Level Design For Games

Aeonite writes "As a content writer I was not heavily involved in the level design process at my last game industry job, but Phil Co's Level design For Games: Creating Compelling Game Experiences accompanied me to work every day. Not only is it a good introduction to the world of level design, but it also provides an excellent overview of the entire game design process." Read below for the rest of Michael's thoughts on this book. Level Design For Games: Creating Compelling Game Experiences author Phil Co pages 352 publisher New Riders rating 10 reviewer Michael Fiegel ISBN 0321375971 summary An excellent introduction to the art of game level design In the past I've been rather verbose when reviewing books about game design, as I wished to provide evidence that justified the often less than stellar score I gave the book in question. I'm pleased that I don't have to do that with this book, which as far as I can tell is a nearly flawless introduction to level design. As such, this review will be more of a recap, so as to help you decide if the book's content is right for you.

Chapter 1, "How Do You Make a Game?," discusses the game development process from Pre-Production through Gold Master by way of showing how level design fits into the overall scheme of things. Also discussed are design documents, basic level geometry, and the difference between alpha and beta, and A, B, C and D bugs (A being "fix this now" and D being "nice to have, maybe later").

Chapter 2, "Defining the Game," focuses on the various types of games on the market and the differences between them, from first-person shooters to platformers, action RPGs to MMORPGs. Also discussed in some depth are themes (fantasy, sci-fi), ESRB ratings and audience age, and system limitations.

Chapter 3, "Enemies and Obstacles: Choosing Your Challenges," is where the book really begins to get into the nitty-gritty of the level design process. This third chapter covers the placement of enemies ("mobs") and objects within the level, the types of levels (hubs, boss levels, etc.), skill trees and the application of skills to obstacles within each level.

With an idea of what needs to go where, Chapter 4, "Brainstorming Your Level Ideas," delves into the creation of concept sketches and reference images, the creation of a level's storyline, the drafting of a level description and the design of the puzzles and scripted sequences within the level (which incorporate the mobs and objects discussed previously).

Chapter 5, "Designing With a Diagram," is where all those ideas and brainstorming begin to take concrete shape. A primary concern here is the scope and order of levels within the game, particularly in terms of a player's progress through each level. Once you know where your level fits into the overall schema, the author tells you to lay it out in diagram format by creating a grid; this is not unlike a Dungeon Master carving out 10' by 10' dungeon corridors on graph paper for a D&D game. You know who you are.

Chapter 6, "The Template," introduces the reader to UnrealEd, a level editor for which a demo is provided in the back of the book. The author walks through the basics of using UnrealEd, from the basic creation of a room and the placement of an NPC within it to slightly more advanced topics such as vertex editing and static meshes. It's a fairly technical chapter, but is laid out clearly with numbered instructions and plenty of screenshots to guide the reader along.

Chapter 7, "Improving Your Level," jumps ahead in time a bit, assuming that you've already mastered the basics from Chapter 6 and have created a level template that can now be play-tested. It focuses mostly on that play-testing process and how to adjust and balance one's level based on feedback in order to make it fun and functional.

The next chapter, "Taking It to 11," is more concerned with polish and quality. Topics include architectural style, the addition of details like trim and borders, the appropriate use of textures and props, and the like. The second third of the chapter takes the reader back into UnrealEd to practice some of these skills, including the creation of new shapes and a radial building technique to create curved hallways an rounded rooms. Finally, the chapter discusses the addition of other game elements, including scripted sequences, ambient sounds and music, and other special effects such as fog.

The final chapter, "Ship It!," revisits the concept of Alpha, Beta and Gold Master in more depth, discussing optimization, the creation of zones (with an UnrealEd tutorial to help the reader along), game balance, and bug testing. It closes off with some discussion of helpful skills and practices one might pick up, including how to file a good bug, why you should archive data, and how to take good screenshots.

On the subject of screenshots, it is worth noting here that the book contains one such shot from Flagship Studio's Hellgate: London, a game which I am downloading from the EA store as I write this review, and which is scheduled for official release on Halloween, 2007. In my experience, many books on game design tend to incorporate screenshots and examples from older games, and it's rare to find a book that includes a screenshot from a game that is not only current, but as of the book's publication was yet unreleased. Indeed, most of the examples in the book are of games released in the past several years (Psychonauts, Half-Life 2, Doom 3), and this gives the book added relevance, appeal and longevity.

Aside from the more technical language involved with the UnrealEd tutorials, the book's clear language and friendly tone makes it quite accessible, even for those not of a technical persuasion. While I can't speak to how much the book would help a more experienced LD, it definitely seems appropriate for a beginner who's eager to learn the craft, or anyone interested in the game industry as a whole. I highly recommend it.

You can purchase Level Design For Games: Creating Compelling Game Experiences from amazon.com. Slashdot welcomes readers' book reviews -- to see your own review here, read the book review guidelines, then visit the submission page.

Comments

Level Design Primer

[TrevorB]

For those of you who don't want to pick up nasty bookseses, pick up a copy of the Orange Box, and play through Half Life 2. Particularly pay attention to the developers commentary in HL2 Episodes One and Two, Portal, and Team Fortress. You'll have a much better appreciation for what level design is and what it means, and (IMHO) Valve is the king of it.

No I am not a Valve Employee.

Re:Level Design Primer

[CastrTroy]

Also, try playing some really crappy games, so you can find out what doesn't work. It's almost as important to understand bad design as it is to understand good design.

Re:Level Design Primer

[p0tat03]

For multiplayer level designers, I'd take a very long look at Call of Duty 4. Where Valve is the king of maps in general, Infinity Ward has a ridiculous way about making use of height, rooftops, and general stacking of levels while maintaining impeccable multiplayer balance and flow.

