Reporting Back from Nucl.AI 16

Posted on Posted in Blog - Company

At Scientific Revenue, we apply machine learning to very hard problems (namely pricing) and we frequently find ourselves on the bleeding edge of machine learning techniques. As a result, we like to collaborate, and we love going to conferences like NUCL.AI, which features a mix of leaders from both industry and academia, all focused on artificial intelligence in gaming and the creative industries.


This year’s conference was the best yet. While I spoke in the Analytics and Data Science track (slides here), I spent quite a bit of time in the Real Time Decisions and Cognition, Bots and Understanding tracks as well.

For me, the highlights included:

  • An excellent talk by Julian Runge (from Wooga) on how to measure Word of Mouth marketing. While the conclusion wasn’t very surprising (the number one factor in word of mouth marketing is how fun the game is), the methodology and examples were compelling. And when you incorporate the equally excellent subsequent talk by Santiago Murillo (from Social Point) on how to measure the impact of television advertising and take steps towards true multi-channel attribution, it’s clear that mobile game companies are leading the way in understanding user acquisition.
  • Two “hallway conversations” on the evolving nature of gaming.
    • The first was a wide-ranging breakfast discussion on the historical trends in gaming, starting with the game of Go three thousand years ago (extremely simple rules, easy to describe and very easy to make a board, even in pre-industrial societies) , heading to games like Agricola (stunningly complex rule set and literally hundreds of different machine-carved wooden pieces) and then to modern video games which have even more complex rule sets that constantly evolve.
    • Then, later that same day, I had a long conversation with some people from Supercell about the evolution of gameplay in Clash of Clans which mirrored my breakfast discussion. Supercell has noticed that introducing new player types (for example, the recent introduction of the miner from Clash Royale) has had dramatic impacts in the way people play Clash of Clans, and the strategies that are effective. The extent to which Supercell tracks the evolution of strategy, and the extent to which the YouTube community also does so, are both very impressive.
  • The fact that everyone is hiring. You hear a lot about game companies laying people off, and about studios shutting down. But it’s clear that, on the data science side of the house, there are a LOT of open positions, and a huge number of gaps in most organizations.

All in all, an excellent conference. I hope to return again next year, and having more great conversations on the state of AI in mobile gaming.