designers to go through these meaningless bugs will waste far more time than it may
save. It makes the most sense to bring in the traditional testers only when the game is
in a state that is truly appropriate for testing. In the end, bringing them in too early will
only delay the game’s progress.
HowtoTest.................................
How you have your playtesters work on your game is as important as who you have
testing and when you have them do it. Game designers will often ruin the effectiveness
of their playtesters by making a number of fundamental errors in how they interact with
them. These are all problems that can be easily avoided, as long as the designer is con-
scious of the way he deals with his testers and what he does and does not tell them.
The most important part of interacting with playtesters is to actually spend most of
your time watching them play instead of telling them how to play. Let them play the
game their own way and see how they fare. The temptation to correct playtesters’
actions is great and can be hard to resist. By the time the traditional playtesters start on
the game, the designer has already played the game so much that he is intimately famil-
iar with what the players are “supposed” to do in a given situation and how the game is
“supposed” to be played in general. When watching over the shoulder of a playtester
for the first time, the temptation is to say, “Go over there next,” or “You want to use the
strafe buttons for that,” or “Why don’t you try to get the power-foozle?” Watching
someone stumble while playing a game the designer is intimately familiar with can
quickly turn him into a teacher.
But the point of the playtesting is to see how players will actually play the game
without the game’s designer coaching his every move. Certainly, the designer cannot
fit in the box the game comes in or even be downloaded over the Internet. A certain
amount of stumbling about and learning the controls is to be expected, and the best way
to playtest is to let the testers do this initial exploration on their own. And if the players
truly do get stuck or if they never seem to be able to master the controls, the designer
needs to ask himself what is causing these problems. Is the game too hard or too con-
fusing? How can it be made simpler so that players have a fair chance of understanding
it and learning how to play? These are the lessons a designer is supposed to take away
from playtesting, but they are lessons the designer is never going to learn if he corrects
the tester’s playing at every step. Beyond that, when I sit in on a playtesting session, if
the testers do not know any better, I try to pretend I am not on the development team to
avoid having them color their opinions to be nice to me. Indeed, most gameplay focus
tests are conducted using a one-way mirror, since people will behave differently when
they think they are alone than if they have someone standing over their shoulders tak-
ing notes, regardless of how quiet that person is or who he claims to be.
While watching the testers play, the designer should try to observe the way in
which they try to play the game. Players may not try the approach or solution the
designer had thought of to a particular situation. The designer must then ask, does the
game support what the tester is trying to do, and if not, could it and should it? The test-
ing period, if started early enough, is a time when the designer can add a breadth of
content to the game that will allow the game to truly be accepting of multiple playing
styles. Up until this point, the people playing the game have been limited to the
Chapter 25: Playtesting 491