John Romero's rules for level design

[philtron]

A while back I stumbled on a list of level design rules that John Romero used when he was designing levels for the original Doom. He mentioned these rules retrospectively in an interview years after Doom and Quake had been released. I couldn't find that interview so I got this list from a Doom wiki.

I thought I'd just see what people think or what people have to say about this, even though the rules already mirror much of the advice I've seen on these forums already.

1. always change floor height when changing floor textures
2. use special border textures between different wall segments and doorways
3. be strict about texture alignment
4. consciously use contrast between light and dark areas, cramped and open areas, everywhere in a level
5. make sure that if a player can see outside that they are able to get there somehow
6. be strict about designing several secret areas on every level
7. make sure your levels flow in such a way that the player will revisit areas several times; this will help them better understand the 3D space of the level
8. create easily recognizable landmarks in several places for easier navigation

It really blows my mind how much modern games could be improved if they were designed with some of these rules in mind. Especially the last four.

[Windbreaker]

That list seems pretty spot on; especially the ones concerning border textures and texture alignment and shading. I'd add a 9th rule, though, that says if you know something sucks and you can't fix it, scrap it and start over. I made a map all the way up to the texturing stage only to realize it was poop, so I scrapped it and started over. No use in finishing something you know is awful, when you could be making something better. Or whatever, I don't know.

[RyokoTK]

It's a good list. John Romero has always been a lunatic, but when he put his mind to it (and had someone holding his leash) he was a decent mapper. His efforts in Doom and Doom 2 are all at the "okay-to-good" range -- he did most of the easy, short levels in Episode 1 of Doom and a smattering of mid- to late-game levels in Doom 2. Nothing offensive. In Quake, he did most of the castle levels in Episode 2, which were all really fun and dynamic (relative to the rest of Quake, anyway).

[irons]

Marathon 1 pretty much always violates #5.

[Wrkncacnter]

I don't even get why #5 matters. He's against scenery?

[RyokoTK]

I think it's an issue of the player wanting to explore stuff. If you see a cool rock formation with a waterfall through a window in a game, you might feel like it's worth exploring.

It was probably more relevant back in the 90s when every map sector counted, since games nowadays can pretty much drown the player in overwrought eye candy with no performance loss.

[philtron]

Yeah, #5 is partly about exploration, but I think about it in these terms: Don't waste the audience's time with stuff that doesn't matter. It's true in writing stories and it's true in level design. If there's an area and it looks like you can reasonably get there then you should be able to get there; if the designer doesn't want the player to explore that area then he should just cut it. It's like if in an FPS you saw a gun on the ground and you couldn't pick it up. Or, if in a book there was a ten page description of a building which doesn't relate to anything else in the book.

I think the issue is even more relevant these days because of how often designers use invisible walls/ceilings or "Return To Battlefield" bullshit to hide the fact that their billion polygon levels are essentially hallways. You might see a nice sniper spot, or a potential shortcut, or a cool place to explore and then, BAM, you're forced back onto the linear path.

if you know something sucks and you can't fix it, scrap it and start over

That's a good rule for pretty much anything you do in life.

John Romero has always been a lunatic

Still, what he was aiming for turned out to be pretty good.

Marathon 1 pretty much always violates #5.

Marathon 6: The Violater

[Wrkncacnter]

It's like if in an FPS you saw a gun on the ground and you couldn't pick it up.

I don't know, the exact sentence you wrote makes it sound like if you're playing a level that takes place entirely indoors, you're not allowed to have any windows. Sure, they don't matter, but they can help make it feel like a real place.

in a book there was a ten page description of a building which doesn't relate to anything else in the book.

That's a good idea, thanks. I'll have to remember that for my own project.

[philtron]

Sure, they don't matter, but they can help make it feel like a real place.

Then they matter.