Re:Level Design Primer

[gEvil (beta)]

Also, try playing some really crappy games, so you can find out what doesn't work.

Got any suggestions?

Re:Level Design Primer

[Anonymous Coward]

or try the original natural selection

fkin sweet maps

Re:Level Design Primer

[quest(answer)ion]

For multiplayer level designers, I'd take a very long look at Call of Duty 4. IMO, the CoD games are also worth taking a look at for single-player level design. they have their drawbacks--overly scripted and railed progression, nonsensical objective-linked enemy spawns, etc--but some of the moments that appear in these games are just brilliant. take the ghillies in the wind level in CoD4: it's a heavily scripted sniper mission, but it captures a sense of tension and realism that makes you believe that if you deviate in one detail, your character would indeed be dead. scripted and inflexible or not, that's an impressive gaming experience.

one of the things that makes the level work (and Valve does a bunch of this in HL, too) is believable and crucial interaction with your partner NPCs. important as the physical details of the environment are, unless the game is about the last dude on earth, other characters have a role as design elements. they contribute both to the feel of the game and to the flow of whatever story is being told.

Re:Level Design Primer

[SQLGuru]

Anything with the word Extreme in the title (esp. spelled "cool").

A list for your (dis)pleasure: http://en.wikipedia.org/wiki/List_of_video_games_considered_the_worst_ever

Layne

Re:Level Design Primer

[Mantrid]

Blacksite: Area 51 featuring mediocre gunplay, then empty stretches of unfinished levels.

Re:Level Design Primer

[hansamurai]

Half-Life 2 (etc.) has some awesome levels. Lately I've been playing Psychonauts and the level design in that game is amazing. Basically the concept of the game is that you're jumping into different people's brains and experiencing and fixing things that are going wrong. The game starts out with a linear go here to here brain but then starts introducing some amazing designs.

Some of my favorite levels that I've played so far are:
A giant war board game! The game places you in a room with two people playing a hexagonal board game, one of them is Napoleon so you have to help the less skilled player. You jump in the board game and are teleported down to the board where you can run around at the size of the buildings and playing pieces. Then you can go down even further and experience the level as if you were a real person on the board. The levels are very seamless and you can jump between them easily. Very fun level to play.

Another level has you jumping into the mind of a giant fish, where the main character is basically Godzilla. All the fish inhabitants of are yelling up at you as you crush their buildings and squash their tanks (you just crushed an orphanage Goggolor!).

Well, those are just two out of something like 15 levels, not sure, still haven't finished it! What's great is that each level is varied in setting and general gameplay.

Re:Level Design Primer

[Thrymm]

I agree, Psychonauts level design is amazing and very well planned out for the storyline at the time you are in them.

Re:Level Design Primer

[vimh42]

No I am not a Valve Employee.

But it appears Phil Co is or was at some point. http://www.valvesoftware.com/people.html

So perhaps the book is worth a look.

Re:Level Design Primer

[Anonymous Coward]

Sorry to troll, but actually I totally hate the Valve Game Design. I did buy the Orange Box and started playing Half-Life 2 Ep 1 (I already played the original HL2 when it came out). But the level design is extremely repetetive. Every kind of puzzle you have always to do 3 times before there comes something new. There are always the exactly same houses, rooms and enemies throughout the whole episode (and I encountered all of them already in the original game). I would take Half-Life 2 rather as an example how you shouldn't do it.

Re:Level Design Primer

[Cathoderoytube]

I remember pretty much every old game by Raven had outright awful level design. Every level is like a maze and it just winds up being an extremely aggravating experience. I always thought Hexen II was an awesome game. The level design was complete shit though.

Re:Level Design Primer

[EmperorKagato]

The Angry Video Game Nerd can provide good insight in a humorous manner.

Re:Level Design Primer

[Jackmn]

Got any suggestions?

Daikatana

Re:Level Design Primer

[Esc7]

I came in here to say the exact same thing. I picked up the Orange Box a little while ago and I'm playing through Half Life 2 for the first time in my life. Never before have I played a game with such a rich linear level design, that simultaneously feels "real" but also subtly guides you along. The areas nearly never get boring, and are all memorable. The commentary tracks, especially in Portal, are invaluable if you want to design a good game. There they tell you, step by step, how to introduce completely new game concepts, present puzzles, and pace the player to create a unified theme and experience. Both are works of art in my opinion.

Re:Level Design Primer

[TempeTerra]

I remember the original Dungeon Siege being appalling. The game was pretty well designed, decent combat mechanics and skill system, good loot system... a lot like diablo II with an entire party. Graphics were very good for the period and the environment was nicely designed. No level-loading either, I think it was one of the first games to stream in nearby areas before you got to them so you could walk from the start of the game to the end without seeing a loading screen. Problem: completely linear. There was exactly one way to walk from the start of the game to the end. A very pretty corridor with some open areas such as large caves and forest clearings, and I don't think enemies ever respawned so at any point there was usually exactly one group of enemies which you had to kill next to advance the game.

I think some people liked it, I recall some critical acclaim at the time it was released, but that may have been due to reviewers snorting too much coke through the hundred-dollar bills they were paid.

Bitching aside, apart from the level design it was a fantastic game and there were some very interesting mods for it; notably the Ultima 5 Lazarus project which was a complete remake of U5 with the Dungeon Siege engine.

Re:Level Design Primer

[totally bogus dude]

That level of CoD4 felt a bit forced to me because of the conversation from your partner. On the one hand it's trying to give you the feel of being an elite sniper on a high-risk mission deep in hostile territory - the kind of mission only the best of the best would be sent on. On the other hand, the team leader is talking to you like it's your character's first ever covert operation. I found that a bit jarring.

