Working with talent: A human approach to analytics

by Simon Demediuk, Lead Game Analyst, Weavr

As a researcher focused on games analytics, I’m not used to working with people to get to the number, model or prediction that I want. Working with esports talent as part of the Weavr R&D team has changed that.

When I talk about talent I mean the broadcasters, commentators, analysts, professional players and coaches who live and breathe esports. These people have years of experience and love the industry they work in. They want to see the output of R&D enhancing both gameplay and the viewing experience, and to be a part of that progress.

Without a doubt, working with talent has helped to drive new research angles. As lead analyst for Weavr, I’m faced with so many metrics to consider for both the back end model and the forward facing app. My team and I look to the talent for their thoughts on what makes a good story and what will make sense in terms of what we want to tell the audience.

Weavr is primarily a story driven application and so we want to make sure that we are in line with the story that the broadcasters and talent are already telling. We want to enhance that narrative rather than compete with it.

For example, understanding what metrics the talent take into consideration when deciding the Most Valuable Player led to us including some additional elements into our Performance Index metric on the data side. Insights from the talent made sure that some metrics we weren’t previously considering were included and resulted in an enhanced model.

We also talk to the talent about the forward facing app. We ask if the app is easy to understand, does it help them in their broadcasts, and is the information what they as a storyteller would give to the audience. We also ask them what metrics they would like to see.

For me, the only small disadvantage of working with talent is that much of their feedback is positive. As an academic researcher I’m used to hyper critical feedback about my work. I guess it’s up to us to try and encourage more critical feedback from talent in the future but I’m not sure yet how we address that.

I think it’s important to add that we also undertake valuable research with the general public and high-ranking (non-professional) players, but our focus with these groups is usually more on the app functionality rather than the back end model.

Working with talent brings a new and interesting vibe to game analytics. As we develop new features that help the talent shape their storytelling, coaching or game play, the talent feedback on what we could do to adapt or enhance those features. It’s a real team effort and I think there’s a long and exciting future ahead.