State-of-the-Art: Design of the Semantics Acquisition Games - Semantic Acquisition Games: Harnessing Manpower for Creating Semantics - page 105

Task distribution

Player challenges

Player capability driven

Data (ontology) driven

Social experience

Task-value driven

Task difficulty

Self-challenge

Greedy

Equally complex tasks

Competition

Random

Discovery Gradually complex tasks

Purpose
encapsulation

Low

Offline player mutual agreement

High

Online player mutual agreement

Bootstrapping

Restrictive rules

Automatic approximative

Automated exact

Mutual player supervision

Helper artifacts

Anomalous behavior detection

Validation of
player output

Anti-cheating
measures

A posteriori cheating detection

Fig. 7.1 Six design dimensions of semantics acquisition games. Each axis represents one design
dimension. The values on the axes represent different types of solutions (design patterns) that SAGs
feature to solve issues that dimensions represent. SAGs sometimes use more than one types of
solutions

as standards by SAGs. Almost each SAG (when differences in purposes are counted)
represents an unique combination of the design patterns and features. Overall the
“design space” represents a system of trade-offs that balance the amount technical,
artistic and marketing work of designers at one side and cold start problems, vul-
nerability to malicious player behavior, attraction, retention and game output quality
and quantity on the other side.
Some of the design dimensions represent critical problems which must be over-
came by each SAG (using one of the patterns), other merely adjust the final output
of the game. As critical, we could consider validation of player output (which must
be done in order to properly score the player), the anti-cheating measures (which
must deal with all threats the game rules impose) and player challenges (from which
at least one must be present in the game). Satisfaction of these is a requirement for
a successful SAG. On the other hand, using a random task distribution or low purpose
encapsulation , won't necessarily disable the whole game concept, but may seriously
decrease the quality and quantity of game's output. At the same time, the existing
SAGs can always be polished along these dimensions (e.g. SAG randomly assigning