Re:Level Design Primer

[Anonymous Coward]

"Also, try playing some really crappy games, so you can find out what doesn't work."

I think I will.

Re:Level Design Primer

[Anonymous Coward]

All Doom games
All Halo games
All Marathon games
All Quake games (except Q3A)

These and practically anything made by id software or Bungie embody shitty level design.

Re:Level Design Primer

[tygerstripes]

A third-party HL2 mod/level that demonstrated impeccable level design was Minerva. People may wax lyrical about the feel and flow of the gameplay, but the physical layout for this was astounding.

Every nook and cranny of the layout was used superbly. When you see the actual map for the level(s), you realise just how small it is and think "No way - I spent a fantastic hour in that room..."

Just sayin'.

Re:Level Design Primer

[Anonymous Coward]

Vista

Re:Level Design Primer

[vuffi_raa]

besides your FPS games some of the most amazing level design in my opinion cames from games like: psychonauts (esp. the crazy milkman level and the black velvet world), sonic adventure and sonic adventure 2\sonic rivals among others. they have a very "non-standard" approach to level design where the player is challenged by the physics and dimensions of the level vs. something that just looks nice (although the look nice as well).

Re:Level Design Primer

[Anonymous Coward]

Quake 1 is good level design, all Valve has ever done is utter boring shit. Nearly all passages are too long and turn sooner or later into an endless tiresome quick-load/save mess. You can't get through if you don't know in advance what will happen.
Try playing any Valve game you already played through again, you will get bored. All Valve has ever done is revolutionary art/gfx/physics you never saw before in any game. But the gameplay? God forbid.

Re:Level Design Primer

[quest(answer)ion]

That level of CoD4 felt a bit forced to me because of the conversation from your partner you're absolutely right. hell, the fact that they're talking at all stretches realism a bit too far for my tastes. i'm not saying the level is perfect, just that elements of it are extraordinarily well done. i think playing that level with the captain's tutorial voice-overs muted would be amazing--the hand signals are actually almost enough to explain what needs to be done at crucial moments, just as they would need to be in the field.

I was going to post a shock site link

[Anonymous Coward]

I was going to post a shock site link but the sheer lenght of this review makes me think that somebody may have put some effort in to it.

meh.

Re:I was going to post a shock site link

[Anonymous Coward]

You were wrong. The review is a summary of the chapters, not a review of the book.

Quick...

[Anonymous Coward]

Someone, give that book to these guys.

Too bad Hellgate: London sucked

[PrescriptionWarning]

I played the demo of the game and found it to be about as fun as whack-a-mole. I guess to me it just felt like they were trying to go for a Diablo like feel, but doing very poorly at it. For example if you took the melee fighter, there was only a couple of swing animations and it was tough to hit enemies properly as it was tough to determine the range to the opponent. If you took the gun fighter, it just felt like a cheap old FPS since the enemies had zero intelligence (much like you'd find in a Diablo like game).

Re:Too bad Hellgate: London sucked

[coolGuyZak]

In addition, they didn't let you play with several hyped features, particularly weapons modding. That royally annoyed me.

I found that Mass Effect has delivered on what Hellgate attempted to sell, sans multiplayer support and dynamically generated levels. (Weapons modding, tons of story lines, a similar seeming combat system (albeit almost wholly ranged combat), with downloadable content on the way.

Re:Too bad Hellgate: London sucked

[Mongoose Disciple]

The game's not bad.

The demo deserves to go in the "Dumb Ass Demo Choices" hall of fame, though. It's the game developer equivalent of purposely choosing clothes that make you look fat.

Re:Too bad Hellgate: London sucked

[smurgy]

Agreed, I was truly disappointed with it. When we were told that "the Diablo people from Blizzard were making what could be considered Diablo III" I was really excited and ready for something above and beyond - instead we seem to have been delivered a badly modelled and textured (I truly found the environment unbelievable) combination of Diablo style mobs, equipment management and levelling with the worst aspects of Elder Scrolls combat.

The NPCs were irritating too. I played a female character and one kept making inappropriate remarks. Far from adding flavour it just plain creeped me out. Realistic maybe, but I don't play games to immersively engage in the thrill that is sexual harrassment.

Fellow Slashbots, prepare to be dazzled!

[Anonymous Coward]

Well, as Samzenpus already mentioned, the name of the book that I read was Level design For Games: Creating Compelling Game Experiences It's about these...levels. Levels... that you design for games... and... improving them... with brain storming... Did I mention this book was written by a guy named Phil Co? And published by the good people at New Riders. So, in conclusion, on the Slashdot scale of 9 to 10, 10 being the highest, 9 being the lowest, and 9.5 being average, I give this book... a ten. Any questions? Nope? Then I'll just sit down.

Level-design after the end of levels

[Anonymous Coward]

A recent Slashdot post suggested that the age of levels in video games was coming to an end, that now game environments are much more open-ended. Does the book discuss how this might affect designs?

Re:Level-design after the end of levels

[AnotherUsername]

IMHO, there is no 'End of Levels' coming anytime soon. The open-ended design does not end levels, it just makes the transition more seemless. Even in 'sandbox games' such as Grand Theft Auto, Bully, and Morrowind, there is still definite breaks in the story, which can clearly be defined as levels. While it may not say, 'LEVEL 1: CLEARED" or something of that sort, it still has tougher enemies, more of them, and allows the player to reach new areas not previously available to them.

The day that there are no new levels in games is the day that storytelling in games ends.

