E.T.: Why it was a victim of its own schedule

By Howard Scott Warshaw. Posted

The tuning phase is where most of a game’s quality is generated. One case which truly illustrates the importance of the tuning phase is my 1982 video game, E.T. I had all of 36 hours to do the design. Of course, that wasn’t just for design: it had to include eating and sleeping and a Learjet ride to the presentation, 400 miles away. Nevertheless, I produced a plan which gave me a path to the finish line. This enabled me to begin a determined march toward a clearly defined destination. Then I started working my butt off toward that goal.

A goal which sat ever so slightly beyond the fog plane for the next four and a half weeks. I had created a design spec which reflected my concept, and all I had to do was reach it. And I did. The fact is, E.T. achieved virtually 100 percent of the original design concept.

What a disaster! Ordinarily, this doesn’t sound like a problem, but I believe it’s actually one of the top reasons why the game has problems.

“Hey there, how’d that project go?”

“We achieved 100% of our original concept.”

“Oooh, sorry to hear that.”

Despite his limited time, Warshaw came up with an ambitious adventure game concept for E.T., which involved collecting items on an open map.

This conversation sounds ridiculous, but the fact is, most products deviate significantly from their original design. Why? 

Sometimes a design is overly ambitious and cannot be realised, but this is more frequently the case for failed products. Successful products tend to deviate from design because they get better. As I move from vision to reality, stuff happens. I learn things. I achieve new perspectives and insights. Some of these provide improvements and alternative approaches I couldn’t see at first. If I’m paying attention while I’m working, my understanding and capability are growing along the way. If I feed this back into the product, my target both moves and improves.

It doesn’t have to go that way, though. Sometimes I’m too busy to improve my methods. Or I choose to remain married to my original concept and ignore any new information or feedback which arises. This way, there’s no danger of veering off target (and no danger of doing better). Ego can be another block to progress. If I over-commit to certain game mechanics or elements, there may not be room for something better to find its way in.

What if I really believe in an idea and choose to defend it? There’s an old saying about editing your own work: “It’s tough to kill your children.” It’s also true that simultaneously holding commitment to a concept and remaining open to new directions is a tricky balance to strike.

For a variety of reasons, it’s best to view original designs as launching points, rather than true destinations. Most developers shoot for a final delivery which is enhanced significantly beyond the initial concept. Many of my games benefited from this trajectory, but E.T. did not. There simply wasn’t time. The E.T. video game was a victim of its own schedule.

Director Steven Spielberg gave Warshaw his blessing to make E.T., following his successful adaptation of Raiders of the Lost Ark.

First Playable

Because the time available was so short, this game was going to have to be released at First Playable. In a normal development, reaching a proposed and accepted design is the first milestone, and the next significant milestone is the point when all the basic game elements are represented, and all the rules are implemented. No real graphics, bells or whistles, just the bare essentials of the game. This is the first time the game can be played and experienced largely as the design intended. This is known as First Playable. It usually occurs somewhere around 30 to 50 percent of the planned schedule.

First Playable is really the starting point, because it’s the first time you get to feel the game, to begin seeing what works and what doesn’t. This launches the Tuning Phase, where you spend the majority of your time fixing, adjusting, and improving the game. Deviation from the initial design should be beneficial.

First Playable is a basis for change. E.T. delivered 100% of its original design concept for one simple reason: there was no time for a tuning phase. That’s the cost of a short schedule, or more accurately, that’s half the cost.

An over-production of E.T. cartridges left many buried in a New Mexico landfill – though not the millions of copies we were led to believe.

With E.T., simply getting to First Playable in five weeks would be a considerable achievement. It was clear there wouldn’t be time to tune the game. This put inordinate pressure on the design to be perfect (which rarely happens). This leads to the next problem with the game.
It’s not just tuning time that is lost. Tuning means refining game mechanics which already exist. But what if, even after tuning, some mechanics still suck? Now I need to let it go for a bit and clear my head. Then come back to the game with fresh eyes, and hopefully new ideas or approaches. This is called rumination time. This is the other major cost of a short schedule.

The lack of rumination time was a major issue with E.T. On a creative project, I need the opportunity to create it, get sick of it, step away from it, forget it, and then come back to it anew. This is where giant evolutionary steps come from. This isn’t adjusting, this is redirecting, and it can work miracles. I never had the chance to take any real break from E.T., and that was a problem for the game.

The moral of the story is: try not to deliver your concept. Deliver the sum of your concept plus tuning plus rumination. Deliver the brilliant symphony you arrive at by the end of the journey which began with your concept. Remember: an initial design is not a goal, it’s a launching point. Unless you’re doing E.T., then it becomes a landing zone.

E.T.'s tendency to fall into steep pits was and remains a common source of frustration for players.

Second Thoughts

There were other games I launched at Atari… and abandoned. I’d work on them for a while, get them set up, and start tuning. Then I’d take a break and do my rumination thing. I’d kick it around, talk with other engineers about it. Make faces at it. Call it names for a while. But sometimes nothing really comes of it. In moments like this, it’s important to remember two quotes: one, “Genius must be allowed to fail.” And two, “The captain who goes down with the ship is dead.” I recommend living to make another game.

More features from Wireframe magazine

Subscribe