But remember that their function is subservient to giving the player meaningful expectations. A window that looks out onto a distant vista: makes the player feel like they're in a real place. A window that looks into a small courtyard: makes the player feel like that's a place they can explore, and if they can't explore it then that makes the world feel less real.

That's a good idea

No, it's a horrible idea, that's why I used it as my example.

Sure, writing ten pages that have nothing to do with the novel they're in is a great idea if you're a Freshman Creative Writing student and you just learned about '60s and '70s avant garde, but it's not a good idea if you've realized that those sorts of ideas lost their novelty decades ago.

While you're at it you should write a novel in Sumerian Cuneiform and bury ten of its pages in a secret location you tell no one. Why not just print a book with white ink on white paper, say it is a metaphor for the human condition, and then crown yourself King of Pretentious Ideas That Serve No Purpose?

[Windbreaker]

That's a good rule for pretty much anything you do in life.

Unless you're raising a kid. Actually, no. It's a good rule even then.

No, it's a horrible idea, that's why I used it as my example.

He probably knows that.

[RyokoTK]

Yeah, #5 is partly about exploration, but I think about it in these terms: Don't waste the audience's time with stuff that doesn't matter. It's true in writing stories and it's true in level design. If there's an area and it looks like you can reasonably get there then you should be able to get there; if the designer doesn't want the player to explore that area then he should just cut it.

I don't agree with this. Video games (excluding visual novels and the like) as a narrative are so far removed from cinema or literature that a lot of the most basic Storytelling 101 guidelines don't apply, and this is one of them. You may have an ultra-linear shooting game where the player is tied to the railroad tracks, but if you're going to put a very high amount of detail in the setpieces within reach, it's going to look very unnatural to not have any other architecture anywhere. Doom 2's downtown level looked acceptable as just a 4x4 grid of square buildings surrounded by a brick wall because that was the, uh, resolution of detailing available to them then, and yeah, every map sector counts.

But if you've played Left 4 Dead, you'll notice -- at first blush, the world seems convincing enough, but there is absolutely nothing beyond the one street you can go down in The Parish or No Mercy or whatever. It feels very flat. Now that works okay for L4D, because L4D makes absolutely no attempt at realism or separating gameplay conventions from the game world, and you want the players to be funneled down a very strict path for multiplayer purposes, of course, so the lack of visual chaff works. On the other hand, Hard Reset is another game that works like this -- every area you can see, you can access, but the way it's all laid out, it feels incredibly cramped and unconvincing to me. The visual design was really elaborate, but the simplistic level design failed to tie it together. Borderlands 2, conversely, has areas you just aren't allowed to access, either by ledges you clearly can't climb or with annoying invisible walls, and while that's sort of annoying too, I totally adore BL2's level design on the whole, and the world feels very complete.

[treellama]

Sure, writing ten pages that have nothing to do with the novel they're in is a great idea if you're a Freshman Creative Writing student and you just learned about '60s and '70s avant garde, but it's not a good idea if you've realized that those sorts of ideas lost their novelty decades ago.

While you're at it you should write a novel in Sumerian Cuneiform and bury ten of its pages in a secret location you tell no one. Why not just print a book with white ink on white paper, say it is a metaphor for the human condition, and then crown yourself King of Pretentious Ideas That Serve No Purpose?

I think it's time for the backstory--the incident in your past that caused all this bitterness. Did your sister run away to join an artist colony?

[Wrkncacnter]

Why not just print a book with white ink on white paper, say it is a metaphor for the human condition, and then crown yourself King of Pretentious Ideas That Serve No Purpose?

I don't really like doing the same thing twice.

[RyokoTK]

You aren't allowed to make jokes on The Pfhorums.

[philtron]

I think it's time for the backstory--the incident in your past that caused all this bitterness. Did your sister run away to join an artist colony?

No. My artist colony ran away to join my sister. Right after it slept with my wife.

I don't agree with this. Video games (excluding visual novels and the like) as a narrative are so far removed from cinema or literature that a lot of the most basic Storytelling 101 guidelines don't apply, and this is one of them.

