Game Development Reference
In-Depth Information
originally aimed to use the network visualized in exploratory search tasks). For this,
we needed only limited number of players playing the game with a same task term.
We have to come up with a task assigning strategy that would not break the direct
competition between players.
The answer is to organize the gameplay within “ladder periods” lasting from
several days to several week (the exact duration depends on the number of players
participating). Each period has a set of task terms assigned to it—the players are able
to play only with them during the period. At the start of the period, the ladders are
reset, so the competition begin anew. This, as a side effect, aimed at attracting new
players to the game.
To prevent the players from playing with just one task term (which is possible,
since it is within player's power to repeat as many attempts as he wants), the game
keeps a separate
task ladder
for each term and then a
joint
(“overall”) ladder. The
arithmetic of the joint ladder simply accumulate the partial results found in task
ladders: for each task ladder the player has a rank in (i.e. he has finished at least one
game with that particular term), he receives points to the joint ladder, based on his
relative position within that task ladder. The set of terms for one period is relatively
small (up to ten terms), so it is within reason to play with them all for a regular
player and have a chance to reach the maximum score in the joint ladder (which is,
naturally, the most prestigious one).
8.1.2 Cheating Vulnerability and a Posteriori Cheating Detection
After the initial design of the Little Search Game and its first deployment, it was soon
apparent that the game is prone to two types of unfair or malicious player behavior:
1.
The use of stopwords
. The LSG scores players according to real web
co-occurrence of terms. Guessing semantically related words to a given task
word is a fair challenge and is encouraged by the game. Unfortunately, there is
a workaround: the player may easily use some very frequently used words instead
and achieve even “better results” (lower number-of-result yield), while providing
absolutely no value to the purpose of the game and possibly ruining the game's
competition. Such frequent words may be stopwords of the language (e.g. prepo-
sitions), corpus (e.g. “software” in a specific domain of software engineering) or
syntactic features of the corpus resources (e.g. HTML tags indexed by the search
engine, such as “table”, or words frequently used for structuring the websites,
such as “menu”).
2.
Interfering with game implementation
. This is a rather “standard” threat of
many regular games, particularly online, competitive games. An interfering with
game's network communication between the client and server is the usual type
of attack. A standard solution to this is executing all of the game's logic to the
server, leaving the client with mere visualization and interaction functions (i.e.
the set of actions available to the player is equivalent with set of actions available
Search WWH ::
Custom Search