The Creation of Q*Bert

By Warren Davis, as told to Coinop.org

Introduction

I am constantly amazed at the fondness people have in their hearts toward Q*bert. I was glad it was successful when it first came out, but I never expected it to be remembered after this many years. It's an honor to have been able to touch the lives of so many people in a positive way, and for that I will always be grateful.

Q*bert was the love child of three people: Jeff Lee who did the graphics and created all the characters, David Thiel who did the sounds, and myself. I was the game's official designer and sole programmer. Back in those days, programmers generally tended to be the game designers since they had to actually implement the game. Many people contributed ideas while we were developing Q*bert, and I had to be the filter who decided which ideas were worthy and which were not. Since Q*bert was my first game, and started out as a programming exercise, "worthy" usually meant "easy to implement."

Jeff created most of the characters in Q*bert before there was even a game. I'm pretty sure Jeff also gave all the characters their names, with the exception of Q*bert himself who wasn't named until the game was completed. (More on that later) Dave created all the sounds, many on his own, some in response to my suggestions. Dave was the one who came up with the idea of using the random phonemes of a speech chip to make Q*bert speak gibberish. Contrary to modern myth, he is speaking only gibberish. Nothing was ever programmed to be said specifically, with the exception of "Hello, I'm turned on." when you power the game up and "Bye-bye" at the end of a game. All three of us constantly gave each other input during the creation of Q*bert, and we all enjoyed and respected each other's ideas which is what made the development process so much fun.

The Creation of Q*bert

I was new to Gottlieb in 1982 and had "learned the ropes" by helping one of our programmers with his game. I was looking for a game of my own to do when I saw that another programmer, Kan Yabumoto (who would later do Mad Planets), had filled a screen with hexagons that consisted of 3 differently colored diamonds. If you chose the colors right, each hexagon seemed to be a 3-dimensional cube. Kan had filled the screen to its edges with this pattern, but for some reason, when I looked at it I envisioned a lot of the hexagons removed so it looked like a pyramid of cubes floating in space. Then I thought of balls bouncing down the pyramid. This was really a thought of convenience, since every time a ball landed it had two choices of which way to bounce. Two choices means 1 bit and that meant in one byte I could determine a ball's path. It was a purely scientific endeavor since I wanted to learn to program randomness and gravity. It was a nice exercise and nothing more.

Once I programmed the pyramid with balls randomly bouncing and falling, people started to comment on how cool and 3-dimensional it looked. So I kept playing with it. Jeff Lee had created a bunch of characters thinking that they might possibly be used in a game. I thought the orange one with the big nose would make a good player character (he looked kind of helpless.) Jeff had originally intended that he would shoot out of his nose. He even wrote up a game description and called it SNOTS AND BOOGERS. However, I liked the idea of the player hopping around trying to avoid bouncing obstacles. Also, I was trying to do something a little different, and there were a lot of games then that involved shooting. Still, a few people tried to get me to have the character shoot. I resisted, but I think there were people at Gottlieb who always thought I would eventually put that in. Because of that, the name Snots and Boogers was sometimes attached to it, mostly as a joke, but the proponents of flying snot-bombs were ultimately disappointed. My name for the game during its development was "The Cube Game". I know, pretty boring. The name Q*bert came much later and is a story in itself.

In implementing the player character, it seemed natural to me that you'd want to move the joystick along diagonal lines, since every time you jump, you're either jumping up or down. So I implemented it that way. But as people came along to play it, they got very confused. I'm still not sure why. They wanted Q*bert to be able to hop from side to side. I stuck to my guns on that one too.

One night, I was sitting at my desk playing with what I had so far, which was Q*bert hopping around the pyramid avoiding balls. It was fun, but it wasn't a game yet. There was no goal, and no way to move from round to round. Sitting behind me was Ron Waxman, our VP of Engineering. Ron can be a very intimidating person, but once you get to know him he's a pussycat. He would do this thing occasionally at night--sit behind someone and just silently watch them work. It's a little unnerving, but I had gotten used to it. So I was just playing with what I had, wondering what to do next. The entire development of the game was very much like that. There was never really a master plan. I would implement something and then start to think... "Ok, what should we put in next?" Out of nowhere, the voice of Waxman behind me said, "What if the squares change color when he lands on them?" This struck me as a particulary brilliant idea, and that is the moment when Q*bert actually became a game.

