Planning and Iteration - Designing Games - page 281

But these aren't the greatest costs of overplanning. The real price of
overplanning is in how it creates a false sense of certainty about the future.
Written plans are often treated as guaranteed visions of what's going to
happen. But they're not—they're laden with assumptions. Work that de-
pends on those assumptions will later collapse when they're shown to be
false.
For example, imagine an initial design document says that the player
character can jump 10 feet into the air. Based on that plan, a level artist
builds a level bounded by 11-foot-high walls. If the plan is correct, every-
thing should work, since the player character can't get over an 11-foot wall
with a 10-foot jump. But then the designer discovers that the game feels
much better if the character jumps 15 feet in the air instead of 10. Now
there's a problem. Either the level needs to be reworked to handle 15-foot
jumps, or the jump must stay at 10 feet where it would feel better at 15. One
choice throws out good work. The other weakens the game.
And it's never as simple as this. Real game designs are networks of
dependencies; changes in one place almost always imply many changes
elsewhere. A simple change in jump height might affect level boundar-
ies, enemy movement (so they can catch a high-jumping player), jumping
puzzles, audiovisual effects applied to jumping, and more. And each of
these changes might imply further changes—changing enemy movement
might involve adjustments in the enemy art and animation. Reworking
a jumping puzzle might then mean changing the plot of a level, if that
puzzle was threaded together with the story. The effects of the failed plan
ripple through the design, twisting art, code, mechanics, and fiction.
Game design is unusual among modern creative pursuits in the amount
of uncertainty embedded in every plan.
As Soren Johnson, lead designer of Civilization IV , wrote, “To be a
game designer is to be wrong.” A designer can guess how a system or a
level will play out, but he can never know. Usually, when it is constructed,
the game system plays very differently than anticipated. That's why great
games tend to change so much during development.
For example, Halo spawned one of the most popular first-person
shooter franchises ever. But in its original form, Halo wasn't a shooter,
and it wasn't first-person. It was a top-down strategy game. Instead of
firing a gun through the eyes of a space marine, the player viewed the
battlefield from above and used a point-and-click interface to order troops