The 3D Realms Vault: 1994 Design Tips from Tom Hall - Part 1

The 3D Realms Vault: 1994 Design Tips from Tom Hall - Part 1

The 3D Realms Vault: 1994 Design Tips from Tom Hall - Part 1

Join us as we dig into the 3D Realms Vault with never-before-seen notes, articles, and material from our 30-year legacy. Our first post covers early design tips from the one and only Tom Hall.

In the early nineties, the legendary game designer worked on classics such as Commander Keen, Wolfenstein 3D and Doom as the Creative Director of id Software.

In 1993, he joined 3D Realms (then known as “Apogee”) to develop Rise of the Triad and support development of the Duke Nukem series.  He had his hands in several Apogee/3D Realms titles then, from Duke Nukem II to Terminal Velocity, Rise of the Triad, Hocus Pocus, and Prey.  A complete list of what Tom did would be longer than this document, but a few selected highlights are worth noting.  Specifically, the deathmatch design, some laser gfx, and the story from Terminal Velocity.  Tom was also responsible for a lot of the "character items" and the story for Duke Nukem II, including the famed merchandise items, which were part of the foundation of the eventual "character" of Duke as we knew him in future titles.

In 1994, there were few rules as to how games were designed. As such, Tom was one of the fathers of today’s design laws and philosophies.

Below is Part 1 of a never-before-seen memo from Tom, in which he shares his game design tips and tricks.


This document is for Apogee developers only.
Copyright © 1994 Apogee Software.


"These kind of "tricks" also make it very hard on "tech support" which makes them want to "kill" the "designers" with "claw hammers." You can quote me on that."

-- Never have treasure you can't get.

-- Never have an area the player can get into that they can't get out of.  The player should never have to kill themself or quit.  The only exception  is a puzzle game, where you can get stuck.

-- Don't throw all your puzzles/tricks at the player at once.  Introduce a new one, let them learn it, let them experience it in combination with all the previous tricks, then introduce another new one.  Later on, add new twists on the old tricks that allow them to use their knowledge in a different, but somewhat familiar, way.

-- Don't make all areas look the same on a level.  The player really needs reference points.  The fun of a game is the challenge, not getting lost.

-- Never require secret tricks or paths for the solution of a level.  The challenge should be there because it is challenging to do, not because, for example, you didn't jump backwards through the third pillar, which, ha ha ha, isn't solid.  These kind of "tricks" also make it very hard on "tech support" which makes them want to "kill" the "designers" with "claw hammers."  You can quote me on that.

-- Never require trial and error where a learnable or logical solution is possible.  It should be "I need a fuel pack to use that machine. Hey, maybe I can use the cell in my plasma gun." not "oh, I was supposed to throw the pie in the face of the polar bear.  Of course. That always happens."

"oh, I was supposed to throw the pie in the face of the polar bear. Of course. That always happens."

-- Don't make the player go all the way to the end, then all the way to the  beginning, then all the way to the end again.  It's more fun to feel like you are making ground.  Have a series of areas to get through, rather than one big awful area. 

-- Never draw words in a map, unless they are very hidden.  Initials are much better because there are small, rather than spelling out, "Hello" to make a level.  You can hide your initials once, but don't take shortcuts to design levels.  All that room could be used for fun, good design.  This also goes for smiley faces, genitalia, and so on.

-- If you have halfway point markers that the players return to after getting killed, make sure the level can be completed from that point. It's okay if they can't get to secret areas (like in Super Mario World, where you need the cape to get to a secret area in the air), but they must be able to complete the level.




-- The background colors should be muted, and in contrast to the foreground characters.  If you can't see the action, you can't enjoy it.

-- Keep a constant light source (usually from the upper left or upper right) for the game--not just in the game graphics, but in the menus and endscreens, too.

-- Things that act different should be at least a little graphically different.  If you can blow away a pillar, it should be cracked a little, so it looks different from the others.  If a bridge piece gives way, it should be missing a pin or something.

-- Don't use all your graphics in the first episode.  There should be a lot of cool graphics to look forward to.  Otherwise, the user feels cheated.  "It's just more of the same."  (Plus you can put the cool graphics in the teaser screen shots.)



-- Make sure the actors all do different things: walk, fly, shoot, climb, steal, give, etc.  Provide as many different acting actors as possible. 

-- Here are the most important aspects of a game:

1. Playability--if the control doesn't feel right, gamers will not play the game. It should feel natural, it should do everything you'd want to do at any point, and it shouldn't be so complex that you can't remember what to do.  Also, it should work decently on each type of input device you support. If you're barely supporting something, don't bother.
2. Variety--if there isn't enough variety, the player will get bored doing the same thing.  However, don't go overboard--the player has to learn and appreciate the new things before the next ones are introduced.  And the variety should fit within the theme and setting of the game.  Just because you have a great kitchen sink, it doesn't mean it belongs in the game.  Save it for the next one.  This could be called "Variety with Coherency."  Having a good story behind a game will provide a rich source of ideas for actors and settings.  The story provides a framework for the game's world to be built on and designed from. It's not really important that the user reads the story--in fact, long stories are not welcome--but the story is important for a) building a coherent world and b) providing a unique wellspring of ideas, producing concepts and actors unlike other games.  The former makes the player feel the game is really "together" and allows another of their layers of disbelief to fall away--since the world is "behaving" correctly (even if it is chaos), they allow themselves to get deeper into it.  The latter is what many designers miss.  There's so much you can do with each subject, but many designers tend to have a wild idea, then implement it just like the last game they liked in the genre.  You want something new in your game, so it'll stand out.  A story is just a handy tool to make the game unique and interesting.

"There's so much you can do with each subject, but many designers tend to have a wild idea, then implement it just like the last game they liked in the genre. You want something new in your game, so it'll stand out."

3. Graphics--sight is the main sense, and a game must be visually appealing to succeed.  Attention should be paid to all the graphics--in the menus, the font, and so on.  And it should all be coherent in style, a style appropriate to the subject of the game, not just, "Hey, I got this cool font!" Watch out for the "angry fruit salad" problem--using all the colors all over rather than colors that look good together.
4. Sound/Music--sounds should be pleasant and appropriate, and bear a lot of repeat listening.  Be wary of voices, because they can be the most annoying.  The music should enhance the game, not try to overwhelm it.  It should create the best mood possible for the action to take place in.

5. Challenge--this must be very carefully balanced.  It must be sufficiently tough to make the player want to keep trying, but not too hard so the player gets frustrated quickly.  Having different skill levels is helpful, but you still have to balance the game so it is progressively harder.  It should be fair.

6. Robustness--every part of the game should be crashproof, and it the case of irrecoverable errors, it should exit gracefully, providing as much information as possible, in a friendly format.  This includes cheats, because if a cheat window overwrites something, you can never know for sure if a problem is with the game, or because you cheated and the cheat messed up memory.

And most importantly: we are here to give the player a great time, not to laugh at them.  We can scare them, and put them in a dangerous situation, but we should never take on the attitude, "Well, too bad!  You jumped there! Ha ha ha!"  There should be a childlike joy in game play, not a childish "I'm better than you" attitude from the designer.  Someone with perfect reflexes should be able to finish your game without dying.  In a nutshell, never screw the player.  You're only screwing yourself out of their money and respect.

Join us next week in Part 2 of the Tom Hall Design Tips!