What makes a level non-linear?

[flubbernugget]

I have been mapping recently and realized that unlike the original doom levels that were always in a maze-like non-linear fashion, my wad's always end up being straightforward in a very linear path. Are there any ways to help fix this, or will it just take a lot of practice?

[Khorus]

well. simple really. rather than making a straightforward map, create multiple entries to rooms and areas to explore.

[EarthQuake]

Well, a really good suggestion would be to make the area look less linear by creating areas that flow back onto the main route, either with one-way doors or drop-downs. Make some areas visible, but not immediately accessible. Make routes that are completely optional, say for like a weapon or some smaller items. Hell, if you really want to go all-out, make the player make decisions on where to go, or let the players decide the order in which to complete the level.

There's lots of things you can do, you just have to put your imagination to work.

[David_Dweedle]

flubbernugget said:

I have been mapping recently and realized that unlike the original doom levels that were always in a maze-like non-linear fashion, my wad's always end up being straightforward in a very linear path. Are there any ways to help fix this, or will it just take a lot of practice?

this is also one of my major problems when it comes to mapping. All I can say is play as much Pwads as you can and see how other people make things non-linear and take those ideas and expand upon them.

A good example is on E1M2 with the computer areas as it is a big space but it goes in a big circle BUT is off the main path giving the player a large optional area to explore.

[Enjay]

Yes, I suppose it's all about giving the player options. A doom map, ultimately, usually comes down to getting from the P1 start to whatever triggers the end. Giving the player choices of routes and strategies along the way is what reduces the linearity of that situation.

Of course, if you really wanted to get cunning, you could provide additional choices such as more than one exit, or teleporting the player to different start locations.

[Torn]

A really good thing is also to create windows to the other areas, it makes a nicer 3d feeling (heh, I stolen your oneliner, essel :P).

[StoneFrog]

I often have trouble making a level non-linear. Actually, scratch that. I often have trouble making the non-linear areas of my level worth visiting and cohesive. Many of them are rather "out of the way", with the linear path to the end of the level being quite straightforward.

As a result, whilst the map has plenty of side passages, only the person really intent on seeking the goodies within will look for them.

One technique I find is good for conceptualizing side paths is, as previously stated, making your rooms look out into other, inaccessible ones. For instance, the Heretic map I'm working on has a bunch of large support beams in the main hallway, but there is a crawlspace above and behind them - with grates allowing you to see this passageway between the pillars.

That's probably the best way to do it. Just start combining your main level geometry with other interesting, albeit inaccessible areas. From there you can start thinking of ways to combine them together.

[Super Jamie]

StoneFrog said:

As a result, whilst the map has plenty of side passages, only the person really intent on seeking the goodies within will look for them.

You pretty much just summed up E1M5 to E3M6 in one paragraph. Except for the obvious boss maps, decent portions of most maps are completely skippable if you know what you're doing.

[40oz]

A lot of people tend to frown on linear maps but I like them. Not to say that I like them opposed to non-linear maps. I think it's an important quality of mapping be able to create a longer and more memorable map while using less rooms. But I also like straightforward maps where I'm simply flying through the map until I get to the exit.

But it's easy to detect how your map is too linear. Utilize more switches that activate doors. Make rooms with more than 2 hallways connected to it.

Something that might also come in handy, when you are mapping, instead of making the starting room, and then every corresponding room leading to the exit, maybe you should start mapping from the middle of the map by creating a "main room" that the player has to get to, and make a lot of optional entrances and exits in and out of it. Mark some of the doors with keycard requirements, and you're set.

[newt]

I recall seeing some fervent debate once on just what "nonlinear" actually means.

The way I see it, the original Doom levels were linear. The original MAP02 and E1M7, for example, may have been very nonlinear by design, but the formula stays the same: You have to get the blue key to get the red key to get the yellow key to get to the exit--or some variation thereof. It doesn't matter if Doom II MAP11 is a big open circle, because no matter how many times you play it, it always progresses the same. It's designed that way.

