UBC Sauder shines in campus-wide Varsity Challenge

Posted 2019-05-17

Three UBC Sauder teams, including two from the Master of Business Analytics program and one from the Bachelor of Commerce - powered through to the finals in the recent UBC Varsity Challenge co-hosted by UBC Sauder and Kabam Games.

Kabam – a world leader in developing entertaining, immersive, and highly social multiplayer games for mobile devices – tasked participants with conducting data analysis and creating an algorithmic model to identify the value of mobile game users.

Zhen Mu, a finalist in the challenge and a Master of Business Analytics (MBAN) candidate at UBC Sauder found the experience intense yet exciting. “My team, Illumination, did exploratory data analysis to find important features for prediction, conducted feature engineering for creating new, supportive metrics; fitted two machine learning models, Random Forest and Long Short Term Memory, and obtained a high accuracy ranking in the top three among the final four teams”, he explained. “At the end, we presented our approaches and business insights to judges from Kabam.”

The Varsity Challenge with Kabam Games was open to all full-time UBC students, as a part of Dean Helsley’s mandate to collaborate with other departments at the university. A team from the Master of Data Science program at UBC won the challenge, collecting a $10,000 prize.

“The response we had was excellent”, said Sven Tapp, Head of Business Intelligence at Kabam.“Our partners from the MBAN program at UBC Sauder were very supportive of this event. They also helped us promote it throughout campus and I feel we ended up with a diverse group of participants from multiple programs.”

In total, 52 teams comprising 139 participants registered for the challenge, and of these, 76 were from UBC Sauder. “Students from the BCom, Master of Business Analytics, Master of Business Administration, and Master of Management programs participated,” said Elisabeth Chin, Manager, MBAN Student Experience.

UBC Sauder was instrumental in organizing the challenge. Right before the first submission, there was a two-day Hackathon conducted at the UBC Sauder Learning Labs, where participants consulted with professionals from Kabam.

Tapp explained that it was a deliberate decision for the competition to be more than just a data analytics exercise. “A successful data analyst must be able to put the findings in context, explain how the results are actionable, and be able to communicate the method, results and recommendations effectively,” he explained. “Kabam's data analysis team experiences this first hand, as they often communicate findings to game producers or executive leaders who do not have strong technical knowledge.”

“The hackathon was an experiential learning opportunity that allowed students to work on a ‘real-world’ problem,” said Professor Harish Krishnan, Academic Director of the MBAN program at UBC Sauder. He said that the challenge taught participants to work through all stages of an analytics project, from clearly understanding the problem, to structuring their approach, performing the analysis, and developing and communicating recommendations.

In addition to the hackathon, the challenge took place in two parts. The first was a quantitative analysis of a data set from one of Kabam’s games. Participants were asked to create models to predict the values of four variables and write a short essay proposing a player quality score metric. Finalists were chosen based on their metric accuracy, and for the second part, they had to present to a panel comprised of Kabam's Principal Data Scientist, Head of Community Marketing, and Industry Consultant to the CFO. The panel was deliberately selected to be cross-functional so teams would be forced to present their ideas in laymen’s terms rather than relying on technical jargon.

Mu, one of the MBAN finalists in the challenge, credits his UBC Sauder studies with helping him reach that stage. “I could never have reached the final without what I learnt in the MBAN program. For example, UBC Sauder has offered programs and opportunities for me to study statistics and machine learning, which I find so interesting. I’ve taken my learning beyond the course work,” he said. “I’ve also practiced many presentations during my program, and this definitely helped me to confidently and effectively present to Kabam.”