At the recent Games Convention in Leipzig, Bruce Shelley shared his perspective in a presentation called "Designing by Playing" given during the GCDC developer event.
The Ensemble Studios senior designer helped start the studio in 1995, working on both Age of Empires
and Age of Empires II: The Age of Kings
-- and "Designing by Playing" is his name for the development process that created the AoE
It starts with the creation of a playable game prototype -- quick and crude -- early in the process, and then evolving the design. Shelley likens the process to "sculpturing out of a block of concrete."
Next comes the playtesting, a time to consider what is working, not working, or missing from the game. Why use it? Shelley says, "You must play to judge a game. I think you need that intuition ... to know if you are going into the right direction."
"Designers are only guessing until they play," he adds, noting that there are too many variables in a game today to get it right the first time.
Shelley highlighted some critical conditions for success. "People who are testing and playing the game .. have to have not only skill, but interest and passion in games," he said, recommending players with good gaming instincts and experience with games. Additionally, a test base comprised of a variety of gaming tastes leads to games with wide appeal, he added.
Age of Empires
was opened up to the entire Microsoft community for playtesting. But at one point, Shelley said the company even forbade employees to play AoE
at some point, lest they lose their objectivity to ensure success.
Shelley stressed quality testing -- play to test, not to win -- with rapid turn-around, meaning an entirely different game when the testers play again the next day. Moreover, testing should be frequent, even daily.
So what can go wrong? First, the designer noted, being decisive is important. "You get this feedback from this playtest; if you're indecisive on how to use this feedback, you'll lose the vision for the game." He recommends avoiding any breakdown in the system of testing, collecting feedback and rapid implementation of feedback.
Shelley also cautioned designers to be cautious that the "offense" approach -- aggressively and decisively implementing changes from feedback -- doesn't switch to the "goal-tending" approach of designers defending their vision.
Shelley's five-step process? All letter "P": Plan, Prototype (quickly!) Playtest (a year or more) Polish (balance, art) and then Publish. "A multiplayer prototype existing at the six month mark is the goal," he added.
He also stressed multiplayer testing, whether volunteer or even "semi-compulsory" (he recalled Ensemble's "Angry Tuesdays"). "We also find other stuff we are not necessarily testing for, things like graphic glitches and bugs," he said.
Specifying the conditions for effective testing, Shelley again highlighted the importance of having gamers on board who have "reasonable instincts" for what is entertaining and fun -- or not. He reiterated a wide range of tastes, and said the playtesters should be those who follow directions well, as the desired feedback is usually needed for very specific areas.
Finally, the quality of the feedback is important -- look for those who provide constructive criticism, including reasons and alternatives. Shelley said, "People who come with problems but also solutions are people you notice, people you promote."
During the design iteration process, the project leads determine what changes are to be made; they implement the changes and create new versions. Immediately, the new version is installed in the test lab for another round of testing.
"Design by playing has its root in board games," Shelley explained. "We sort of had cardboards and pieces and pushed the pieces around." Ensemble Studios, he noted, started with little design experience but lots of playing experience, adding, "The design by playing process turns the playing experience into a development asset."
The strengths of the approach include that it reviews the ever-changing experience of playing. It means "lots of different eyes and minds engaged with a game, which leads to games that are fun for your players," Shelley said. If you first start with casual players and make it fun -- and then, add the hardcore players and make it fun for them, too, a game develops a broader appeal. It's good for developers, too; Shelley says the experience helps them take ownership of their game. "If you get people to playing, they become better at what they are doing," he said.
But design by playing also has some weaknesses. It can be difficult to predict when the design process will be finished, which can have implications for cost. For AoE
, the designer noted, "We were at a good spot at that time, because Microsoft didn't have much experience in game publishing ...they were learning to become a game publisher, while we became a good game developer."
Design by playing also results in a game that suits the taste of those who played it -- whether the audience is narrow or broad becomes a big factor, and players will influence each other's taste. There will always, of course, be someone who dislikes something; "You can't iterate forever," Shelley pointed out -- but even a "lite" model of design by playing works, he adds, if feedback is used constructively.
To conclude, Shelley listed some examples of designers other than Ensemble's studio who use the design by playing method -- his old Microprose colleague Sid Meier and World Of Warcraft
creator Blizzard, to name a few. "Blizzard makes us look like amateurs, the way they collect feedback and design by playing," he said. "They are also very secretive about it."