Ten Things I Learned Building a Quake Level in 2017

(November 19, 2017)

On August 15, 2017, I started work on a map for Quake. It wasn't anything big. I had a vision of a temple in my head, a pad of graph paper with a rough layout, and a few photos from Google Images. I remember being on a Discord call with a friend as I built it. I drew out a pyramid shape and two walkways.

Finished maps are second only to fulfilling discussions in the VDU chat in rarity. I almost gave this one up. I've decided to write up this piece in case you have issues finishing your own maps, and what I learned about overcoming it.

1. Have something of a plan going in.

For Temple of the Strange, I had dug up a pad of graph paper and found that I had drawn out a partial temple layout for a story I was working on with a friend some years prior. The story was never finished, but the layout stuck around. I don't like letting plans for things go to waste, so I figured I'd use it here.

I deviated a lot from the plan (I can't make out what half of it is either), but it did help to fill in some of the blanks on what a "temple" was. In particular, the hallway containing the megahealth secret was lifted from a hallway containing jail cells. At the end of that hallway is a torture room, which inspired the room where you press the button to lift the cage off the key. The fallen, cracked pillar and the hole in the wall in the spawn area were lifted from the actual worship area of the temple.

Even if you're not actively drawing out your level, knowing what you want before you go in can save you a lot of time later. I knew I wanted two branching paths the player could take at the start, hence the two walkways. Plan ahead.

2. Restrict yourself.

Restrictions breed creativity. I honestly believe one of the biggest challenges facing the game industry today is the ability to do literally anything with the technology. Tight restrictions in available technology or skillset requires you to get creative with how you make your ideas come to life. The end result tends to be a lot more inspired as a result, provided you pull it off well. Thing is, you can arbitrarily impose restrictions on yourself and still get the benefits.

I limited myself to one texture set (4-5) per area, and I tried to build the level as much as possible like an original id level. I was gonna pay attention to my r_speeds, I wasn't gonna go overboard on secrets or monster count. The cool thing about that is that you can pay attention to details when you restrict yourself. You get creative with the way you build. You can't rely on lots of different ideas and quirks to carry your level. It becomes about experimenting with a few in particular.

Put another way...

"The songs are really simply structured. A lot of them are purposely like that. That way you can get away with a lot more when you mess around with the contents."—Kevin Shields on recording My Bloody Valentine's Loveless

You don't need to necessarily restrict yourself in these ways. Build something grand, but find some way to limit yourself and your ideas. That way, you can focus. Big, open fields are rarely pretty and even more rarely playable.

3. Build things with a purpose.

In the same vein, not every map needs every gameplay element. Not every Counter-Strike map needs sniper posts. Don't force gameplay elements where they don't belong just because you like them. You can design stuff in, but it has to fit.

In general, elements should have a reason for being there. Temple's watery side cavern is a good example; not only does it feature monsters, a secret, and a rocket launcher, but using it brings you right back into the flow of the map, without having skipped a single enemy. Speaking of flow...

4. Players shouldn't have a hard time navigating.

Quake is a pretty fast-paced game, but even slower games have some sort of flow through their levels. In a multiplayer game, I should know where the capture points are, or where the bomb sites are. If there's an end to the level, it should be pretty obvious where the end is. Temple's a pretty linear level, but everything leads to the end. Even going backwards is linear. You can't really get confused and go in circles.

In the above visual, the player spawns in the bottom left and the exit to the level is to the right. Note the giant lava chasm.

You can challenge players on enemies, puzzles, and most of everything else, but navigation shouldn't be one of them. If I can't find my way, I can't play your level. That doesn't mean put up neon arrow signs and big flashing text saying "IT'S THIS WAY DICKHEAD", but subtle hints, especially in more open-ended areas, aren't a bad idea. Players sometimes miss stuff. Guide them back.

5. Textures can help you measure walls and keep your brushwork clean.

In the middle stages of the level, I used bands of skulls on the wall to break up the monotony. Thing is, these skulls actually helped me keep my measurements consistent.

Textures probably shouldn't split off in the middle. If the sides of each wall had half a skull, it would look wrong and sloppy. Because I made sure there was a proper quantity of skulls per-wall, my brushes were kept on the grid and the brushwork became much neater as a result. Things like that add a layer of polish to a map. (J.A.C.K. also has a UV-viewing mode, seen above, that helped me out in aligning textures.)

6. Reuse areas.

Reusing areas creates a sense of continuity in the level. It recontextualizes an area, adding depth to what you're building, and it's a good way to increase the length of the level without building a lot more of it. It's all over id's maps, too (E1M4, E2M4, E4M5, among others).

For Temple, I wanted the player to blow through the map up to the staircase in the middle, come to a locked door, and subsequently go searching for the key. They loop back into an adjacent room through a lift, grab the key, fight off a shambler, and come right back out in the direction of the staircase, all without stopping a single time.

