Information Technology Reference
In-Depth Information
4.2
Offensive Positioning
The selection of positions for players in offensive situations (team owns the ball)
typically consists on finding the most suitable position to: i) dribble the ball, for the
ball owner player; ii) receive a pass; or iii) score a goal.
In 2008, Kyrylov and Razykov [83] applied the Pareto Optimality principle to the
selection of these positions based on the following set of partially conflicting criteria
[84] for simultaneous optimization: i) players must preserve formation and open spac-
es; ii) attackers must be open for a direct pass, keep an open path to the opponent's
goal and stay near the opponent's offside line to be able to penetrate the defense; and
iii) non-attackers should create chances to launch the attack.
In the same year, Nakanishi et al. [85] proposed a method for marked teammates to
find a good run-away position based on Partial (approximate) Dominant Regions.
This method divides the field into regions based on the players time of arrival (similar
to a Voronoi diagram based on the distance of arrival), each of which shows an area
that players can reach faster than others.
One year later, Gabel and Riedmiller [86] proposed the use of a Simultaneous Per-
turbation Stochastic Approximation (SPSA) combined with a RPROP learning tech-
nique (RSPSA) to overcome the opponent's offside trap by coordinated passing and
player movements. The receiver of the pass that breaks the opponent's defense starts
running in the correct direction at the right time, preferably being positioned right
before the offside line while running at its maximal speed when the pass is executed.
4.3
Dynamic Positioning and Role Exchange
The Dynamic Positioning and Role Exchange (DPRE), and Dynamic Covering [1, 2,
4, 19, 20] was based on previous work from Stone et al. [49,50,79] which suggested
the use of flexible agent roles with protocols for switching among them. The concept
was extended and players may exchange their positionings and player types in the
current formation if the utility of that exchange is positive for the team. Positioning
exchange utilities are calculated using the distances from the player's present positions
to their strategic positions and the importance of their positionings in the formation on
that situation.
4.4
Setplay Framework
Robotic cooperation demands coordination at team level, requiring planning at differ-
ent abstraction viewpoints and situations. Setplays are frequently used in many human
team sports, e.g. rugby, basket- ball, handball, soccer and baseball. Certainly, there
are considerable differences between robot soccer and standard sports, but Setplays
were, even so, always expected to have a considerable impact on team-level coordina-
tion and cooperation.
Tactics and skills of robots are always improving, and thus opponent teams try to
adapt to new playing patterns and react to them. It is thus convenient to be able to
define Setplays, through freely editable configuration files, or even a generic Setplay
graphical editor.
