In his retrospective at last month’s Game Developers Conference, John Salwitz looked back at the creation of coin-op classic, Paperboy, and working for Atari in the 1980s.
John Salwitz joined Atari in 1981 as a programmer, spending ten years with the company and working on games including 720º and Klax. Much of this work was alongside artist and designer, Dave Ralston. It was, he says, one of the best jobs he ever had.
“We worked at one of the few centres of video game making on Earth. And the games were all made within spitting distance of each other. You could, and would, regularly walk out of your lab and go to an adjacent lab to play somebody else’s game,” recalls Salwitz.
“This was an incredibly creative atmosphere and you could learn about what it would take for your game to be successful, and frankly get a little intimidated in the process. I loved it.”
One of the first titles Salwitz and Ralston worked on together was an unreleased arcade game called Akka Arrh. Following a poor field test – where the game was outshone by Eugene Jarvis’ Robotron, Salwitz thought his time at Atari might be up. However, Atari executive, Dan Van Elderen, instead said: “Great try for a bunch of rookies. What are you going to do next?”
It was Ralston who first originated the idea for Paperboy, coming up with the concept during one of Atari’s regular brainstorming sessions in 1983.
“Dave showed us this idea for a kid riding down a street, delivering papers on a single transparency for an overhead projector,” Salwitz remembers.
At the end of the presentation, Ralston was awarded three ‘attaboys’ from the Atari leadership. Praise indeed, and an unofficial nod to develop the concept further.
Salwitz was initially reluctant to work on the game – preferring the likes of Missile Command or Asteroids – but he came around to the idea and joined the development team.
“At Atari, to get a game started officially, you had to survive what was known as an initiation meeting,” Salwitz says.
“You would present a thin game description, some idea of your hardware, and maybe a schedule.”
Based on this more detailed presentation, Paperboy was quickly approved, and development started almost immediately.
Paperboy’s hardware was designed another long-time collaborator, hardware engineer Doug Snyder. Linda Sinkovic joined the team as technician, and worked with Snyder to debug and improve the arcade hardware throughout development.
“A technician back then was just as much a part of your team as a build engineer now,” says Salwitz.
“Unfortunately, and not to the credit of Atari, Linda was laid-off not once, but twice during the development of Paperboy. She was, however, absolutely instrumental in the making of the game.”
At the heart of that hardware – which later became known as the Atari System II – was the DEC T-11 microprocessor. It was basically a PDP 11 on a chip, and after working with the 8-bit 6502 microprocessor, it was a significant step up for Salwitz.
One of Paperboy’s biggest innovations was in its display technology. At the time, typical raster-based arcade games were displayed at a resolution of 320×240 pixels. Doug Snyder developed a new ‘medium resolution’ monitor that could display 512×384 pixels.
“I suppose, compared to high-definition, and certainly to 4K, this is downright quaint, but to us at the time it was awesome,” says Salwitz.
This innovation extended to the game’s control scheme. Atari often pressed its development teams to try out new ideas for its arcade games, and Paperboy was no exception with its distinctive bicycle handlebars.
The ancestry of the controller can be traced back to through 1983’s Star Wars, Atari’s Bradley Trainer military project, and Battlezone. In fact, the Star Wars yoke controller – with a pair of red handlebars glued on – was used for the initial hardware prototypes.
“Through this controller, we were able to prove that you could control the motion of a kid riding down the street,” Salwitz says.
“The biggest iteration was the ever-shrinking width of the handlebars. This was because you could apply an enormous amount of torque on the controller through the handlebars.”
There was also an issue with some of the buttons not being visible to players of a shorter stature, such as children. This resulted in some people playing the game without ever throwing a news paper. Not an ideal scenario for a game called Paperboy.
“This we fixed by putting the Start buttons on the controller itself, so that at least you had to find the button to start the game,” says Salwitz
Paperboy’s distinctive isometric design was inspired by another arcade game, Sega’s Zaxxon. However, significant changes had to be made for this format to work.
“Zaxxon presented us with a bit of a problem because it is a 60-degree isometric perspective,” explains Salwitz.
“First, it made scrolling for us more complicated because we were scrolling at an odd angle to the play field hardware. That meant drawing was a lot more complicated for us. Also, it meant that we would be showing just as much of the sides of the houses as we were of the front of the houses. The sides are not playable surfaces so it was a waste of screen real-estate.