The idea of nonlinearity, to me, is that you can go in whichever direction you want in whichever order you want to get to the end. There can either be multiple paths to a common end, or multiple paths to different ends (much more difficult to accomplish in a sensical way if you're mapping Vanilla), or a series of areas that each must be visited in any order before the end can be accessed.

A few maps from Action Doom are good examples of the first; there's a fork in the path and you can go either way, but they each lead to the same end.

Likewise, Action Doom 2 as a whole is a great example of the second, though on a much more complex level than what you might need. There are several paths you can take and the way you play through it determines the ending.

Vader's Blackrock from ZPack is a perfect example of the latter, and the model I used for creating MAP23 ("Magma") of NDCP2. There's a single goal, but you can approach it in any order you please.

But, I wouldn't stress too much about making everything nonlinear. If you can switch up the linear get-key-to-get-the-next-key-also-doorswitches formula, you can make a perfectly good linear map. As has been said:

-Have areas looking into other places that can't be accessed until later
-Have the map progress in such a way that certain areas are revisited (and utilize teleporting monsters or traps to prevent revisiting from being boring)
-Include non-essential areas for exploration, ammo, health, bonuses, secrets, 100% kills/items, etc.
-Use loops from time to time--in moderation, mind you.

And I know the examples of nonlinearity I gave are all really complex (and all ZDoom, heh), but should you attempt to make a truly nonlinear map, those ideas can still be applied by simpler methods. (Granted, multiple endings doesn't really make sense in Vanilla, but the other two work just fine.)

[Gez]

If you're a flowchart-and-diagram kind of guy, you can present your map in that way. Let's take MAP01: Entryway and MAP02: Underhalls as an example:



This is a very linear level. There is only one way to the end, and you have to jump from the ledge, go through the corridor, then in the big room and from there the exit room. You can take detours to visit optional and secret areas, but it's absolutely not needed.

MAP02 is pretty linear too, but it is a bit less so:


You can see that while it's still pretty linear, you have choices of different routes in places. Those choices are pretty minor and don't really change the flow of the level, but they're here.

Now those two examples are in very tiny maps.

Let's imagine a map where you need to get two keys to reach the exit. Both the blue and the red key are available from the start through one-way teleporters. If you go get the blue key, you'll also find a chaingun. If you go get the red key, you'll also get a shotgun. Once you have either key, you can return to the start area, and once you have both you can accede the exit room. This is the blueprint for a really non-linear level. Suppose you play from a pistol start, your choice of which area to go through first will also determine with which weapon you'll play the rest of the level.

So, non-linearity:
- Multiple orders in which the level can be completed
- Multiple paths from point A to point B (and multiple return paths to A)
- A different flow depending on choices made

[PhilibusMo]

newt said:

-Have areas looking into other places that can't be accessed until later
-Have the map progress in such a way that certain areas are revisited (and utilize teleporting monsters or traps to prevent revisiting from being boring)
-Include non-essential areas for exploration, ammo, health, bonuses, secrets, 100% kills/items, etc.
-Use loops from time to time--in moderation, mind you.

All those are excellent things ot have in levels, they can make an actually quite compact and linear level feel like a vast and very real area as in real life there is very rarely only one way in which you could proceed to reach your intended destination, and so it should be in games if you wish to retain realism.

[Creaphis]

PhilibusMo said:

All those are excellent things ot have in levels, they can make an actually quite compact and linear level feel like a vast and very real area as in real life there is very rarely only one way in which you could proceed to reach your intended destination, and so it should be in games if you wish to retain realism.

Indeed. Modern shooters are a tad schizophrenic in their approach to realism, as while the graphics aim to be ever-closer to believably real, the layouts are as far from realistic as ever. I mean, hypothetically, suppose you're on a martian base that suddenly becomes invaded by hellspawn, and you need to assail certain key objectives in certain locations quickly. You'd prefer not to die, if feasible. Would you:

1. Find yourself in a large and branching complex where every wrong move could get you killed, and therefore you are forced to intelligently choose the safest, fastest and most secret routes, and avoid side treks at all times, unless you need to risk extra travel to reach areas which would logically contain useful resources such as security offices and infirmaries?