This is nonsense. Video games are not some super unique medium completely removed from all others. All storytelling is fundamentally the same. The objects used may be different, but the methods, forms, and narrative devices are identical across all mediums. Point is, the Storytelling 101, 215, and 376 guidelines are equally applicable to literature, video games, or standup comedy. Standup is an especially good example of the rule "don't give the audience things that don't matter"; in a well constructed joke there is no room for useless elements, everything has to contribute to the punch otherwise it won't work.

You may have an ultra-linear shooting game where the player is tied to the railroad tracks, but if you're going to put a very high amount of detail in the setpieces within reach, it's going to look very unnatural to not have any other architecture anywhere.

Then don't put a high amount of detail in the setpieces. Or conversely, don't make it ultra linear. Either way, make the two aspects of the game complement each other not work against each other. A good artist accepts and works within his limitations, he doesn't use his limitations as an excuse to produce poor work. Going off your Hard Reset example, they should have made the visual design less elaborate in order to match the simplistic level design since the disparity between them made things seem unconvincing, which seems to be your point, but it's also my point as well.

I haven't played B2, but if it's anything like Borderlands 1 then I won't like the level design.

[treellama]

Why are we arguing about #5 which, who cares, when #1 and #6 are clearly BS?

[RyokoTK]

All storytelling is fundamentally the same. The objects used may be different, but the methods, forms, and narrative devices are identical across all mediums.

This is such an incorrect point of view even when you compare, like, cinema to television, or fiction to poetry, let alone any of those to video games. The key difference in video games is that the player has an active role in the narrative -- pressing buttons and winning levels -- and gameplay conventions typically interpose themselves into the storytelling. The bit between cutscenes in Call of Duty games -- the part where you shoot brown people -- is what makes this medium different from others, and there's an expectation that the game playing experience is still complete and enjoyable regardless of what the game is. Look at the reviews for the new game Spec Ops: people generally praise the story and characterization for being really dark and gruesome, toying with people's expectations of the genre (middle east murder simulators), but the gameplay itself is so bland and unsatisfying that it's still getting middling reviews overall. If you just wanted to tell a story about Heart of Darkness parallels in Dubai, make a movie. I had similar complaints about Eternal; Forrest had quite a story to tell, but the whole video games medium did not work well with it.

Point is, the Storytelling 101, 215, and 376 guidelines are equally applicable to literature, video games, or standup comedy. Standup is an especially good example of the rule "don't give the audience things that don't matter"; in a well constructed joke there is no room for useless elements, everything has to contribute to the punch otherwise it won't work.

I think you're missing the point. Window dressing does matter in games. It all contributes to the sets, and the gameplay experience, which I was talking about above. Even a game like Portal 2, which is pretty finely distilled pure gameplay, has quite a lot of that window dressing to make the setting seem more complete. Compare the level designs from Portal 1 to Portal 2. The first game is a pure puzzle game with basically no relevant setting (it's a bunch of white rooms that you make portals in), but the second game really fills that out and completes the picture, even though the game itself is no more complex.

Then don't put a high amount of detail in the setpieces. Or conversely, don't make it ultra linear. Either way, make the two aspects of the game complement each other not work against each other. A good artist accepts and works within his limitations, he doesn't use his limitations as an excuse to produce poor work. Going off your Hard Reset example, they should have made the visual design less elaborate in order to match the simplistic level design since the disparity between them made things seem unconvincing, which seems to be your point, but it's also my point as well.