Q*bert was written in 8088 assembly language and was started with a framework supplied by Tim Skelly for the game Reactor. When we started the game, we were working in Bensenville, Illinois, in a "think-tank" kind of environment. The offices accounted for only a small part of the building we worked in. It was mainly a manufacturing facility, but it wasn't yet being used. Howie Rubin, our VP of marketing, used to come through the offices yelling, "OK, stop working everybody! Time for football!" and we'd go into the plant and toss a football around. This had to be the coolest job ever, largely thanks to Howie and Ron.

At some point during the development of Q*bert, we were moved from Bensenville to Gottlieb's main pinball plant in Northlake, IL. We used the first available IBM PC's when they came out. The floppies held 180KB and the hard disk (when it was finally available) held a whopping 1 MB! The game program was downloaded to our own hardware through the PC's bus. Where the arcade games had ROM, the development system had RAM. I seem to remember that images still had to be burned in ROM all the time.

We actually started on an Intel development system called THE BLUE BOX. (I think that was our name for them, not Intel's.) It used, if you can believe this, 9 inch floppy disks. They looked exactly like the flimsy 5 1/4 inch disks which are all but history now, only bigger. The BLUE BOX itself was big, about the size of 5 modern PCs stacked on top of each other, and blue. Kind of pretty actually.

For a while I had a monitor sitting on a table with some wires connecting it to the game hardware. The joystick was mounted to the bottom of a plastic bucket. (The bucket sat upside down). I've got some pictures and videotape of this somewhere. This may have been why people had trouble with the joystick being at a diagonal. Because the joystick was attached to a bucket, they kept trying to straighten it out! Eventually, as production increased, game developers got actual game cabinets to develop with, but when Q*bert was started, Reactor (Gottlieb's first in-house video game) wasn't rolling off the production lines yet.

I'm still amazed at how great all the characters in the game look. This is considering that Jeff had to plop down each pixel in each image one at a time. There were no drawing tools like the ones we have today. Each character was a 16 by 16 pixel grid, (Coily was actually made up of two 16 by 16 blocks) One of the little things Jeff did was make an image of Q*bert with his knees bent so he actually absorbs the shock as he jumps from cube to cube. This kind of thing, though small, helped the characters come alive.

The same can be said of the game's sounds which contributed enormously to its appeal. Dave's work was just brilliant. From the pitiful sound of Q*bert plunging off the pyramid, to the clinking sound the game makes when you drop a coin in the coin slot, Dave was a master at using sound to enhance the game experience.

One other thing should be credited and that is the knocker which would bang the cabinet after Q*bert disappeared from the bottom of the screen after a plunge. This was the idea of Rick Tighe, one of our technicians. He put a standard pinball coil in a test cabinet so that I could trigger it at the appropriate time. I didn't actually like the knocking quality of the sound. I wanted more of a thud, so it would sound like a sack of potatos being dropped. With a little experimentation, we found that a small piece of foam glued to the right spot produced what I thought was the perfect thud. Unfortunately, the gluing of the small piece of foam was deemed too labor intensive for production and wasn't done. But the effect was still fantastic. The knocker was controlled by a dip switch, so an arcade operator could turn it on and off.

I think the programming of Q*bert started in April of 1982. By June or July, it had evolved to the point where it was put on the schedule as an actual game. The game took about 4 months to complete after that and production began in October or November of 1982. When the game was on test, we would go to arcades and watch people play it. Most people thought it was cool. Some didn't. Some people found it natural and easy to play. Some didn't. Almost everybody hated the fact that you could jump off the pyramid and die. Some people would put their quarter in, jump off the pyramid three times in succession, and not even realize why their game was over! (Granted, not many) Normally, this would be grounds for going back to work and immediately changing the game so you cannot possibly die without knowing why, but I didn't do that. Because for some reason, those people saw other people playing the game with ease and came back to try again.

