Analytics & Data Science (Day #2)

Join experts from Blizzard, Scientific Revenue, and leading mobile or games studios for a day-long course on the state-of-the-art of analytics techniques for understanding and enhancing player experience and design choices—whether for games, bots or apps—using combinations of machine learning and clever modeling tricks!

Now in its 8th edition, nucl.ai brings together hundreds of developers from creative industries and interactive media to resolve the biggest technical and design problems with Artificial Intelligence. One of the event's tracks focuses specifically on “Analytics & Data Science” for an entire day.
(Day #2: Tuesday, July 19th)

Regular & Current Attendees

  • Blizzard Entertainment
  • Nordeus
  • Scientific Revenue
  • Riot Games
  • Creative Assembly
  • Ubisoft Studios
  • Wooga
  • Unity Technologies
  • CD Projekt Red
  • King.com
  • Square Enix
  • Jagex
  • Microsoft Game Studios
  • Natural Motion
  • Mind Candy
  • Epic Games
  • Zynga
  • CCP Games
  • ...

I've never seen so many industry figures, academics, and students happy to discuss a range of topics around AI in games. It was really, REALLY helpful.

Matthew Bedder

PhD Student

University of York, UK

It was great to return to #nuclai15 this year, I met loads of interesting people and came away buzzing with new ideas.

Vicky Smalley

CTO

Small Jelly, UK

It was my first time at #nuclai15 and I heard about it very late through co-workers. I was mostly interested in the Character Animation panels. I really hope to be able to attend every year from now on! Great information, great people, great time!

Laurent Delayen

Senior Gameplay Programmer

Epic Games, USA

For the first time in my life, I left a conference with concrete and crystallized plans to improve my current project based on the talks I attended and the conversations I had with people. Incredibly. Useful.

Sergey Mohov

Game Designer

Tequila Works, SPAIN

First time visiting a conference like this, and #nuclai15 made a very strong first impression. Great sessions, great people, great time!

Aaron Oostdijk

Lecturer & Researcher

University of the Arts Utrecht, NETHERLANDS

An invaluable resource of ideas and enthusiasm. Long may it continue.

Jim Brewster

Technical Lead

NaturalMotion, UK

Spent months of saving and talking to Alex on twitter – it was more than worth it. I cannot express how this experience changed me and might as well have shaped my future. From the networks I've created to the feedback needed to get back on track I am extremely happy for this opportunity and #nuclai16 is on my calendar.

Mohamed Serry

Gameplay Programmer

Creability, UNITED ARAB EMIRATES

#nuclai15 was really unique and cross-discipline, bringing feature film, game AI, animation and machine learning people and talks together makes it the perfect opportunity to learn for everyone.

Alexander Shafranov

Animation Programmer

Remedy Games, FINLAND

Inspiring talks, dedicated people, things to play with, that's a conference to my taste.

Anika Uhlemann

Engine & Game Programmer

GOODGAME Studios, GERMANY

Attending conferences like Nucl.ai brings me in the perfect mental state. I always try to invent something crazy in the plane back home...

Simon Clavet

Animation & Physics Programmer

Ubisoft Montreal, CANADA

Wide variety of topics, great speakers, interesting conversations and lots of very very cool stuff to see and try - Nucl.ai easily beats other conferences!

Kevin Schmidt

Head of Data Science

Mind Candy, UK

#nuclai15 was again a wonderful opportunity to focus on game AI and the intersection of game design and programming. Great talks and great conversations.

Jurie Horneman

Independent, FRANCE



Applying analytics to game design is not about driving by looking in the rear view mirror while ranting about KPIs. Analytics can be used to apply real-time or just-in-time intelligent design judgement to the logic of who sees what when, and to the multidimensional dynamic challenges of balance and matchmaking. Our experts explain their creative analytical solutions to a variety of challenges in game design and experience design, across a variety of platforms and genres. Our goal is to give you some knowledge you can take home and apply. Come join the fun!

Keywords

  • adaptive systems (pricing)
  • pattern recognition
  • real-time user profiling
  • machine learning

Format

  • Keynote
  • Presentations
  • Panel Discussion
  • Expert Q&A

Track Schedule (Day #2: Tuesday, July 19th)

Track Schedule (Day #2: Tuesday, July 19th)

Day #2: Tuesday, July 19th

Agent Behavior & Coordination

Procedural Content Generation

Analytics & Data Science

Alex Zook, Blizzard Entertainment

12:00 A Brief History of Matchmaking in Heroes of the Storm (Amphitheatre)

Tickets!

Join developers from all over the world and secure your ticket for this year's nucl.ai now! Three days of artificial intelligence theory and practice packed into three parallel tracks per day with many networking opportunities are waiting for you in Vienna.