Paring down the details leads to a very anemic-feeling game. Which is great in a game like Quake 3 or Left 4 Dead, because these are foremost about PVP competition. I don't think I understand your point about how an ultra-linear game can't have that kind of world-building if it's not actually in the field of play. As long as the game design is clear about what is in play and what isn't, I don't see the problem. In a good game, your attention will be focused on what's in play. Take the canal and airboat levels in Half Life 2; there are several areas -- buildings, bridges, whatever else -- where Combine troops can attack you from, and you can't get up there. The game is pretty clear that it's out of bounds, but it's still part of the set design, and it takes the game beyond just a level in a stone trench floating in space. Now it's a drain pipe in a city, and it's much more convincing, even though Half Life 2 is linear as shit.

Why are we arguing about #5 which, who cares, when #1 and #6 are clearly BS?

They made more sense in Doom. Particularly #1. Marathon 2 has a lot of textures that have visible seams -- tiled floors, for example -- so you can have two textures abut at the seams and it looks completely natural. Doom doesn't have a lot of these. If you just put a grass texture and a rock texture on a flat floor touching each other without some kind of visible border, it looks weird. Having an elevation shift works nicely for that. Nowadays, of course, you can just have textures blend into each other and it's okay. It's basically point #2 applied to floors.

#6 is about encouraging and rewarding map exploration, I guess. Which is good.

[treellama]

#6 is about encouraging and rewarding map exploration, I guess. Which is good.

OK but you don't need them in every level to accomplish that.

I will admit, when I first replied I only had net maps in mind--because (as the scenario contest showed) who is making anything but net maps now? So, anyway, less BS, but still a little BS.

[philtron]

This is such an incorrect point of view even when you compare, like, cinema to television, or fiction to poetry, let alone any of those to video games. The key difference in video games is that the player has an active role in the narrative

I'll respond to this in a different thread because I feel it deserves it's own topic.

I think you're missing the point. Window dressing does matter in games.

I never said it doesn't.

Window dressing can be important in any medium, even music or poetry, but only if it is used correctly. If it's not used correctly (or I should say, used skillfully) then it just wastes the artists's and the audience's time.

Paring down the details leads to a very anemic-feeling game.

Disagree. Paring down the details leads to a cleaner, more succinct, and more focused feeling game.

Elegance through simplicity.

I don't think I understand your point about how an ultra-linear game can't have that kind of world-building if it's not actually in the field of play.

My point was that the various aspects of game design should complement each other. Creating inaccessible side areas to a linear level works against the linear design; it draws the player's attention away from the focus of the level, it wastes the player's time by creating false expectations, it's a waste of resources, and it's a false representation of the virtual reality that's been constructed. Ultimately, creating an area that looks playable but which is outside the field of play for the sole purpose of world building or storytelling is lazy; that world building could done within the field of play, it's just that it would take more creativity and effort to do so.

There are a few exceptions that I can get behind. Take the Citadel in Half-life 2 for example. It's outside the field of play and is important for storytelling and world building; it can even be used as a landmark by the player to orient himself when he gets lost. However, it's also really clear that you can't reasonably get there, it might as well be part of the skybox (if it isn't already). I can get behind that style of world building. Or consider a situation where you're looking into a room you can't access, seeing a scene play out, and then you can access the room once the scene is over. I can accept that since the area becomes playable after a time. However, this sort of thing should be used sparingly or not at all.

As long as the game design is clear about what is in play and what isn't, I don't see the problem. In a good game, your attention will be focused on what's in play. Take the canal and airboat levels in Half Life 2; there are several areas -- buildings, bridges, whatever else -- where Combine troops can attack you from, and you can't get up there. The game is pretty clear that it's out of bounds, but it's still part of the set design, and it takes the game beyond just a level in a stone trench floating in space. Now it's a drain pipe in a city, and it's much more convincing, even though Half Life 2 is linear as shit.

I agree with a lot of what you say here. It's good that Half-life 2 (usually) makes it clear when it's possible to reach an area and when it isn't. That's good design. But even better design, in my opinion, is if any area the enemies can reach is an area the player can reach as well. In the situation you mentioned above, the player is forced to use certain weapons and certain tactics which makes not just the level design linear but it also makes the gameplay more linear (yes, I understand that limiting the player's options can add enjoyable challenge to an area but I don't think this is such a situation). If the player wants to attack those enemies with a crowbar or a short range weapon then they should be allowed to do that. You can still make the area convincing from a world building perspective without sacrificing gameplay options which is what this situation (enemies attacking from inaccessible areas) does.