Or

2. Find yourself in a complex, where, coincidentally, only one avenue of movement is ever possible, except for where small, dead-end side paths exist, which you are rewarded for traveling on, even though this is technically a waste of valuable time, because these side paths always contain cashes of ammo and health even when they're highly illogical places for such resources to be? Oh yeah, and there are PDAs out the wazoo.

That said, I like linearity, when the series of sequential challenges is involving enough, when the game art is immersive enough, when the straight-forward cinematic story is told skillfully enough. However, I learned while playing the original Thief games that using your brain to choose your own path and your own tactics, then executing your own plans successfully, gives you a good feeling. I know of no modern games that give you this feeling. So, I'm going to say there are two ways to do it:

1. Make a linear gaming experience, that's really good. That means providing a great deal of variety in encounters, and careful balance and pacing, among other things. I played Portal for the first time yesterday, and for the second time, yesterday, with director's commentary. That's what good linearity looks like. The commentary makes me marvel at how much thought and testing went into every individual aspect of the pacing and the environment so that the players would be guided through the puzzles while still being allowed to feel like they solved it themselves.

2. Make a non-linear gaming experience, that's really good. That means, for me, creating an environment where some areas are optional, and perhaps where some areas are actually best avoided altogether. The environment should contain logical cues so that the player can guess at where to look first for supplies and what route would be safest while traveling to some further objective. Among other things. Of course, non-linear levels in Doom typically have an equal difficulty level across all possible paths so that the player is safe regardless of where he goes and these levels are also designed so every single monster is the level can be killed with no great strain as opposed to there being rooms best avoided and there's never any clear destination objective in these levels because an exit room can be tucked away anywhere, so you can follow that formula if you want to I guess. What are we talking about again?

[Jodwin]

1. Draw a bunch of rooms.


2. Connect them with hallways (and other rooms).


3. Add locked doors and keys in a manner which gives the players choices for routes.


Or, look at other maps for examples. Just keep in mind that sometimes you might end up taking nonlinearity too far, which isn't always a good thing either. :P

[myk]

Creaphis said:
Modern shooters are a tad schizophrenic in their approach to realism, as while the graphics aim to be ever-closer to believably real, the layouts are as far from realistic as ever.

Heh, I'd say that's through the paradoxes of technology. The technology that allows realistic visuals, effects and complex game behavior needs so much work that scenario layouts have to be made simpler or stupider in some respects, both because you don't want intelligent AI doing something stupid in an environment that outsmarts it and because the art and design become expensive and time consuming tasks. You can also give the players more options, but then you have to check that each of those options will actually work and not make a mess or require restrictions that look very arbitrary. Like you can do all this weird shit like talk to people and type in consoles but if you shoot your rocket launcher at your commander's computers or directly at him, all that might happen is that he might say "ouch" if you hit him and then "the people are waiting, champion, head on to glory!" just to ensure the "plot" continues as required and to save the devs from adding side effects.

[Graf Zahl]

Lesson of the day:

Games with overly restrictive plot lines suck. Nothing new here...

[Creaphis]

myk said:

Heh, I'd say that's through the paradoxes of technology. The technology that allows realistic visuals, effects and complex game behavior needs so much work that scenario layouts have to be made simpler or stupider in some respects, both because you don't want intelligent AI doing something stupid in an environment that outsmarts it and because the art and design become expensive and time consuming tasks. You can also give the players more options, but then you have to check that each of those options will actually work and not make a mess or require restrictions that look very arbitrary. Like you can do all this weird shit like talk to people and type in consoles but if you shoot your rocket launcher at your commander's computers or directly at him, all that might happen is that he might say "ouch" if you hit him and then "the people are waiting, champion, head on to glory!" just to ensure the "plot" continues as required and to save the devs from adding side effects.

I observe that some retro-styled games with much lower tech requirements are starting appear again, to positive response, with Megaman 9 being a prime example. I'm tempted to predict that, at some future date, the genre of anachronistically simple games will rediscover the year 2000 and produce 3D games with better development priorities.

[Jodwin]

Or we'll end up in a situation where game engines are made by one kind of companies, art assets by an other kind and then you have the actual game devs buying engines and assets from these companies, so that they can dedicate their own time on gameplay rather than tech or art...