Re:Level-design after the end of levels

[Urza9814]

What about RPGs? I mean, take a look at games like Baldur's Gate. There are no real levels. I mean, there are certain things you must do to unlock certain areas of the map, but you can attempt those at any time. Or not. And if you took that basic style, but removed that limitation, you would have a game without levels. You could try to attack the 'boss' at any time, but you would be very unlikely to succeed without doing the lower level quests first. But at the same time, there are no required quests. If you wanted to, you could just go into the forest and hunt until you get the XP you need.
Basically, look at MMORPGs. Though they don't really tell a story, you could put one in. Basically, 'This happened to you, and because of that you must kill this guy. But you should probably spend some time getting more experienced and better equipment first. Go.'

Re:Level-design after the end of levels

[Anonymous Coward]

'This happened to you, and because of that you must kill this guy. But you should probably spend some time getting more experienced and better equipment first. Go.'

It's "Kill Bill," the game!

Re:Level-design after the end of levels

[grumbel]

Look at EF2000 for a game without levels. It is a flightsim that simulates a complete war from start to finished. You still have missions that you have to fly which one could call levels, but the war simulation never stops and the missions aren't prescripted, all the tiny units are simulated all the time and missions simply emerge out of where the front lines currently are or you can even design them yourself. So there really isn't much of a real start and finish of a level, the difference is only if you sit in your plane or watch the tactical screen.

A different example is Operation Flashpoint, while this one does have clear predefined levels that reset the game world to a known state, the game world is completly open at all times, you can walk anywhere at any time in the game and approach missions objective any way you want. So while the levels are still there, the gameplay emerges for most part from the AI and not from the scripting or restrictions placed up on the player. This causes a mission to be played completely different basically each time you play it.

the cliches of level design

[Speare]

All too often, I have seen games where the level design consisted of the following cliche decisions. Level 1 should be garden-themed, level 3 should let you swim (if you're ever allowed to swim), level 4 should be slippery ice, level 6 should be raging lava which kill you if you touch it, and level 8 should be a screwed-up-gravity level that lets you walk on the ceilings or reorient yourself in space.

What's funny is that these same gameplay decisions are leaking into the storylines of modern adventure movies. For example, the plucky racing scene in Cars, or Star Wars I. Or the sidescroller robot factories in Minority Report and Star Wars II. Or the "jumping on floating bits across lava" scene in... uh, Star Wars III. The transitions in Lord of the Rings from "ice" to "fire" to "water" to "forest" areas actually seem to make sense, but only because they take place over 36 hours of video, or 1600 pages of text. Cramming it into a single game or movie with almost no transition just makes it seem ridiculous.

Re:the cliches of level design

[JoshJ]

Not as bad as the "Mandatory Video Game Minecart Ride". (See: Tony Hawk's Pro Skater 3)

Re:the cliches of level design

[Anonymous Coward]

Kalima will rule da world!

Re:the cliches of level design

[CRCulver]

Or the "jumping on floating bits across lava" scene in... uh, Star Wars III.

That Anakin Skywalker was disfigured in a battle on a lava planet goes all the way back to Lucas' novelization of the first Star Wars film in the 1970s. I do agree that the cinematic realization was inspired by video games, though. Reminds me of this Onion article.

Re:the cliches of level design

[Threni]

> The transitions in Lord of the Rings from "ice" to "fire" to "water" to "forest" areas actually seem to make sense, but only because they take
> place over 36 hours of video, or 1600 pages of text. Cramming it into a single game or movie with almost no transition just makes it seem
> ridiculous.

No, it makes it fun. When I play games I want fun now. Watching lord of the rings is ok if you've got 3 or 10 hours to watch `explore a bit..fight a bit...explore a bit...mystical crap about fairies...fight a bit...`, but if I'm playing sonic or..well, anything, then I want something a little more entertaining.

Re:the cliches of level design

[SQLGuru]

That Indiana Jones game did it first (Temple of Doom)

Layne

Re:the cliches of level design

[192939495969798999]

The movie scenes you're referring to were likely added so that they make sense when they appear in the video game, for people too young (or ignorant) to realize what a video game about the movie "Cars" would probably look like. Anytime you see a "video game-y" scene in a newer movie, you can be sure that 1) there will be a video game based on the movie, and 2) it will feature gameplay similar to the in-movie game trailer you've just been forced to watch. /rant

Re:the cliches of level design

[tieTYT]

The transitions in Lord of the Rings from "ice" to "fire" to "water" to "forest" areas actually seem to make sense, but only because they take place over 36 hours of video

Is that the Director's Director cut version where each film has 9 hours of extra content?

Re:the cliches of level design

[CCFreak2K]

All too often, I have seen games where the level design consisted of the following cliche decisions. Level 1 should be garden-themed, level 3 should let you swim (if you're ever allowed to swim), level 4 should be slippery ice, level 6 should be raging lava which kill you if you touch it, and level 8 should be a screwed-up-gravity level that lets you walk on the ceilings or reorient yourself in space.

That sounds a lot like this game.

Re:the cliches of level design

[JoshJ]

I have no idea which game was first. In THPS 3 if you did a grind down a series of mine-cart tracks, you got a gap (bonus points + combo multiplier) called "Mandatory Video Game Mine Cart Ride". They were poking fun at the cliche.

Re:the cliches of level design

[SirSlud]

What you have describe is "game design". You design a game, and it has X gameplay mechnics by the time you have access to them all.

Level design is a separate discipline entirely. It's about taking the game design, and crafting progression within those limitations.

These cliches you're talking about are more along the lines of game design cliches, which forces the land of level design to introduce gameplay elements at a rate which feels challenging, but not overwhelming, impossibly, or narratively nonsensical to the gamer.