OK but you don't need them in every level to accomplish that.

No, you don't need them. Nor do you need them in every level. But they definitely add a little depth to each level they're in and they reward the player for investing extra time and interest in your game. So, every level that has a few secrets is a little more enjoyable to play.

[RyokoTK]

My point was that the various aspects of game design should complement each other. Creating inaccessible side areas to a linear level works against the linear design; it draws the player's attention away from the focus of the level, it wastes the player's time by creating false expectations, it's a waste of resources, and it's a false representation of the virtual reality that's been constructed. Ultimately, creating an area that looks playable but which is outside the field of play for the sole purpose of world building or storytelling is lazy; that world building could done within the field of play, it's just that it would take more creativity and effort to do so.

It's not really any of these, actually. It's certainly not working against a linear level if it's inaccessible since, y'know, it's inaccessible. If anything, once a player is wise to the fact that a game is super linear, he's not going to be worrying about the sniper ledge the baddie is hiding on. If it's accessible, he'll find his way there eventually, and if it's not accessible, there's nothing to worry about. Also, creating level architecture outside of the world of play just for world-building is pretty much the complete opposite of laziness -- it's extra work for a marginal payoff.

Here's an example for you. Say you're making a level that takes place in a city street. In the real world, not only can you see the houses on the street you're on, you can usually see the houses on the next block over, and taller buildings off in the distance, and you can usually see the moving lights of traffic on adjacent streets, and so forth. If, in your game, all you model is what's on the street you're on, it looks and feels like a theme park: all facade with no content. That's what L4D's level design is like. It's unconvincing, like exterior sets in sitcoms. When they film at a tight, close distance, everything seems normal enough, but there's absolutely nothing beyond that. The difference between TV and video games is that the player controls the camera, so he can look and move around the set and see how flimsy it looks; on TV (and in film) you can control what the viewer sees.

But even better design, in my opinion, is if any area the enemies can reach is an area the player can reach as well. In the situation you mentioned above, the player is forced to use certain weapons and certain tactics which makes not just the level design linear but it also makes the gameplay more linear (yes, I understand that limiting the player's options can add enjoyable challenge to an area but I don't think this is such a situation). If the player wants to attack those enemies with a crowbar or a short range weapon then they should be allowed to do that. You can still make the area convincing from a world building perspective without sacrificing gameplay options which is what this situation (enemies attacking from inaccessible areas) does.

This is getting a little far afield of Romero's point, I think. I don't really disagree with this point, but I don't think monster sniper ledges are what he was referring to. There are a lot of Doom 1 levels in the first episode where the entirety of the level proper takes place inside a building, and around the building is a courtyard. If Romero made that level, that courtyard would be accessible and there'd usually be an item out there to collect to reward exploration, but it works just fine as window dressing too. Helps keep you from feeling like the building is just floating in space.

[irons]

Pretentious

QUOTE(skraeling)

Te Occidere Possunt Sed Te Edere Non Possunt Nefas Est[/quote]

[patrick]

quod erat vale, demonstrandum

[philtron]

@irons
How so?

Also, creating level architecture outside of the world of play just for world-building is pretty much the complete opposite of laziness -- it's extra work for a marginal payoff.

Come on. The amount of "stuff" in the final product is not a measure of the amount of effort put into it. Yeah, adding extra architecture outside the play space is more work than not adding anything, but it is less work than adding something to the playspace without interfering with gameplay. If you wanted to achieve the same world building effect within the play space as outside of it you'd have to put a lot more thought into what you're doing, do more iterating, and do more testing. The extra thought put into the design alone means that it's more work than adding stuff outside the field of play. And that's why I say avoiding that extra effort is lazy.