Yeah, one can always hope.

[Creaphis]

Huh? At first I was under the impression that you intended for that to be an image of a games industry even more dystopian than today's. But you're saying that would be a good thing? If these first two sorts of companies are selling their engines and assets to developers at a profit, doesn't that leave even less monetary momentum for the production of content? Wouldn't such a corporate introduce severe disconnects between engine, art and game design, making it impossible for any products to emerge with functional and aesthetic coherence?

[myk]

Graf Zahl said:
Games with overly restrictive plot lines suck. Nothing new here...

I mentioned plots but that was just part of my point, as the complexity I was talking about goes well beyond character interaction and the inclusion of a story. A very invasive plot, on the other hand, would work in a genre where that is the focus, but not really in an action game.

Creaphis said:
I observe that some retro-styled games with much lower tech requirements are starting appear again,

Those will always be around. Sometimes they'll make their way to more marketable positions, depending on the situation. The development of cellphones and hand-helds has given them additional value, for example.

Mine wasn't so much a complaint at today's games, but rather an issue technology has always had and will never be able to avoid because it's a cultural artifact of practical use. Good games that stick to their main assets intelligently, such as DOOM, and properly fit what the technology can do comfortably, will end up being relatively strong. If not, the technology will still allow the author to follow an idea, and will encourage certain design directions once it's put in motion, but only to help make a bunch of inchoate impressions to the user.

[Super Jamie]

myk said:

Good games that stick to their main assets intelligently, such as DOOM, and properly fit what the technology can do comfortably, will end up being relatively strong. If not, the technology will still allow the author to follow an idea, and will encourage certain design directions once it's put in motion, but only to help make a bunch of inchoate impressions to the user.

I think things are a bit different today than in 1993 though. Like, Doom WAS the limit of what the hardware can do. One can conceivably think of things that technology wasn't capable of yet, room-over-room for example.

To make a game these days, there's almost no limit to what engines and systems can't do, short of beaming the game directly into someone's mind. I think there's alot of options when making a game and alot of technological hype and team sizes need to be alot bigger to pump out a game in a reasonable amount of time and many other factors, it's increasingly more difficult to decide on a vision and stick to in unerringly.

The end result comes across as something like a noob making his first megawad, way too many options and all the good parts spread too thin over filler material that someone else thought would be a good idea, or has worked well before but in a different situation.

[myk]

Super Jamie said:
Like, Doom WAS the limit of what the hardware can do.

Maybe you could put it that way, but going beyond that limit is the issue, not being that limit. Perhaps that's what Tom Hall leaving id meant. While Tom is a capable designer, he wasn't putting into DOOM what it required to reach its full potential, so he lost out and left.

I don't think there's no limit to what can be done nowadays. Technology is like language, its nature always puts limitations on itself. Newer tech can do more things or do things in a more potent way, but that's still limited by what it can do best.

The end result comes across as something like a noob making his first megawad, way too many options and all the good parts spread too thin over filler material that someone else thought would be a good idea, or has worked well before but in a different situation.

That could be said of many games through the history of video games. We tend to forget because we keep the good stuff on our systems and have long deleted most of the crap. State-of-the-art games are more expensive to make today, certainly enough, and thus the initiative to make them has a larger scale, but it still happens that some are of quality, if managed by intelligent developers and producers. If that weren't the case, movies, skyscrapers, satellites and military operations would all be junk made by noobs too, because they're also made through large scale, expensive and complex projects.

[Jodwin]

Creaphis said:

Huh? At first I was under the impression that you intended for that to be an image of a games industry even more dystopian than today's. But you're saying that would be a good thing? If these first two sorts of companies are selling their engines and assets to developers at a profit, doesn't that leave even less monetary momentum for the production of content? Wouldn't such a corporate introduce severe disconnects between engine, art and game design, making it impossible for any products to emerge with functional and aesthetic coherence?

