Fans of football computer games could soon be playing
against more realistic opponents thanks to thewinning entry in an artificial intelligence
competition.
The De Montfort University Leicester team won the
2008
Machine Intelligence Competition, run by the British Computer
Society (BCS), which was held at Cambridge University this
week.
The competition took place during the British Computer Society's
28th SGAI International
Conference on Artificial Intelligence (AI-2008).
The annual conference brings together experts in the field to
discuss future developments in artificial intelligence.
Martin Rhodes and Simon Coupland's winning software uses
programming inspired by Darwinian evolution to enhance the way
virtual players in football games take free kicks.
The computer-controlled opponents in the system take free kicks
in a more varied and organic way than those in existing games.
This means their actions are less predictable and makes them a
more challenging opponent for the person playing against the
computer.
Rhodes, who is a PhD student at De Montfort University,
developed the system as part of his MSc in intelligent systems and
robotics, which he completed last year.
He said, "I am very pleased that my
MSc dissertation work has received this level of recognition.
Taking the software from drawings on a piece of paper through to a
final, award-winning product has been a brilliant experience."
Rhodes presented the system alongside Coupland who is his PhD
supervisor and an expert in computational intelligence working in
De Montfort's Centre for Computational Intelligence.
They were chosen as the winning team by the conference
delegates, who voted in a secret ballot following live
demonstrations by each of the finalists. Their prize was
£1,000.
Coupland said, "In video games, scenarios tend to be scripted,
rule-based, deterministic systems. This means if you get the same
situation twice, the exact same things will happen in the exact
same sequence.
"Any hardcore video gamer will tell you this is a significant
limitation of current games. This new approach gives a diversity of
good computer game play and means the computer agent players are
unpredictable."
The system uses a computer model of Darwinian evolution which
chooses the best options from the numerous possibilities for a shot
on goal and can also decide where to place defenders and the keeper
for a free kick situation.
The team believes the system has a lot of potential for
improving future football computer games releases and could easily
be integrated into gaming technology to help improve the experience
for gamers in the future.