This is getting a little far afield of Romero's point, I think. I don't really disagree with this point, but I don't think monster sniper ledges are what he was referring to. There are a lot of Doom 1 levels in the first episode where the entirety of the level proper takes place inside a building, and around the building is a courtyard. If Romero made that level, that courtyard would be accessible and there'd usually be an item out there to collect to reward exploration, but it works just fine as window dressing too. Helps keep you from feeling like the building is just floating in space.

True, the conversation's left the original point. So, going off your example of the building with the courtyard around it:

I disagree that the courtyard works fine as window dressing. Despite the goal of using it to make the world feel more real or expansive these areas have always made the world feel more fake to me and have only made me feel trapped and frustrated as I try to reach them and fail. I mean if you go through the trouble of making the area in the first place then just take the extra step and let the player travel there. And if you're not willing to do that then do you really need that window dressing? I really don't think the absence of external views will make the player feel like the building is floating in space, it never has for me.

The best analogy I can make is that seemingly-playable areas that can't be reached make me feel the same way as when a game features fake doors and toilets that don't flush. If you're going to add a toilet to give a certain feel to your level then make it function like a real toilet. Otherwise it doesn't even function as a set piece. And fake doors, good god, I hate fake doors. Every time I go up to a door and realize it's not real it is a moment of disappointment. Why is that door there? To make the world seem more real? But the door isn't real, so now the world feels even more artificial than if that texture wasn't there in the first place.

[patrick]

@irons

Jon 19:22 What I have written I have written.

[RyokoTK]

Come on. The amount of "stuff" in the final product is not a measure of the amount of effort put into it. Yeah, adding extra architecture outside the play space is more work than not adding anything, but it is less work than adding something to the playspace without interfering with gameplay. If you wanted to achieve the same world building effect within the play space as outside of it you'd have to put a lot more thought into what you're doing, do more iterating, and do more testing. The extra thought put into the design alone means that it's more work than adding stuff outside the field of play. And that's why I say avoiding that extra effort is lazy.

I'm not really sure that you understand what kind of damage additional side-areas can be to gameplay, depending on the design philosophy behind the game. If you're making some kind of prison level and there's a guard tower just out of reach, sure, it may be frustrating that you can't get into that guard tower, and if you're the level designer, you may decide, hey -- what harm can it be if I put a little side passage that leads to this guard tower? Now I can put a sniper rifle and some ammo here and reward people that go off the beaten path.

The thing is, every time you add a little branching path off of your "main route" (assuming we're talking about a more linear game and not Skyrim or whatever), you're now diverting attention from that main route. You're slowing down the pacing. And not everyone is that good at level navigation, so with every little secondary road, you add in another way to confuse some players. Particularly in games that don't really have tangible rewards -- what are you going to find in Halo as a gift that can be of help? Yet another overshield? A rocket launcher with its paltry two shots? Great. Not to imply that these side areas and bits of exploration are bad, mind you. I think they're great. They're just not necessary, is all. On the other hand, I do think that kind of window dressing typically is necessary, or at least more so, assuming you're trying to affect a somewhat realistic feel.

The best analogy I can make is that seemingly-playable areas that can't be reached make me feel the same way as when a game features fake doors and toilets that don't flush. If you're going to add a toilet to give a certain feel to your level then make it function like a real toilet. Otherwise it doesn't even function as a set piece.

I'm going to go after this example in particular. What kind of feeling does a toilet give to a room in a level other than that you're in a bathroom? When I see a toilet, I know I'm in a bathroom. I don't need to hear a stock toilet flush sound effect to affirm that opinion. In this case it maybe is a case of laziness, but still, the fact that the plumbing doesn't work doesn't detract from the appearance of a bathroom whatsoever.

But I mean, I can't even take a crap in the toilet, so what function does it serve, right?