Game Development Reference
In-Depth Information
on the domain and quality of tags that are validated. Therefore, we could not really
employ any gradual difficulty policy in game, except the fact that levels (sequences
questions) for higher ranking players were composed from more tasks and therefore
the player spent more time with them and put larger effort to acquire bonuses (which
were awarded for no-mistake answering).
From all of our games, most effort was made to encapsulate the CityLights'
purpose. Since a musical riddle is a common type of game and players did not have
to type in any words, just use the cursor, most players have not realized that the game
has some higher purpose (unless they read the game description). A solid effort was
also made by devising an appealing graphical design of the game and also a simple
metaphor of player wandering through the night city. These features too helped to
encapsulate the purpose a bit.
The CityLights design also implicitly ruled out cheating threats. As a
prevention
,
the game rules were set in a way that random player behavior would cause loosing the
points in the game, instead of winning some—there was no other way of winning the
game and gaining rank, than correctly answering questions. A hacking eventuality
was also ruled out by using a thin client and moving all of the game's logic to the
server side of the game.
8.3.1 Validation Games
The game is designed tominimize the cold-start problemvia single-player. The single
player nature of the gamewas enabled by the existence of the tag assignments featured
in the game. The score in the game is inferred from the existing facts, however, these
facts (assigned tags) only approximate the true solutions (correctly assigned tags) and
the scoring is truly
independent
on the quality of artifacts. Therefore, we categorize
the CityLights artifact evaluation mechanism as
approximative-algorithmic
scoring
with a subsequent offline cross-player validation.
Still, this assignment does not fit so much, it is not an “approximative algorithmic
evaluation”, it only approximately scores the player while for the computation an
existing metadata set is used, unusual to other SAGs.
We believe that CityLights may mark a new breed of semantics acquisition games,
oriented towards validation of existing artifact assignments, e.g. descriptive metadata
or term relationships. On abstract level, each of such games would operate with the
same principle—a question—which of course could be incorporated into the game
by various means and forms.
The game constructs a question to the player with several possible answers. Both
questions and answers could be any entities, relationships of which are to be
validated. For example, in case of validating of descriptive metadata, the question
can be represented by resource sample and answers by metadata, but it can be also
other way around, e.g. the question may be some metadata sample and the answers
some resources that can possibly match the question.
Search WWH ::
Custom Search