Game Development Reference
In-Depth Information
(e.g., A S D F J K L ; SPACE), whereas a gamer lays his her hands on
the keys most notably associated with gaming (e.g., W A S D SPACE
SHIFT). Don't believe me? Look at your keyboard. My WASD let-
ters are literally worn off. What this tells the program is if players use
WASD to move without prompting, suppress the moving tutorial. If
players attack an enemy without first taking a hit, suppress the attack-
ing tutorial. This is extremely important in managing cognitive load.
Learners who are already skilled face the expertise-reversal effect
of having instruction actually increase their cognitive load instead of
decreasing it. To avoid this, we need to detect player success as well
as failure, and rescind our teaching accordingly. This lets our skilled
players be skilled immediately.
Reward Failures
his seems at odds with other things I have told you in this topic. I
want to clarify what I mean by a certain type of failure as opposed to
unwanted behavior. The failure we want to reward is the failure that
occurs when we attempt to do something for the first time, or with
very little training, and do it poorly. This should be rewarded because
we tried, and to some extent hopefully learned something. On the
other hand, an unwanted behavior is something like being unable to
dodge bullets, return fire, use the cover mechanic, manage the inven-
tory, and so on, after having previously done it successfully. In order
to execute any punishment, we have to be sure that our learners have
learned how to perform the task. For example, it is useless to punish
an infant for babbling. They are trying to talk, and if you want to
encourage talking, you have to give them what I like to call correc-
tive rewards. This takes something that needs to be corrected, like
babbling, and removes it while still encouraging the core of what the
baby is trying to do—talk. A corrective reward for this kind of thing
might be the baby looking at a bottle and saying babbsbdbsdbpppgth! ,
to which I will reply, “Yes, BOTT - LE. B O TT L E , after which I
will give the baby the bottle. I have made a clear and present correc-
tion by enforcing the word bottle repeatedly and distancing the request
from the reward, but haven't relinquished the reward entirely, merely
delayed it. In Skull Island , we did this by making the AI prioritize AI
targets until the player makes his or her first kill. his way, the player
Search WWH ::




Custom Search