Just about every year, Steve and I coordinate a great from the gaming industry to visit our day job and give the team a day-long talk on game design. We like to focus on classic game designers because we firmly believe that their experience in creating an entire industry from scratch will benefit our game creation process. If it has been programmed, they have programmed it, if it has been created, they have created it, and if it was a mistake, they have learned greatly from it. We have had Chris Crawford and Bill Kunkle visit in the past. This year, we had Rob Fulop join us for a lively discussion on games and online communities.
If you don't know Rob's name, I'll fill you in a little on his experience. He was a game programmer and designer at Atari where he started by programming arcade conversion: Night Driver (Atari 2600), Missile Command (Atari 2600), Space Invaders (Atari 800). He went on to Imagic where he created arguably the best 2600 game of all time, Demon Attack. Wanting to start creating games that didn't only rely on shooting, he created Cosmic Arc, Fathom, the unreleased (but soon to be in limited release) Actionauts, and Cubicolor (another recent limited release from the Imagic Days). Actionauts did see the light of day as a C64 release. When AOL was first starting in the mid 80's, he created the worlds first online gambling game, Rabitjack's Casino. After creating some titles for the ill fated CDi, he created both Sewer Shark and the controversial Night Trap for the Sega CD. The furor over that game caused him to create the nicest, most enjoyable game he could think of, Dogz. Coming out at about the same time as the Japanese toys with the same premise, he virtually invented the computer version of a virtual pet. His company, PF Magic (Pure F*cking Magic) was sold off to The Learning Company in the 90's. He currently focuses on online titles, social interaction games and consulting.
Rob is full of stories from the classic era. One of the best is how he got his job at Atari. Be was in a backgammon tournament, and his opponent happened to be Nolan Bushnell. He really wanted a job at Atari, so he proceeded to try to let Nolan win. Backgammon is not a game that lends itself to trying to lose, so Nolan was amused and befuddled by what appeared to be his opponents attempt to lose the game on purpose. Sensing that he was choosing the wrong path to employment, Rob went on to win the game, and convinced Nolan to hire him anyway.
On the title of his company, he said : We called it Pure F*cking Magic because that was our response when people ask us how our technical wizardry was accomplished! It was also the only comment Rob would leave in the code of his Atari 2600 games.
I don't want to give away's Rob's farm, as he usually gets paid for the information I am about to give you, but with his permission I would like to give you a taste of his game design ideas. My hope is that you will be tantalized enough to invite him out to your game shop and have a fun day of discussions with this game industry great. These are not verbatim, but my interpretation of some of his ideas.
Rob Fulop's Game Design Essentials, as Paraphrased, summarized, and added to by Jeff from 8bitrocket.com
1. All games need a premise of some sort.
Without one, there would be no reason for the game to exist and it wouldn't be much fun to play.
Common premises are: Treasure Hunt, Find Secrets (or keep secrets), go fast, clean up, chase with the knowledge that you can turn the tables, fight (any kind of head to head competition where one much fall), etc. Without a premise, you really just have an activity or a puzzle. Not that activities or puzzles are bad choices for online entertainment (Sudoku, and Barbie Dress up are two great examples).
Pacman is fun not simply because you are being chased, but because you can turn the tables and chase back, Most 'extra' weapon space shooter games are fun because of this simple idea: The bad dudes may have the edge at the beginning, but you can attain enough power to get them back at some point.
The premise of racing games is simply to go really fast. Games can have more than one premise, so a game like Burnout 3 has more than one: Go Fast, and break things.
Mario games don't simply have exploration as a premise, but also they are a supreme treasure hunt. You collect coins, find secrets, and rescue Princess Peach. It was highly successful because it combined a number of basic premises into one awesomely complete game and it revitalized and entire industry to boot.
I think it might sound obvious to most that a game needs a premise, so the concept is not overly revolutionary, but the message is this: Sit down and think about the premise (or multiple) before you do any other game design. Without a solid premise, you probably will not have a good game.
2. Design Risk
By choosing a tried and true concept, you limit design risk. The more unique and untried concepts you put in a game, the more design risk you incur. You tend to be a marketing person's dream if you simply copy what was successful before (and who wants to be that?). The key is the balance between tried and true design decisions and new concepts. A space shooter, for example, was done 1000 time before R-Type was created. R-Type, though, took a tried and true concept (Space Invaders), turned it on its side, scrolled it, and added in massive bolt-on weaponry the likes of which had rarely been seen in a game of this type.
It is very very difficult to think of a brand new game type - Most puzzle games of today are Tetris games, most shooters are Doom clones, most RPG's are Dungeon Master, and most RTS games have Dune II and Command and Conquer to thank for their success. Who would have thought Brain Age, Guitar Hero, or Cooking Momma would have been as great as they are. I am sure the first guy/gal that pitched those games was laughed at by a marketing genius who wanted another FPS or RTS. So, The second key to making a good game is to decide on the level of design risk you are willing to take before you start creating the game.
3. Game Characters
One of the most important parts of a game is the main character. Rob cited the latest version of Linerider, where the rider has been programmed to react to his line riding actions with verbal and physical cues, as a great example of giving in-game characters the ability to create a relationship with the player. The response that virtual characters give to user about his/her environment (even breaking the wall between game and user) makes the character more real to the user. When you begin to design a game character, you should make sure that the character has some sort of identifying characteristic(s) - like the 7 dwarves, or characters on Gilligan's island: Each had one defining characteristic. Prominent in-game characters (player and NPCs) should have at least on define characteristic. Reactions to the environment should be believable and based on this/these characteristic(s).
Note: The above is not straight from Rob's mouth, but my interpretation of what he said. Consider them my design-risk adverse version of his game design ideas I don't want to give Rob's entire discussion away, so if you are interested in what he had to say, even if you disagree (especially if you do), I encourage you to visit his site, email him, maybe even invite him out to discuss game design with your team. It will be well worth your time.
I recommend visiting his site http://www.robfulop.com to check out more details on his games and his consulting services. His is a wealth of knowledge on game design, and all current and future game designers can learn a lot from his experience.