The staircase now becomes the place the player fights off the ogre behind the door. You're now fighting up, rather than down. I could've just had that room become another passageway, but because the door's locked, the player has to stop and survey their surroundings. The apparent length of the level's also been increased, without much extra building.

Even in a multiplayer context, the rooms overlooking the courtyard next to CT spawn in de_cbble reframes the area. Snipers make use of the area differently than players on the ground. Rarely does a space serve only one use in real life; you can reuse areas in a game in much the same way.

7. Make what you'd wanna see as a player.

Every game has specific traits and aspects of gameplay you should account for in your map. In Quake, I like to keep moving. I like well-oiled layouts and layouts that are easy to follow. That's not gonna work for every game, and some levels disregard that altogether to great effect, but it works for Temple.

Build things you'd like to play into a level. If it's not a level you find fun, it's probably not gonna be fun. I couldn't care less about shiny maps that are a slog to play, and neither should you.

8. You'll probably need to redo part of it.

After it was decided that the finished level was way too hard, I deleted all the monsters, ammo, health, and armor and redid their placement from scratch. This time, the level was much better balanced. I already had the flow solidified and I could focus just on what the player was to fight, what they needed, and where.

As an example of the ridiculousness, at the start of the level, I had players facing off against an Ogre with only the Shotgun and maybe the Nailgun if they thought to grab it first. There was a vore, a spawn, and two death knights near the end. On Normal. It was totally unbalanced, and I could barely beat it myself.

If something's not fun, delete it. If something needs to be redesigned, delete it. Oftentimes, it's an inevitability. It might seem like a catastrophe or a sign that something's gone wrong, and you might be tempted to just rework the area. If it's not working, though, sometimes it's just best to cut it altogether.

Put another way...

"We have a saying at Valve that everyone will have their favorite feature cut from the game. While it's not true, it does help everyone prepare themselves for the fact that they will have features they like—or that they spent some to a lot of time on—cut."—Greg Coomer at Valve

9. Doodle.

After the staircase room, I knew I wanted to have varying heights and puzzles, something that would require players to slow down and explore a bit. I wasn't sure what I wanted to do with it, so I pulled out that pad of graph paper and started doodling:

That became this area.

If you're stuck, turn to doodles before you commit to brushwork. Even stupid, simple stuff can help you figure out exactly what you want out of an area before you have to actually make it.

10. Good documentation is a necessity.

This might seem like kind of a cop-out end to the list, but making Temple of the Strange actually inspired this entire website.

I had to rely on the Internet Archive for a lot of the documentation I needed to make the map. Writeups, example maps, and utilities for Quake, all lost to the Aether, only living on in archived form. It was a lot of digging that ended up distracting me from actually making the map. I'm not alone, either; just this morning, someone in our Discord mentioned how hard it was to find Quake II mapping information. His grand quest to find ArghRad was something to behold.

Thing is, that's all completely preventable.

Allow me to introduce you to a man named Jason Scott. Jason Scott is, in no uncertain terms, a digital preservationist. He works for the Internet Archive, archiving and storing boxes upon boxes of everything from CDs and hard drives to, recently, Apple II floppies. He also heads the Archive Team, who swoop in and spider and archive entire websites right before they die, and hopefully, long before that.

Jason Scott's entire ideology is built around the preservation of our modern digital culture. There's a very real chance that at least 90% of all the data we generate today will be lost to time. Our lives are lived out on the internet now. It's thousands of hours of work and consumption from people just like us. It's history. It's culture. The web is the dusty terracotta pots of today for the historian tomorrow. If you've got a minute, read these two posts from him and tell me you're not convinced. Aside from just plain enjoying his writing and speaking, he's honestly a personal hero of mine.

Jason made me realize something. We've all been sold a lie about how safe our data is. Companies can die in a wink, and all those photos, music, stories, and memories you uploaded over the years that they told you was secure in the ~cloud~? Gone. When people shut down their sites or abandon their profiles because life gets in the way? Gone. And when it comes to an ad-hoc modding community of fan-created tools and information on games so old, they don't even need a 3D card to run?

How many times have you run into some obscure problem with your tools, and your only resource has been a long-dead forum thread with three replies and a Megaupload link? I've seen it, you've seen it. We might not be able to preserve it all, but that's no excuse to not try.

So, I mobilized, and with the indefatigable work of my webmaster, Jax, the Valve Developer Union was born. In the three months we've been working on this website, I've archived from shiny, pressed-up fansites and from fucking OneDrive accounts alike. Through it, I've met people all over the world, with different skillsets, backgrounds, and mental disorders. We're still not sure of our own future, but we're dedicated to making this happen. And it is happening.

And it all happened because I had a vision of a temple, a pad of graph paper, and a few photos from Google Images.