The best level designers can take a game with a very very small set of mechanics, such as portal, and turn it into a set of puzzles that are fun to play while you learn to exploit the gameplay mechanics.

In platformers (which is really the genre you're talking about), there are usually a ton of mechanics, and so its also quite tricky to present them in a way where you don't blow you're entire budget on creating levels that simply bring the player up to speed.

I think Galaxies did a marvelous job of that. The "this is how you do that" moments of the game are absolutely minute compared to the, "oh sweet, I can apply that knowledge in this non-obvious obstacle puzzle" moments. Sure there are fire levels, ice levels, yadda yadda, but game design is building the car. Level design is presenting you with fun places to drive it.

Re:the cliches of level design

[Mr Z]

And here I thought you were talking about Sonic the Hedgehog games at first.

Re:the cliches of level design

[Cygfrydd]

Or the "jumping on floating bits across lava" scene in... uh, Star Wars III.

That Anakin Skywalker was disfigured in a battle on a lava planet goes all the way back to Lucas' novelization of the first Star Wars film in the 1970s. I do agree that the cinematic realization was inspired by video games, though. Reminds me of this Onion article.

... by Lucas, you mean Alan Dean Foster, of course.

Re:the cliches of level design

[BBadhedgehog]

*> The transitions in Lord of the Rings from "ice" to "fire" to "water" to "forest" areas I know, that bloody Tolkien copying the themes in video games. It's the pong rip-off that winds me up the most.

Team Fortress level design

[InvisblePinkUnicorn]

TFC in my opinion has some of the best levels when it comes to design and purpose. There is a delicate balance between providing enough points of attack for the offense while providing enough cover for the defense, all while keeping the level small enough to allow users to quickly and easily get back into battle.

Re:Team Fortress level design

[DustyShadow]

Yea those levels are nice but it would have been nice for Valve to release a few more. The game is quite repetitive with the same few maps over and over and over again. I've played maybe 2 decent user created maps but most of those are pretty weak so far.

Re:Team Fortress level design

[Floritard]

See he's talking about Classic, not the recently released TF2, which means he's probably talking about the dozens of user-created maps.

Re:Nearly two years old

[Anonymous Coward]

The books are reviewed here so the poster can collect a bounty for the referral link to amazon.

Re:Nearly two years old

[Otter]

Your two complaints seem somewhat contradictory...

Multiplayer level design

[Anonymous Coward]

On a side note, one thing I would be interested in seeing would be a level design which would allow more randomness to multiplayer games.

Take counter-strike as an example. Next, look at the Italy map. I haven't played in a while so that is why this is a reference to an older map. In the Italy map, if you are a counter terrorist, you can go down the alley to the left, through the apartment straight ahead, or through the market to the right. Those are the only choices you have and all the paths are always the same. If the level is designed such that the apartment changed interior layout, the market changed configuration, the terrorist house changed configuration, you would have a lot more random choices available.

One way that this could be achieved would be to create sections of the map which would either rotate or move themselves veritcally. For example, the terrorist house is located on the map in a corner. Imagine a house which is actually 4 times as large as the terrorist house but only 1/4 of it is seen on the map. If the house is rotatable, then you could have 4 different terrorist house configurations. Now imagine that the house can move up and down one story with one store in the ground and one in the air above the "sky". If you move the building up and down along with rotating it, you could have many different configurations for just that one portion of the map.

Now do the same thing with the apartment and the market place. Something along these lines would open up the multiplayer experience such that you don't have people camping in the same places every round. You can come up with an experience where the map is changing all the time and each round would be different from the previous round depending upon the random choices concerning rooms etc...

Just a thought.

Re:Multiplayer level design

[Mordicain]

Problem with this is you can only really do it on a map change. Otherwise you'd get people loading 4 times the regular amount stored on the map. On lower end computers it would hurt and you'd have to design the game with less bells and whistles.

Re:Nearly two years old

[nautsch]

Yeah. Because its THE "Anonymous Coward".

What about for Non-Games?

[GreatRedShark]

So this is level design for games... what about level design for non-game situations?
I can imagine using game level design for architecture:

Architect: "So I think we should have lots of cardboard boxes in this hallway"
Sr. Architect: "Why?"
Architect: "For better sniping spots during the shootouts!! And all good levels have cardboard boxes and crates!!!"
Sr. Architect: "WTF are you talking about?"
Architect: "And there should be a flamethrower canister in the Men's room"
Sr. Architect: "You're fired."

hehe... actually, I remeber hearing ages ago that Oni levels were built with architecture tools, not standard level design tools.

Re:What about for Non-Games?

[UbuntuDupe]

Something similar happened for Halo.

Level Designer: "Let's see, after going through this hall, they'll probably be beat up pretty good, gonna need some health."
Writer of companion novel: "Hm, okay, I can work in a dead-marine-carrying-health-packs-in-that-room into chapter 3..."

Re:What about for Non-Games?

[vimh42]

"I remeber hearing ages ago that Oni levels were built with architecture tools, not standard level design tools. I read that too. It's too bad the game was a disappointment. Combat and such was cool but I recall being very bummed when I finally played the game.

Re:What about for Non-Games?

[Chabil Ha']

hehe... actually, I remeber hearing ages ago that Oni levels were built with architecture tools, not standard level design tools. I remember the tools used to create Doom 2 levels looked more like AutoCad tools more than anything.

Re:What about for Non-Games?

[Kjella]

Hehe, you should visit one of the places I worked. The customer areas were extremely formal - forget wearing anything but a suit and tie and they were effectively sealed off from the rest of the building. Likewise there were "polished" areas for external workers. We however worked there so long we were in the back end - and it was "anything goes". Several hallways actually did actually have all sorts of junk and cardboard boxes and shit in them, wear anything you'd want and a cluttered desk policy. It was actually several buildings merged together, so the floor numbers didn't always match and it was a confusing mess. Apart from not seeing a flamethrower canister, it was perfectly suitable as a game level. And it's the only building complex where I'd really like one of those mini-nukes to demolish it too.

