Game Development Reference
In-Depth Information
be a problem. But what happens when playtesters refuse to flee from the
machete man, or they laugh at the cheesy message in the mirror? The
subordinates can't solve these problems on their own because they don't
really understand what they're doing, and they have no authority to make
changes. So they have to go back to the leader. The leader will probably give
them a bad solution to the problem because he lacks intimate knowledge
of the level and he wasn't actually at the failed playtest. The subordinates'
specific knowledge is discarded, and the process flails.
Instead, the leader should be communicating intent: “This level should
teach the player all the basic controls except weapons, which should be left
until the next level. It should introduce the dead girlfriend backstory so
that we can play on it later. It should also introduce Machete Man. Since
the next level starts outside the abandoned house, this one should end
coming out the front door of the house. It's the first level, so you can start
the player wherever you like. Finally, don't characterize the protagonist,
because I want him to remain a mystery at this stage.” Now the people in
the trenches can go to work. They might start with the same design as the
leader originally suggested. But during iteration, they'll likely find a better
way to achieve the same goals. Playtests and brainstorming will reveal
layers and layers of new problems and solutions that the leader could never
have discovered alone. Being freed of a bullet-point list of things to do, the
developers can seize opportunities and solve problems as they arise. The
design will change, but since they understand its role in the larger struc-
ture, they can make sure it fits in well and properly sets up the following
level. And the leader is freed of the burden of details.
To explain his intent, a leader must first know what it is. This means
understanding the structure of the game and the purpose of every de-
veloper's work. Since this requires some difficult mental gymnastics,
bad leaders often revert into micromanaging nitpicking mode. They give
vague intent and then criticize the details of whatever their subordinates
come back with. This kind of “judgment from on high” adds little to a
game. The leader is the only one with the power to perfect the game's
broad structure. He should be thinking about, analyzing, and iterating on
that every day. If he's doing that job well, he won't have time to nitpick the
lower-level work of others. If a leader must judge others' work to maintain
quality standards, he should limit his comments to generalized state-
ments of deficiencies and offers of help, not specific marching orders on
what to do.
