Siren conflict resolution game wins Best Serious Game award
Consider this scene: Jamie, a boy in Holly’s class, keeps calling Lucy “fat”. Lucy tries to argue back, and then they end up getting into an argument (normally in the classroom). Eventually Mr Devenish (their class teacher) will tell them off, but they still keep on arguing, but more sneakily – sitting down and giving each other hostile glances. Holly thinks that afterwards they feel angry with each other, and a bit resentful that they didn’t manage to get their point across and conclusively finish the argument. This has been going on since Holly joined the school (almost a year ago). Now replace the names with Yannis, João, Marie or Olaf, and the reference to Lucy’s appearance with cultural origin, social or religious background or athletic ability and you end up with a whole range of situations that are very likely to happen in any school across Europe. In today’s multi-cultural and socially diverse Europe, confronting conflicts and coping with them is part of social life, especially since conflicts seem to arise in almost every context and developmental stage of human life, from scuffles in schoolyards, to bullying in the workplace and even to international warfare. The Siren project (http://www.sirenproject.eu) aims to tackle such problems, by creating a serious game which supports the role of teachers and helps them educate young people on how to understand and resolve conflicts. World-class research groups from Greece, Denmark, Portugal, UK and the USA and an award-winning game design company from Denmark worked together to design and develop an interactive environment which benefits from recent advances in serious games, social networks, computational intelligence and affective computing to create uniquely motivating and educating games that can help shape how children think about and handle conflict. The game developed by the project is able to automatically generate adaptive conflict scenarios that fit the teaching needs of particular groups of children with varying cultural background, maturity, technical expertise and the desired learning outcomes as specified by teachers, enabling the system to be used by school teachers all over Europe, without specific technical training. The main objective of the Siren game is to teach players peaceful and constructive ways for resolving conflicts, knowledge that can then be transferred to other domains. The players face a conflict situation together, with the conflict domain being relevant to the interests, maturity, and level of general knowledge of the participants. In order to increase the players’ level of familiarity with the aesthetics and mechanics of the game and, thus, maximise their level of interest, researchers chose two gaming genres popular among the selected age group (10-14-year old students) and created Village Voices, a collaborative version of a farm game designed by ITU, resembling social games played in social networking websites, and My Dream Theatre, a role-playing game designed by INESC-ID, where the player becomes the director of a theatre club, assigning roles to actors played by non-player characters (NPCs) and attempting to work out the conflicts arising between unappreciative characters or as a result of events external to the theatre company. Students play these two games in successive stages, tackling an increasingly complex scenario each time; each scenario contains one or more goals, which players need to achieve, a number of obstacles, and means to overcome the obstacles. In terms of game mechanics, these scenarios are formalised as collaborative puzzle solving with constraints, where each participant has incomplete information about the overall state of the game. All of these elements support the learning objectives of the game by immersing players in the conflict, facilitating a critical approach to their assumptions about the conflict and allowing them to explore new perspectives other than their own. The different game scenarios that players need to solve correspond to dealing with situations that produce conflict among the particular age group. The reasons were identified after thorough user studies in the U.K. Portugal and Greece, coordinated by the University of Birmingham and INESC-ID; these studies included extensive questionnaires and interviews with students and teachers, hands-on exercises and cultural probes. Researchers sampled conflict situations, intensities and intervention approaches across schools from different cultural backgrounds in each country, so as to capture cultural differences not necessarily related to ethnicity. In the following, findings from user studies were transformed to scenarios in the context of the two game genres. Property and resource management, relations with other peers, cultural differences and reputation/rumours were among the most popular reasons for conflict. In order to make the game relevant, game mechanics and objectives were included in the game narrative, corresponding to those factors. In addition, conflict intensity is measured via in-game sampling and non-verbal estimation of the players’ affective and cognitive state. The idea here is that the game needs to be informed about the relevance of each particular scenario instance for each player, whether its conflict intensity is exceptionally high and may result in excessive stress for the player or whether the player is actually engaged in the game. This is achieved with robust results by combining the player’s game behaviour and choices, with estimates of facial expression and attention, produced by lightweight software components developed by NTUA using images captured with a plain web camera mounted on top of the player’s computer. Combining information from the players’ facial expressions and head movements with how a game level has been created and how well players do in the game can help game designers produce quests and challenges fitted perfectly for each particular player, creating games with potentially infinite gameplay, where players are never bored or frustrated, the required skill level is matched with the player’s skill and current affective and cognitive state and new challenges are added when the perceived skill level and engagement increase. NTUA and ITU have tested this approach on platform games and first-person shooters and are now taking it to other games genres, such as social and role-play games. Visual and gameplay features are taken into account in order to procedurally produce the quantitative properties for the next mini game, aiming to maximise the player’s satisfaction and engagement and, along with that, the potential to achieve the game’s learning objectives. This advanced game technology recognizes the emotional reactions of its players and the conflict level of the game at any time; driven by culture-aware player models, ITU’s game AI component generates automatically the next quest for each player so that conflict stays within appropriate levels. The player-driven procedural quest generation in Siren offers personalised learning experiences and player-specific exposure to conflict resolution which results in effective training of social skills. Content generation is accomplished in real time, with the aid of genetic algorithms trained with data collected during user experience tests in Portugal, Greece and the UK. Collected data were used to map the properties of each mini game to conflict intensities, as perceived by the student, and then converted to user and adaptation models, catering for cultural adaptation and personalisation. Besides cultural adaptation, the game’s interface has been localised to all four participating countries by SGI, catering for easier adoption from schools. Evaluation carried out by the University of Bath and INESC-ID in the UK and Portugal showed that the game was interesting and motivating for the students, as measured by the Intrinsic Motivation Index which indicates that participants enjoyed the game and felt competent in it. In addition, the analysis of children’s conflicts when they played Village Voices revealed that children over time were more likely to use more positive conflict resolution strategies: during the first sessions, the children begun stealing of each other and they found that this unregulated stealing led to no one progressing very far and also was very upsetting, with one of the children crying at the end of it; in the next session, they tried to regulate the stealing amongst themselves and found that this led to them all progressing further. Overall, three out of five of the groups of students developed their own rules for dealing with conflicts; this is a major contribution of Siren as a safe ‘sandbox’ in which children can experiment with different approaches to conflict. Siren was voted European Learning Game of the Year 2013, by the Games and Learning Alliance Network of Excellence, reached the final round of the games competition at the European Conference on Games-Based Learning 2013, and was ranked fourth in the list of ‘Great Examples’ of ‘Games for Impact’ of the Office of Science and Technology Policy of the White House.
Keywords
Serious games, learning
Countries
Denmark, Greece, Malta, United Kingdom