Re:What about for Non-Games?

[typidemon]

Having to loot the remains of a fallen soldier is a plausible storyline mechanic. At least compared to "there's some neatly stacked ammo in some random corner".

Re:Nearly two years old

[Otter]

Ummm, this isn't Wikipedia. Everything in a single comment is pretty much guaranteed to be by the same person.

See TF2

[Wizworm]

You need to try out this Orange Box thing, all the cool kids are talking about. Specifically listen to the developer commentary and play a few games on Hydro.

Re:See TF2

[Senzei]

Hydro or Dustbowl. The balance on Hydro is all over the map, some sections are awesome, some are (in my opinion) the worst (barring 2fort, which is just lame) out of everything on TF2. Dustbowl is a little more consistent. Either way the TF2 maps are worth studying, almost all do a good job of having multiple useful paths to important points in the game, and reward those who put in the time to learn the map well without allowing them complete domination because of it.

The Crates of Wrath

[Dirtside]

Pfft. Everyone knows that the quality of a game's levels depend entirely on how long you have to go before you see a crate.

Re:The Crates of Wrath

[omfglearntoplay]

Thank you, very funny. On a more serious note, how much emphasis is placed on levels for multiplayer vs levels for plain old single player? Very very different things going on in those two worlds.

Isn't it time to say goodbye to 'levels'?

[Dutch Gun]

Perhaps I'm just being pedantic, but I'm really tired of the concept of 'levels' in games. I don't want to see loading screens telling me "loading new level..." I'm tired of rocket launchers that don't break wooden doors (or even glass doors. yes, I've seen this) because that would ruin a design's carefully crafted puzzle. I don't want to hear on the back of the box "Twenty Action-packed Levels of Pure Mayhem". Isn't it time the game industry retires this term? I'm fed up with buildings designed like mazes because the game designer simply wants the players to run through a maze. This is what I think of when I hear about 'level design'. I'm sick of levels.

Why do I care about the term 'level' at all? Honestly, I think it encourages game designers, environmental artists (a better name than level designers) and even publishers to think too linearly, and to be intellectually lazy. Publisher: how many levels will this game have? Game designers: 30 levels. Publisher: Fantastic, that's 10 more levels than our competition has. Gamer: yawn...

So what do I want instead? Give me environments! Give me worlds! I want freedom to explore, to find out of the way nooks and crannys, and more than one way of getting from point A to B. I want to solve problems using logic, not by playing "guess what the game designer wanted me to do or go next"? Game designers: Create a living, breathing, interesting world, and then let your players enjoy their time here. Stop shoving the player along a conveyor belt.

Obviously, it's not fair to pin this on the term 'levels'. But it just seems like a term that emphasizes aspects of games I'd love to see the industry move beyond.

Re:Isn't it time to say goodbye to 'levels'?

[gauauu]

Strangely enough, I enjoy the opposite. I want games to quit pretending like they have real worlds in them. The closer they get to looking like a real world, the more annoying it is when you don't understand what you are "supposed to do." I feel like I'm trapped in a fake world.

Older style games where things were much more confined annoyed me less -- they were overtly trapping you, and I didn't get frustrated.

For example, in a 2d side-scroller, it doesn't bother me that I can't break out the windows in the background and go through them. That background is just art, and I can safely ignore it.

But more immersive games blur the art and the world more, and you spend more time wondering if that glass window is something you're supposed to be able to break and go through, or maybe it's indestructible and the developers were lazy. It just feels like, because the world is supposed to seem MORE real, that all the ways that it isn't real and interactive are more annoying than if you just went back to purposely limiting the game.

In that sense, I enjoy game cliches (levels, crates, stupid enemies, etc), because I know what the rules of them are, and don't waste time wondering "am I supposed to interact with this thing somehow?", instead, I can focus on having fun playing the game.

Re:Isn't it time to say goodbye to 'levels'?

[AnotherUsername]

If you remove levels, chapters, parts, sections, and basically any other separator from games, you will also lose all story. How long can you do basically nothing in Grand Theft Auto until you start to get bored and wonder what else is to the world? I mean, if you don't want levels, you will basically just have a game where you run around doing a whole lot of nothing. And when that comes around, game designers will cease having to create 'living, breathing' worlds, and simply create a couple of blocks, fill it with prostitutes, and give you an arsenal of weaponry, and let you have at it. It will become boring very quickly. Levels are simply chapter dividers in stories. Removing them removes the story.

Re:Isn't it time to say goodbye to 'levels'?

[Anonymous Coward]

If you want a living, breathing world, heres a hint: there's one right outside your mom's basement. Get out there and invent your own challenges to test your wits and problem-solving skills. In a short while, you'll easily thing up exciting achievements such as: Sleep With Friend's Girlfriend Without Him Finding Out, Fix Wobbly Screen Door, and, always a favorite this time of year, Find Decent Parking Space At Mall.

Re:Isn't it time to say goodbye to 'levels'?

[Ford Prefect]

So what do I want instead? Give me environments! Give me worlds! I want freedom to explore, to find out of the way nooks and crannys, and more than one way of getting from point A to B. I want to solve problems using logic, not by playing "guess what the game designer wanted me to do or go next"? Game designers: Create a living, breathing, interesting world, and then let your players enjoy their time here. Stop shoving the player along a conveyor belt.

