding RT's in the middle section. I think this might also suggest why pubs very rarely request mineshaft, its just too complex. But how does it perform on the competitive scene?
I didn't really want to talk about my own map but its interesting so I'll share the very first layout (which was a complete bloody mess) alongside the latest which I'm starting to feel confident about. It's still a few weeks off the first playable build but the main body of the layout wont change much from then. I feel it now holds up to this design theory.
Very first floor plan:
Latest:
Also Mouse : If you get the time could you please break down a NS2 map with this design theory... or get your friend to do it! It does look very interesting by it'd be nice to see it in the context of a map we understand.
New
vs old
Mine shaft is very complex compared to the others. the north east half of the map is the problem the other side seems to be very balanced. What's the competitive consensus on the map?
Kouji_San : it seems players- especially pubs, shy away from the more complex areas that don't read well. There's a lot of loops running through that section compared to the rest of the map. More overlapping loops (I would think) would mean more traffic but instead the opposite seems to be true. or it could be the door! maybe both its funny how little of the maps traffic flows through that part.
Here is my spline-loops interpretation
I think the quality here should be judged from how much loops does touch each other, thus allowing migration from one loop to another.
The other thing to consider with the circuits you're drawing is the flow through them, which isn't represented in the circle diagram - look at the North Tunnel area in Mineshaft, for example. That |\ shape (forgive my crudeness) is an extremely awkward angle to move through, and the type of structure that's liable to get people turned around or confused. One of the things I did in Eclipse and Veil that I've kept as a personal rule ever since was avoiding > 90 degree junctions in layouts. Natural player progression is forward - you want players to come to those junctions and be able to clearly read their options on where to go next without having to move through hairpin turns (and if you want a hairpin turn in a level, it should be something clearly readable like a switchback staircase/ramp or a wider loop that curves or is made up of smaller angles a la South Loop in Eclipse).
Also i would argue that you could actually add two additional loops into cavern/cave/north tunnel part. Although the routes and pretty much redundant, I do like to sneak through them as a marine and its always quite easy. the crusher machine 'vent' is not a main route but the fact marines can easily use it adds to the whole complexity of the northern half of the map.
KungFuSquirrel : It is just that, a very simple at-a-glance check of flow and connectivity, It overlooks loads of other elements which define the map! It really shouldn't replace established design methods, but serve as a quick check. its interesting to hear how other people interpret the diagrams. Could you elaborate on the what you perceive as deficiencies of veils design? It seems to me the EAST and WEST sides pose a nice contrast to each other.
need help offer
hallo ich habe ein problem ich habe eine map erstellt und die commander kamara ist auf der map alles so wie im youtube beschieben gemacht aber so balt man in der command station ist kommt man mit der ansicht nicht raus
brauche hilfe bie
EDIT: pretty off-topic, I know, but the guy needs help! :P
You are probably new to this type of forums. It is not a chat, please start your own topic or use personal messages.
Ich denke, dass du neu zu diesen typ des forums ist, hier nicht ein "chat" ist. Bitte, start ein neues thema (topic) oder schick eine persönliche Nachrichte (personal messages).
And here's my own layout.
Rudy & Lokee have used them (somewhat arbitrarily) to identify smaller loops, most of which pass through narrow/small corridors. These diagrams also ignore vents. Panademic has instead used yellow loops them to identify loops that pass through vents (in his first overview at the very least).
In the interests of building something resembling a standard language for spline-loop diagrams, how do you guys feel we should use yellow loops?
Personally, I'm torn. I do think we need to somehow acknowledge vents, but there can definitely be 'secondary' paths through a level.
This isn't a proven method just my assumption on how to make designing maps more manageable in the early days. Do you guys draw them in from the start or put them in later- like me, to counter any balancing issues?
I personally think looping the vents defeats the main idea behind doing this- having a at-a-glance method of analysing the connectivity and complexity of a map. For me it just doesn't give the same clarity although it is interesting.
I think Rudy's method works best, using the yellow just makes it easier to read and smaller areas, but like you say defining what routes are secondary is difficult. Personally i'd just introduce the yellow to keep it readable when things get tight.
One thing I did notice is that the yellow splines often only link two or less 'important' locations (TP, RT, DBL RT). I also wondered if it might be worth trying to 'weight' the splines with variable thickness according to the strength of those locations, could help with clarity. I may mess about with this a bit later on.
(Edited for bad spelling, doh!)
But we can continue the splines idea.
For the methodology - I think each loop should start from a tech-point and continue to nearest expansion(s) and then try to find the closest way back around, following the main routes. The yellow loops indicate secondary routes suitable for flanking. I think every loop should indicate (somehow) how many RT's and TP's connects. I am not for including the vents in the loop analysis. First I think that vents are mainly helpers for Aliens to maintain the balance vs Marines and they long range weaponry. Second, the loop-metod should focus on analysis of the high-level gameplay flow.
But drawing loops around is nice, but why should it be of concern? What to look for and what it tells us?
My (!) interpretation:
For example - in the old Mineshaft there are loops that connects only two TP's (Cavern-Ore; Crushing-Or) and one connecting 2xTP and RT (Operation-Drill-Gap). From gameplay point of view this looks like an imbalance in design. Also not all the loops meet in central area. The traverse from MS loop to cave loop have to be done trough two secondary loops, what could reflect the situation, that navigating trough this area some times leads to taking the wrong turn.
Ideally adjoining loops should intersect at RT, which will the subject of combat and for the winning side becomes foundation for further expansion.
But maybe I am totally wrong.
Improving the amount the data the loops carry should be the aim; by showing things like the connection count and points to other loops, what nodes (rt's, tps') they overlap, the size of the corridors and rooms they overlap and the weight of their flow (secondary/primary).
However I feel if we lose clarity or the 'at-a-glance' feedback I think we're taking it in the wrong direction. As such Personally I'm going to be experimenting with only colour and thickness of lines, Might add some Dots of colour. But I won't be adding icons lines connecting stuff or anything else that'll complicate the simplicity of the loop-diagrams.
But like i said there's no rules at this point we need to all pursue our own ideas then converge with what works.
Hopefully we can develop this into a nice standardised, documented, level design aid for ns2 maps.
I am still interested in the value of this has for other MP games, but personally I am going to try to develop all ideas around NS2 level-design.
I'm not sure what everyone else's workflow is like, but I tend to create a rough layout and then begin creating rooms one by one - not always with a particular location for them in mind.
As a first step, the initial layout could be overlaid with a simple loop diagram - perhaps one that is only comprised of red loops.
Then once the individual rooms are being made, keep an eye on how the secondary loops are forming within it and how the red loops segments are passing through it. The secondary loops should give you a rough idea of the complexity of that section and the red loop segments should give you an idea of where it can fit within the overall layout.
@KungFuSquirrel Could you share the diagram you made for veil and point out these "huge deficiencies" that you mentioned in your post?
Nice stuff.