A PORTAL Postmortem: Integrating Writing and Design
Posted March 3 1:59 AM by Kelly Murdock
A PORTAL Postmortem: Integrating Writing and Design For the last hour of the conference, I decided to attend the Portal Postmortem, but I wasn’t alone. Of all the sessions for the entire GDC 2008, this was the most heavily attended session that I saw. The lines to get into this session extended all the way around the hall and into the next hall. I’m not sure if they turned anyone away from the session, but I ended up sitting on the last row of the hall and they’re were many people standing along the walls of the session and every single chair of the 6000+ hall were full.
Portal, of course, was named game of the year in the award’s ceremony just a couple of nights before and has become the belle of the ball for GDC. This session was presented by Kim Swift and Erik Wolpaw from Value and focused on how the writing and design elements were used to create the phenomenon of Portal.
Kim began the session by thanking all the attendees for playing Portal and explained that the session would contain spoilers for those who haven’t played the game, but I saw no one leave despite the warning.
Kim went on to explain that many of their design decisions were forced do to the constraints that they were under. For example, one of the reasons for making GLaDOS into a simple voice was that they didn’t have the resources to do any other choices.
When combining story and design, Erik described that the game has a story and the gameplay also tells a story and that the difference between these two stories needs to be consistent in order for a game to succeed. This is what the Valve team called a Delta value. Games with a high delta value have a tough time engaging the player. As an example of a game with a high delta is Clive Barker’s Undying. The game featured a story of zombies, but the gameplay had people standing around that weren’t running from the zombies. This disconnect only confuses and frustrates the player.
Kim then explained the team’s process. The first step is to playtest the level. By watching people play the game you can see how they react. It also helps to keep you objective. It also lets you know if the story is working. The team actually started playtesting the game the first week they were at Valve.
Erik commented that when writing funny dialog, it is hard to stay around all the tester where he could hear his dialogue over and over without any laughing. This is like torture for a writer, but by playtesting, he could hear reactions to his dialogue, which helped to keep his sanity.
When previewing the playtesters, it is important to not get too attached to any specific part of the story. As any example of this, Kim and Erik described how they designed the Marathon Box level. They started with a box that would get stuck in the goo if dropped, but this didn’t work too well. The second idea was to have the player always face the box and some dialogue was included to make the player take care of the box. They even put a heart on the box as a clue, which seemed to work out much better.
As another example, Kim and Erik described how the final boss battle with GLaDOS was designed. Through this process, they learned that more complex puzzles typically suffer from poor pacing. This was realized through playtesting where they were able to look at how the Firepit level succeeded. They were then able to use the lessons they learned from the successful levels to design the final battle level.
The final statement for the session was presented as a morale. The first point was to “embrace constraints as fuel for creativity,” and to “have faith in your writing, in your team and in playtesting.”
Southwest Gaming Expo