I wouldn't describe myself as a level designer, rather just a mapper (or map designer if I want to sound posh). But I've made the moderately popular MINERVA single-player mod for Half-Life 2, which people keep claiming is quite good - it's got an interesting world to explore (if you can see it in the distance, you can get there pretty much always), and while it's fairly linear it does have a couple of branches, double-backs and efficient use of architecture. It's definitely not a long, solitary corridor with no turning back...

The book review is interesting because the book sounds like it's managing to focus on things which I've never really thought about - I've never decided to build ' Also in my eyes, parts of the list are the wrong way round - don't take the chapters' ordering as some kind of specified route as to how things must be constructed, rather than aspects to the whole process. Theirs:

• "Defining the Game"
• "Enemies and Obstacles: Choosing Your Challenges"
• "Brainstorming Your Level Ideas" ... "delves into the creation of concept sketches and reference images, the creation of a level's storyline, the drafting of a level description and the design of the puzzles and scripted sequences within the level"
• "Designing With a Diagram" ... "the scope and order of levels within the game" ... "lay it out in diagram format by creating a grid"
• "The Template"
• "Improving Your Level,"
• "Taking It to 11" ... "architectural style, the addition of details like trim and borders, the appropriate use of textures and props, and the like"
• "Ship it!"

My utterly awkward route:

• Think of a setting. Metastasis was basically ISLAND!, while the upcoming Out of Time is CITY! - this helps with...
• Think of an ending. This is the most important part of the whole experience, and will define the beginning. For Metastasis, the beginning is basically the reverse of the ending.
• Think of a middle, and of the whole plotline. This defines what the available enemies will be (the selection needs to be plausible to fit into the plotline - no striders 500m underground, or zombies wandering around a tightly controlled and maintained Combine facility).
• Define and build the architecture. Build it appropriate to the setting and plotline, and with an eye on enhancing details which provide interesting gameplay. You know the beginning and end, so you need something to constantly drive the player towards the ending. Give them something to fight for, and they will. Pull, don't push. Metastasis's architecture changes throughout - and tells a lot of the story though how things are constructed. The game is a result of the architecture, not the other way round. I've got some ideas involving architectural styles imparting major themes and plot-points in Out of Time too - except this time it's borrowed from different areas of the real-world Warsaw...
• While doing all that, add the gameplay! In Metastasis, there was a section which I never bothered adding any enemies to - it was far too atmospheric and interesting just wandering through a seemingly abandoned World War Two-era base to spoil. If I'd decided 'NEED COMBAT-FREE SECTION NOW' before I'd bu

Re:Isn't it time to say goodbye to 'levels'?

[Anonymous Coward]

Perhaps I'm just being pedantic, but I'm really tired of the concept of 'levels' in games

Perhaps I'm just being pedantic, but I'm really tired of people misusing the word 'pedantic' in Slashdot posts.

Re:Isn't it time to say goodbye to 'levels'?

[Mr Z]

You won't like my latest game then. Fortunately, there are enough people who do.

Re:Isn't it time to say goodbye to 'levels'?

[Anonymous Coward]

MOD THIS MAN UP ALREADY

Re:Isn't it time to say goodbye to 'levels'?

[the phantom]

The uncanny divide of game design, no?

Re:Isn't it time to say goodbye to 'levels'?

[MrNiceguy_KS]

Just wanted to say that I really loved your first 2 episodes. Hadn't known the 3rd was out, but I plan on downloading it tonight. When I see you describe your design process, it really makes sense that they were designed that way. It seems to boil down to either, "Create a setting and storyline, then fill in puzzles and enemies as fit," vs. "Think of some cool enemies, then design a story around them and figure out where the story should take place." I have a feeling I'll be keeping that in mind when I'm playing games in the future.

Deus Ex

[Scudsucker]

Valve does have good levels, but all their maps are rat tunnels - there is only one path to take and only one way to solve a puzzle. That was fine in the first game, because you're in a damaged underground base so it makes sense that your movement options are limited. Not so much in the HL2 as you spend most of the game above ground in cities or traveling in the country.

That's why I think Deus Ex is the better example. On most of the maps you not only have multiple paths you can take, but there are multiple ways to accomplish your objective. In one level, your mission is to take down a power generator being used by the "terrorists". You can disable cameras and turrets and sneak through a tunnel, climb onto the rooftops and snipe, or you can go in the front door, guns blazing.

It's the players, not the designers.

[C10H14N2]

People like easily recognizable numbers to indicate status. Remove "levels" and how will people get the instant gratification of knowing they're superior without actually having to prove anything? It's like when car manufacturers went from names to alphanumerics. It was hard to tell who was better, the guy driving the New Yorker, Royal, Saratoga or Imperial. But now, obviously, my 500 must beat your measly 300.

Valve may be good...

[jagdish]

but not the king. You should really have a look at this and this.

Re:Valve may be good...

[EvanED]

They learned their lesson with the flashlight though... starting with EP2, sprint and flashlight power are decoupled.

This is one of my favorites from that strip that gives commentary on what you can and can't do within the HL universe.

Alternatives?

[coldcell]

While some of the content does seem largely design-based and awesome, the inclusion of UnrealED chapters to think about 3D alone seems limiting for the HUGE range that "level design" could cover.

Is there anything that offers up more of the design discussions, without any of the specific (or at least just a lot more varied) implementations.

I'm thinking MUDs, Space Games (no physical obstacles, many degrees of freedom), and D&D levels. Anything out there that could apply to such a range?

Re:Nearly two years old

[Anonymous Coward]

If you read the comment referred to, the (single)poster requested that the book not be reviewed here until there were more book reviews. Well, you can't have more reviews until you post them. Why wait?

