Game Development Reference
In-Depth Information
get a food pellet if it pushes the lever at least 15 seconds after the last time
it received a pellet. In classic deathmatch games, a health pack might not
reappear until exactly 60 seconds after it is taken, at which point it can be
grabbed again.
Players respond to fixed interval schedules with high motivation when
the reward is available and low motivation otherwise. For example, they'll
ignore the health pack until it has almost reappeared, then start checking
more often until it appears, after which they grab it and ignore it for a
while longer. Those gaps in motivation make fixed interval schedules poor
motivators.
A
variable interval
reinforcement schedule makes the reward available
a randomized amount of time after it is taken. This schedule creates con-
tinuous motivation, but at a lower intensity than the variable ratio sched-
ule. In both cases the player could theoretically gain something with every
attempt at collecting the reward. But in the variable interval schedule the
chances are lower the more often the player checks, so he has less reason
to quickly repeat the action.
The
differential
reinforcement
of
low
response
rate
schedule is like the
fixed interval schedule, except that if the player tries to collect the reward
too early, the interval is restarted. This reinforces low, steady rates of
activity.
With the
differential
reinforcement
of
high
response
rate
schedule, the
player must do a certain amount of activity within the interval to get the
reward. For example, you must defeat five enemies within one minute to
get the reward.
And there are countless others reinforcement schedules, each with its
own characteristic pattern of motivation.
suPeRimPoseD ReinfoRCement sCHeDules
Most reward schedules create dips in motivation at certain times. These
points are a problem because they create a window where the player may
decide to stop playing. To eliminate these motivation gaps, a game can run
several reinforcement schedules at the same time.
The power of reinforcement schedules isn't in any one schedule—it
is in superimposing them so that there is always at least one that is
producing high motivation.