Finally, the handlebar controllers sat at an unintuitive angle for players, meaning that if you pushed the controller forward, the 60-degree angle of the paperboy on-screen felt unnatural. The answer was to adjust the angle slightly.
“We crushed the X-axis and brought down the horizontal, parallel to the bottom of the screen, and also to the top of the screen. This solved few problems for us,” Salwitz says.
This made the controller parallel to the front of the houses, and the kid was now only 45 agrees off-angle. It felt better, and it also made it easier to aim newspapers.
The initial prototypes featured a much stranger cast of characters than those in the final product. Street traffic included rolling pianos and boats. There were ginormous snails in the yards of houses, and ducks in business suits on the sidewalks. The response from early focus groups to these designs was not positive.
“It did not go well,” recalls Salwitz. “We’d built a game that did not resonate with the players because the reality we created was unreal. We were, of course, mortified.”
Things changed with the arrival of product marketing lead, Don Traeger, and new project lead, Russell Dawe. Traeger made the team focus on the few positive comments from the focus group, and Dawe energised the team from a leadership perspective, says Salwitz.
“He immediately injected positiveness, organisation, brainstorming, and was basically the jolt of energy our team needed.”
The cast of characters was changed to become heightened versions of people you might see in real life, although Death and the Wolfman were among the more abstract characters to escape the cull.
The street play field was also reworked as a result of the focus group.
“Up to this point we had built the street house by house, but this was not going to scale,” says Salwitz.
“Dave’s answer was to construct two five-foot long foam storyboards, and to transfer the line drawings of our inventory of houses to those storyboards. He then put a plastic sheet on top of that so we could draw the characters, obstacles, the paper bundles, the paths of the objects, and triggers – anything that you would interact with on the street.”
The development team then spent a sunny Saturday afternoon outside in the atrium of Atari’s office, planning and creating the game.
“Both Dave and I remember this as being one of the great creative days in our careers,” Salwitz says.
The final game features 60 distinct street addresses made up from 11 base house designs. This is achieved by changing and reusing various assets such as doors and windows. For example, the first house you see on Easy Street is in the game six times with subtle changes.
“As with all games in those days, we had a problem. ROMs were expensive, and we were running out of space,“ says Salwitz. “The good news is that Paperboy is largely a pattern-learning game, and giving players familiarity by seeing the same house designs was really not a bad thing.”
The game was also innovative – particularly for an arcade title – in the way that it gave players a simple moral choice.
“A byproduct of these game systems gave the player a choice between good and, well, evil. We never did really think of the boy as a hooligan, but you did have a choice,” explains Salwitz.
“You could spend your game trashing every house on the street, and just save one, or you could focus on being the best paper-deliverer possible.”
Crunch is a well-known problem with modern game development, but in the mid-1980’s it was just as much part of the process. Salwitz says that he worked seven days a week for eight weeks straight to complete the game.
“In the end, us putting this game together was a labour of love,“ he says.
In 1985, a field test provided enough reassurance of Paperboy’s earning resilience to convince Atari’s leadership to mass produce the game. Around 3500 units were produced at a cost of $2500 each.
Seeing Paperboy in production gave Salwitz an enormous sense of satisfaction, despite a few bugs that were discovered just after the first batch of units were distributed.
“I cannot tell you how ridiculously wonderful and singular experience it is as a game maker to see your game on a manufacturing line,” Salwitz says.
Salwitz also touched on the Nintendo Entertainment System version of Paperboy.
“In every reasonable, independent, quantifiable, measure, the Atari System II crushes the NES,” he says. “It has more bits, it has more speed, it has more pixels, it has more RAM, it has more ROM. Hell, we had a custom controller, a beautiful medium resolution monitor, speakers and everything. Every single metric except one.”
And that metric was sales. When Tengan and Mindscape ported Paperboy to the NES in 1988 it sold over two million copies. Its success was a concrete sign of how the nature of video games was rapidly changing.
“Obviously, we were all very glad of this. We were very excited, frankly, and quite blown away. It took me personally, years and years after this, to ever realise that the days of the arcades were numbered.”
Since then, the game has been ported to multiple systems, including, much to Salwitz’s amusement, a handheld version from Grandstand.
“Unless your game has been ported as an LCD video game, you have not really made it!”