Uneven Level Design Primer

[Anonymous Coward]

"It's almost as important to understand bad design as it is to understand good design."

Why would I want to play through slashdot?

What about for Architecture?

[Anonymous Coward]

"So this is level design for games... what about level design for non-game situations?"

You do know that the latests Crysis engine is being used by an architecture firm, don't you?

What I'm Waiting To See Implemented In A Game....

[pandrijeczko]

... Duke Nukem Forever.

Re:What I'm Waiting To See Implemented In A Game..

[Anonymous Coward]

There's only one thing that differentiates man from the animals - we're not afraid of vacuum cleaners.

I don't know why, but when I read that, I heard it in George W. Bush's voice. I don't know what that means, but it means somethin'...

Some reflections

[Preacher77]

In the "good 'ol days" of Quake I was fairly involved in the multiplayer side of leveldesign (Aerowalk) and what strikes me is that there's different kinds of leveldesign these days. You have the "casual"-gaming target audience and you have the "e-sport" audience, and designing for the them is completely different.

For the casual gamer, immersion, not getting stuck, being apropriately difficult etc etc is key. For E-sport (or hardcore gamers) leveldesign needs to be challenging, have a learning curve be really balanced. Immersion and nice architecture isnt always as important since many (atleast in my time) turn off all effects etc to get better performance. I remember spending hours aligning textures only too see players playing it tweaked to show only flat surfaces, aaaargh :)

Another issue is that designing competitive levels is really really hard, it's something of a miracle and blind luck that some of the original quake levels were as good as they were. Hardcore gamers will find new tricks, new possibilitie, shortcuts etc etc. That means that optimally the designer needs to be as good (or in the vicinity ) as the competitive players to be able to understand and design for them, and finding people that are both great designers and great players isnt easy :)

Nowadays realism and graphics in first person shooters seems to be everything, atleast it's what sells. It's a bit unfortunate in my opinion because there is a place for games that are games, and not real world simulations. Sometimes not having totally realistic physics makes for a better and more skillful game, in quake 1 for instance you maneuver while in air. Not realistic but made for some awesome tricks.

Graphics in itself arent a bad thing of course, the bad thing is when graphics influences game design too much. That you put in 20 "cool" weapons just to be able to show off different effects, rather than putting in 5 basic but unique weapons that you can master. And remember in the time of Amiga, C64 etc where you didnt have great graphics and game contrent but really unique game ideas. When one programmer could make a somewhat successful game and a small team definitively could? Nowadays you need a 100+ team and a hollywood budget :)

Anways, I realize that I sound as an old fart bitching about the good 'ol days so I'll stop now :)

Re:Some reflections

[omfglearntoplay]

Excellent post. I find myself closer to the e-sports crowd on my favorite games, and I take map making very very seriously for multiplayer player against player like in Warcraft 3 or Starcraft or Unreal Tournament (including the newest). Lately I've picked up Team Fortress 2 and it is another instant classic. One thing that's bothering me, while I'm on the subject, is the way that some games apparently don't cater enough to the general crowd to get a huge following. The ones I named do a pretty good job of that, and that keeps them fun. But now that xboxes and the like are trying to compete with PCs, it's breaking up the crowd even more. Causes some grief. Thank god Starcraft 2 won't be on the Xbox (right?).

Craft of the Adventure

[the_arrow]

While tailored to interactive fiction, the "book" The Craft of the Adventure is a good read for many people wanting to design levels even in modern games. Especially section 3, Bill of players rights, I think is valuable even today.

The Craft of the Adventure can be found on the IF-archive. While there, another good read is the authorship-guide.

Re:Nearly two years old

[k_187]

This is horribly off topic, but when I first started reading /., I was amazed at all the comments that anonymous coward guy was making.

Hardcore players

[IdeaMan]

I designed several Warcraft 3 maps for the hardcore audience incl. Traps & Towers TD. I did mostly Tower Defense type maps, but I did work on a couple of RPG's. What I found is that play-testing the maps takes humongous gobs of time, but was pretty much the only way to significantly balance the levels out. It is important to get feedback from other players, listen to their complaints, but you need to remember that they Want it to be challenging. It follows that the game will have a higher replay value for them if they lose in a way that is preventable (given more time, knowledge, skill).
My key points to level design:
Unlock new skills at appropriate times.
Increase the difficulty of the tasks in such a way as to require the player to use his skills in a synergistic manner. In the later levels you need to assume that the player has mastered the basic skills, and design the game such that the player is experimenting with different combinations of skills.
Each skill should be tuned to work best under its own set of circumstances. For example if the skills are guns, then one gun would be a sniper rifle, another a shotgun. Each has their uses, each can be considered equally powerful in their prescribed usage, but each would be a poor substitutes for the other.
For Co-op play, set it up so that playing together is about 1/3 easier than playing solo. The reasons for this is that communication between players is not 100%: if you made it the same difficulty the players would end up letting each other lose more often than playing solo. If it were 100% easier then it wouldn't be challenging enough. You gotta aim for the players losing slightly less often than playing solo.

In order to satisfy all those, it takes tons of feedback and play-testing. Keep a notepad handy when you play the levels, and jot down notes to remember which parts you need to fix. Then you won't forget to fix them when you're working on the level.

One thing I wish I had added to my maps were a change of pace. The other maps had little mini-games to give the players a break in the action. Extra puzzles, different types of games, levels without any monsters, switching between co-operative and deathmatch all are methods to give the players a break so that when they come back to the level it's fun again.

Oh and about realism being the be-all and end-all, have a look at the (unrealistic) atmosphere in Team Fortress 2, it's AWESOME!!! Only drawback is that TF2 is a dumbed down playskool version of TFC.

- EMPY