How to make educational games more effective

Changing the world with games is hard.

Speaking at the 2014 Games for Change festival, Zoran Popović — University of Washington associate professor of computer science and founder of its Center for Game Science — outlined why drastic transformative change with games is difficult to achieve, and what we can do to come closer to doing so. Popović discussed two "key" world domains that often look to games, scientific discovery and scholastic mastery for a wider range of students, detailing why games haven't revolutionized these areas yet.

Popović said transformative change requires a large amount of effort over an extended period of time, in addition to real-world synergy behind it and a design focused on outcome to begin with — including outcomes that are unknown. A game by itself cannot achieve transformative change on its own — people need to commit to it and put their real-world energy behind it, whether that means testing the game in lab experiments or real-life applicable.

There is hope, said Popović — although we're not too close to any of the outcomes we want to see from games right now. According to Popović, there is an emergence of a system that will lead us into the process of changing these problems with games, though it's not necessarily on the immediate horizon.

In order to make a transformative game for education, Popović said people need to understand the system it will work in. This system includes students, teachers, curriculum and classroom tools and how these tools extend from the school environment to the home environment. Popović and his team are developing the Engaged Learning Platform, which uses an "infinite curriculum" — this means, he said, that nothing in the game is finalized and there is always material suited to a student's current level of learning. The platform is designed to find the optimal pathway from each teacher to student, a pathway that differs from teacher to teacher and student to student.

Thinking of educational games in this way broadens the design space, Popović said. Assessing and building the entire system and tailoring it to adapt to different learners can work for both math and text-based problems.

Data in an educational game also has to adapt towards the best possible outcome. Creators should identify what the smallest dimensions that promote learning are, and then build them outwards. Problems brought up and taught in the game need to be able to translate into real-word problems with solutions application outside of the game. Just giving students random concepts to deal with, without clearly demonstrating how important certain variable are, will result in students' failure to retain information.

Games must also positively reinforce in students that they are working towards a worthwhile goal. Telling them that they "worked hard" rather than they "did well" will improve their performance over time, Popović said, and will help them to develop what he called a growth mindset. A growth mindset is when children's persistence and drive in wanting to solve a problem increases, with the children themselves being the driving force behind their own desire to learn. In his team's studies, this mindset, Popović said, showed that children who struggled with certain concepts were more likely to keep trying to solve problems under the growth mindset system.

Popović and his team are working on another game, NanoCrafter, that maps biological proteins. The game is designed for collective creativity, a crowdsourcing game of sorts through which he team hopes to crack the code of building certain proteins.

"If you want to design something for maximal learning, you need a finer scaling with which to asses the curriculum," he explained. "We develop a thought process language — how to solve specific puzzles or equations or prove a theory — and generate challenges that develop specific though processes or present automatic explanations and track players thinking."