Well, I don't think that many people would mind if all of the shovelware WW2 shooters shared the (almost) same textures and player models. Of course the game devs could always have a few artists of their own to give their game a little flavor, but for the generic assets it probably wouldn't matter so much. Provided the engines would be designed as generic and modular, it shouldn't be too hard for the devs to edit them to their liking so that they could easily create the gameplay experience they want (for example, thanks to Doom being very generic it's reasonably easy to create a lot of different kinds of levels (pure action, horror/tension, adventurous levels...) in Doom, and they don't even feel out of place in the engine).

Also, given enough competition between engine and art providers, buying them should be cheaper for the devs than creating everything by themselves. A lot of companies even today buy engines for their games anyway, and avoiding creating their own engines also avoids having to pay for the best tech experts, paying for engine testing and so on. Similarly there are companies selling art (usually 3D models) for other companies, it just isn't as big yet. And, of course, some game devs also outsource modelling to other companies (I remember some racing games buying their realistic (?) car models from Russia).

[40oz]

Jodwin said:

1. Draw a bunch of rooms.
2. Connect them with hallways (and other rooms).
3. Add locked doors and keys in a manner which gives the players choices for routes.

Where's your work's cited asshole?

[Jodwin]

JohnnyRancid said:

Where's your work's cited asshole?

I said:
too far

That map was basically done by first making a bunch of rooms and then putting them together in a too non-linear way. :P

[Ixnatifual]

The more logical routes the player may take through a level, the less linear it is.

Features you can use to help a non-linear design have been mentioned earlier - interconnectivity between areas, giving the player multiple ways of reaching the same areas, adding extra areas that aren't needed to complete the map, but provide other benefits (new weapons, powerups, health etc). Placement of locked doors can also help the non-linear design. If you have a red door and place the yellow door behind it, as long as you place both the red key and yellow key outside the red door, it doesn't matter in which order the player fetches them. He could visit the red key area and then the yellow key area. Or the yellow key area and then the red key area. Or the red key area, open the red door, visit the area behind it, go to the yellow key area, and then return to the red door area to find the yellow door.

Interesting simplified showcase flow charts comparing Titan map01 and Titan 2 map01. Guess which map is linear and which isn't?



Titan map01 is very linear. It really only has one logical route through it. Begin at "Start", go to "Ridiculously Dark Crate Maze". Go to "Balcony" and flip the switch allowing entry to "Red Key Area", fetch the key from "Red Key Area", unlock the red door to "Bright Craze Maze" and go to the "Exit Area". Every time you play the map you repeat this sequence as there really is no alternative.

With Titan 2 map01 there are a myriad of different routes you might take and it would require large amounts of text to describe them all. The design of the map also requires that you visit all areas eventually, with the one exception of "Chaingun Area". However, the player is rewarded with the Chaingun for visiting this area, so it's a good idea to go there at some point anyway.

[Doom Marine]

Jodwin said:

1. Draw a bunch of rooms.
http://www.akinomori.com/stuff/images/nonlinear1.jpg

2. Connect them with hallways (and other rooms).
http://www.akinomori.com/stuff/images/nonlinear2.jpg

3. Add locked doors and keys in a manner which gives the players choices for routes.
http://www.akinomori.com/stuff/images/nonlinear3.jpg

Why stop there?

4. Connect the hallways with hallways so you can hide in hallways while hiding in hallways


Joking aside, the idea of non-linearity is linked to these questions: how many choices can a player make in a level? How many choices should the player make?



Shown here is the Unholy Cathedral, with four main routes embedded with multiple subroutes interweaving each other, the choice of hitting any 9 of 16 pentagram triggers in any order, and the choice of obtaining either the yellow or blue key to exit.



It's not difficult at all in principle to make a non-linear map: spurring routes in all directions would meet the criteria. But of course, maze maps is not the goal here. Nobody wants to get lost in the Sea of Saragossa of routes that disorients the player in all directions.

For a pleasing non-linear design, non-linear elements are hybridized into the linearity. Take e1m1 for example: you have the sidequest choice of killing zombiemen and sergeants to win body armor. Or how about e1m2: the dark computer hallways are completely optional but rewards the player with a chainsaw.

Combine those ideas with interweaving loops and branches, and you have yourself a pretty non-linear map with a strong gameplay foundation. This isn't a rule or principle, but something to think about.