Tweaking the game was extremely difficult because although some people just got the knack of it quickly, most people at Gottlieb thought it was too fast. So I kept slowing it down so it started slow enough to allow people to learn it. Unfortunately, once it was released, those who got the hang of it were able to stay on it for hours at a time. This was one of the great aspects of the golden age of video arcade games... the ability to stay on a machine for many hours for 25 cents. Many people aspired to do this and were treated like heros when they succeeded. Alas, those days are gone.

How Q*bert was Named

As best as I can recall it, we had a meeting to name the character, since I resisted coming up with a name. Management saw licensing possibilities and wanted something cute. Howie Rubin, VP of marketing, and wanted the name of the game to be what was in the cartoon balloon when Q*bert dies, the gibberish swearing. Most everyone thought this was the stupidest idea we'd ever heard. How are you going to talk about it when the title is unpronouncable? But he was adamant, and a bunch of engineering samples actually went into arcades with that cartoon balloon in the marquis as the game's title (I have one.) After this test, Howie seemed to agree that the game should have a real name and it should be the name of the character.

We passed around a list to everyone in the company to suggest names for the player character. We got about fifty. The only one I remember is Arnie Aardvark. I don't know why I remember that one because it seemed so absurdly inappropriate. In fact, none of the suggestions on the list really worked. So we had another meeting and someone was playing with the word CUBE. Someone else suggested HUBERT since it was a real name and rhymed with CUBE in a way. Then someone else had a brainstorm and wrote CUBERT on the blackboard.

Everyone kind of liked the sound of it and we were all sitting there nodding and getting used to it, when I think it was Rich Tracy, our art director, who rewrote it as Q-BERT and somehow the asterisk got stuck in there and we all knew we had found it. It felt right.

It's hard to imagine a bunch of grown men sitting around a conference table throwing out ridiculous sounding names at each other, but that was the reality of it. A lot of times, in the midst of meetings, I would take myself out of the discussion and listen to it as if I was a complete outsider and I'd have to laugh at how goofy everything sounded. It was a fun time.

Faster Harder More Challenging Q*bert

I still have one of the original engineering samples of Q*bert along with a version of the program which was never released. It was an intended sequel called Faster Harder More Challenging Q*bert. It's basically just that. The discs move so you have to time your jumps. After a few levels of Coily, you get a large female Q*bert-like creature called Q*bertha who "wants" Q*bert much as Coily does. There are some other harder features which kick in after a couple of levels and the game is just faster from the start. There is also a bonus round where tons of Slick and Sam guys fall and you try to jump on as many as you can in a certain amount of time. You also want to let them change as many cube tops as you can and you score points for those when the round is over.

It was intended to be a follow up to the original Q*bert. It is faster, harder and more challenging, but they tested it about six months after the original came out. Most people were still trying to get the hang of the original even though some mastered it to the point of spending hours at a time on it. It didn't test any better than original Q*bert (no surprise given the timing) and they scrapped plans for the enhanced version. I still have it on my machine at home, though. It is my personal Q*bert of choice.

I was asked to make a Q*bert sequel, but I wasn't particularly interested. Neil Burnstein came up with an idea which he made into Q*bert's Qubes. Gottlieb also did a Q*bert pinball game, which I haven't seen in years. (I can't remember the name of it.) I've only seen a little bit of the SNES Q*bert 3, although I know Jeff Lee worked on it. I thought it looked and played very much in the spirit of the original.

I was very honored to see Q*bert have a big screen close-up in the movie Moscow on the Hudson. This may have been due to a rumor I heard that Robin Williams was a big fan of the game and would go around imitating all the Q*bert sounds. There was a connection there since Gottlieb was owned by Columbia Pictures and Moscow on the Hudson was made by... you guessed it.

I used to have a lot of the Q*bert paraphernalia (i.e. junk), but much of it has fallen into the crevices in my closet which lead to another dimension. Too bad, since I had to pay for most of it with my own money.

--Warren Davis