Will the penalty taker shoot for the right, left or center of the football goal? Professional goalkeepers know that they must take into consideration as much information as possible to prevent the score: the history of the rival player, the position of the taker before contact, and so on. By doing so, goalkeepers are in fact generating a model to improve their prediction about how to stop the hit, though they might not even realize the cognitive process at play. This general idea —that one could call the Goalkeeper’s Dilemma— is potentially the basis of a renewed contribution to the understanding of brain functioning, at least according to members of an ongoing research project at FAPESP’s Research, Innovation and Dissemination Center for Neuromathematics (RIDC NeuroMat).
NeuroMat members are developing “The Goalkeeper Game,” at the same time a computer game and an ambitious transdisciplinary research project. In the three-level computer game, an individual takes up the role of a goalkeeper that faces a penalty taker who is ready to shoot. As the game evolves, the expectation is that the individual will be able to make sense of the strategy of the penalty taker and have a high rate of success in thwarting scores. A beta version of the game is available here. The scientific questions that are associated to this project revolve around the learning process and the decision-making model of the goalkeeper.
The development of the “The Goalkeeper Game” started in July, 2015, and four NeuroMat professors have taken part in it. NeuroMat’s coordinator Antonio Galves, a professor at the University of São Paulo’s Mathematics and Statistics Institute, has led the team, that also includes Claudia Domingues Vargas, chief of the Neuroscience and Rehabilitation Research Center of the Institute of Neurology Deolindo Couto at the Federal University of Rio de Janeiro, Maria Elisa Pimentel Piemonte, from the University of São Paulo’s Department of Physical Therapy, Speech Therapy and Occupational Therapy at the School of Medicine, and André Frazão Helene, from the University of São Paulo’s Institute of Biosciences. Since the beginning of the project, meetings were held once or twice a month, that also included a game designer, an illustrator and members of NeuroMat's technology-transfer and scientific dissemination teams.
“The Goalkeeper Game” remains the joint-result of different areas of activities within NeuroMat. It connects directly to the RIDC’s scientific project, to the extent that it is informed by the conjecture that the brain is able to retrieve statistical regularities from stimuli. Specifically, NeuroMat researchers are interested in showing that, given the time spent to make a decision about an action in the game, it might be possible to recover the structure of the context tree that was generated at a given level of the game. Context trees are an extension of Markov chain models, in which a random variable may be predicted in a given sequence based on the minimum necessary previous actions, contrarily to a classical Markov chain in which redundant past might be incorporated in the prediction to the extent that it might have an order that is equal to the length of the tree. This is the core of the current research by NeuroMat member Bruno Monte de Castro, who relies on outputs from the game to tackle this issue. Challenges of the ongoing research agenda related to the game include the understanding of parameters to include in the model (i.e., level of information), the probability distribution that can be derived from game outputs and hypothesis tests to retrieve the tree from the time spent to make decisions.
The technology-transfer activities pertaining to “The Goalkeeper Game” are at least twofold. Firstly, there are specific technology questions on the game design within a scientific framework. There is a programming goal to generate an option in which players may devise their own context tree that would generate the sequence of penalty kicks. Outputs that will be generated should be held in a data base that is still in the making and connects to NeuroMat’s broader agenda of technology development. Secondly, an experimental protocol directed to rehabilitation of patients with Parkinson is being thought of, that might shed new light on learning processes of this type of patients.
According to Prof. Pimentel Piemonte, the game is been designed to contribute to the understanding of how learning processes occur. “Each level of the game presents a new challenge and requests a new strategy to be completed. What is interesting is to look at how users will come up with forms of learning that will make them succeed each level.” The game might also be associated to research on how background impacts learning processes, specifically related to identifying deterministic and probabilistic stimuli.
“The Goalkeeper Game” is expected to be included in the massive online course that the NeuroMat scientific dissemination team along with research members are setting up around the general topic The Statistician Brain. This course will generate an educational and a communicational outcome, specifically an open-access sequence of lectures on learning processes and modeling and a series of activities to spread out the general topic that informs the course. The game has had a pilot during a scientific dissemination activity called Virada Científica, at the University of São Paulo, in October, 2015. Then, the general public was able to test the game, and the RIDC development team was able to improve the general features of the software with reports from the users. Before this pilot, “The Goalkeeper Game” was tested by 14 graduate students enrolled to Prof. Piemonte’s class on “Sensory-motor learning: important foundations for rehabilitation,” at the Psychology Institute. At this closed test, students’ feedbacks were positive, especially pinpointing that they had been motivated to play the game. Students were asked to complete a three-level version of the game, and all of them were able to identify the specific pattern that was designed for each game level. Students also gave feedbacks on topics pertaining to the game tutorial and the general visual identity of the game.
The game is still at the developing stage, and updates will be posted on NeuroMat’s social-network pages.
This piece is part of NeuroMat's Newsletter #25. Read more hereShare on Twitter Share on Facebook