47 GENIUS INTERVIEW Dr Florian Block
your game by 10%”, so you get a hyper- personalised interactive experience that weaves through the main broadcast. FEED: Can you tell me more about the different types of data you’re capturing? Are you collecting physical data about players too? FB: No physical data, although that would be interesting. We have toyed with heart rate monitors, but we’re mostly capturing the in-game data – the whole state of the game as it unfolds, how people move around the virtual environment, what actions they take. In a fantasy game, what spells they cast, what interactions they have with other players. Some games give you very good access, both in real time and retrospectively, about every little detail that happens in the matches. You can compare this to tracking data in football and other traditional sports. But whereas you have to spend a lot of money for visual tracking technology in those, in esports you get all the tracking tech built in and inately pumping out data, so you have a much richer set of data to work with. More importantly, the sector is very permissive and companies are usually quite generous about sharing the data. As a result, you can
get innovations that are unthinkable in football where you have to pay millions and can only get it if you are Sky Sports or the BBC. But everyone can get this data, so it’s a very exciting space. FEED: What games have you been working with? FB: We’ve been working with DOTA 2 and Counter-Strike: Global Offensive . We’re starting to work with League Of Legends and Hearthstone . FEED: Is there anything about these games that is particularly interesting from a data perspective? Or do you work with them because they are popular? FB: It’s both. Yes, these are all very popular games. But DOTA has the most comprehensive free access to data. Valve, the publishers, are very permissive in their data use. Hearthstone is a really interesting game in terms of vision technology. It’s easy to scan the video feed and look at what’s going on, because there are lots of static elements and lots of visual effects that make it easier to do image recognition. What I would like to see in the next five years is an open data standard
DECIPHERING THE DATA Complex patterns can be translated into easy to read content for the audience to enjoy
into something that adds value for the audience. It could create engagement for existing fans, but also potentially make esports more palatable for people who may be new to it. And with the rapid growth of esports you are beginning to face a mainstream audience. FEED: So you really are doing content creation, although it may not be via a linear, literary story. FB: Yes. So far we’ve injected narrative elements on top of the game narrative. But we are currently exploring how to use the rich data space you have that surrounds a live game and historic matches, plus a viewer’s own matches and data. We are interested in going into hyper-personalised, interactive experiences. On the viewer side, you may have the linear narrative that the commentator is giving – the one-for-all story – but you could also have a customised, tailored narrative that takes into account the viewer’s player data, with updates that say “This player is doing better than
feedzine feed.zine feedmagazine.tv
Powered by FlippingBook