Conference* (3 Days)

All Content* (3 Days)

Single Day Ticket

Early Bird

A limited number of these tickets was available.

Sold Out.

Sold Out!

-

Early Worm

These tickets were available for a limited time.

Expired.

Sold Out.

-

Regular

Important: All Access sold out. All Content available instead.*

€ 369

€ 999

€ 399

Student Pass**

Available until Friday July 8th

60% off above

60% off above

60% off above

Indies or Unemployed Devs***

Available until Friday July 8th

20% off above

30% off above

20% off above

* All Access tickets give you access to all rooms on all days of the conference, Single Day tickets give you full access, but only on the day of your choice. With a main conference ticket you can access the amphitheatre and the open laboratories. See here for more details about ticket types and discounts.

Keynotes and Presentations

A Brief History of Matchmaking in Heroes of the Storm

Matchmaking is a perennial problem for competitive game design. Conceptually the problem is simple: assemble two teams of players to create a fun match. In practice there are a wide range of challenges. How do you create fair match ups? How do you model player skills? What do you do for new players? What metrics help track the health of a matchmaking system?

This talk will cover several of the core iterations of the Heroes of the Storm matchmaking system. We'll cover the ways matchmaking has worked, the problems we've encountered, and the solutions we've developed to address these problems. Along the way we'll cover some of the simulations and data-driven models we've developed to deliver high-quality matches to Heroes of the Storm players.

Alex Zook, Blizzard Entertainment

Alex Zook is a Senior Data Scientist at Blizzard Entertainment and PhD candidate at the Georgia Institute of Technology. As a Blizzard data scientist he's work on problems including identifying gameplay styles of World of Warcraft players, analyzing map design in Overwatch, and refining matchmaking in Hearthstone and Heroes of...

Read full biography »

Alex Zook, Blizzard Entertainment

Alex Zook is a Senior Data Scientist at Blizzard Entertainment and PhD candidate at the Georgia Institute of Technology. As a Blizzard data scientist he's work on problems including identifying gameplay styles of World of Warcraft players, analyzing map design in Overwatch, and refining matchmaking in Hearthstone and Heroes of...

Read full biography »

Beyond Churn Prediction: An Introduction to Uplift Modelling

In the games industry, a common approach to diminishing player churn is to use machine learning to score users based on their churn potential and to then target them with specific messaging or offers. However, this approach may be ineffective since it does not optimise what is called "true lift" or "uplift" : the expected incremental gain when the potential churner is targeted.

This talk aims at introducing uplift modelling and will cover most of the theory (how to create the machine learning model and how to evaluate it) as well as practical concerns (when uplift is profitable, the data needed, A/B test design...). It will be illustrated with a real example from a mission at Ankama for reducing player churn in the MMORPG Dofus.

Pierre Gutierrez, Dataiku

Pierre Gutierrez is a senior data scientist at Dataiku. As a data science expert and consultant, Pierre has worked in diverse sectors such as e­business, retail, banking, but also gaming. He has experience in various topics such as fraud detection, bot detection, recommender systems, or churn prediction.

Read full biography »

Pierre Gutierrez, Dataiku

Pierre Gutierrez is a senior data scientist at Dataiku. As a data science expert and consultant, Pierre has worked in diverse sectors such as e­business, retail, banking, but also gaming. He has experience in various topics such as fraud detection, bot detection, recommender systems, or churn prediction.

Read full biography »

Capturing, Analysing and Imitating Emergent Playstyles in Card Games

AI Factory's Spades game on mobile is currently the top rated Spades game in the Google Play store. In collaboration with the developers, we have collected gameplay data from 27 592 games and statistics regarding wins/losses for 99 866 games using Google Analytics. Using the data collected, an initial study analysed the difficulty and behaviour of an Information Set Monte Carlo Tree Search (ISMCTS) player we developed and deployed in the game previously. In doing so we noted significant differences in the playstyle of the ISMCTS player when compared to the emerging playstyles of the human playerbase. In this talk I will focus on how we clustered decision tree representations of player behaviour and biased the ISMCTS to reduce this difference whilst maintaining the playing strength of the previous un­biased ISMCTS.

Sam Devlin, Digital Creativity Labs

Dr Sam Devlin (http://www-users.cs.york.ac.uk/~devlin/) received an MEng degree in Computer Systems and Software Engineering from the University of York, UK, in 2009. In 2013, he completed his PhD on multi-agent reinforcement learning at the University of York and visited Oregon State University funded by a Santander International Connections Award. His...

Read full biography »

Sam Devlin, Digital Creativity Labs

Dr Sam Devlin (http://www-users.cs.york.ac.uk/~devlin/) received an MEng degree in Computer Systems and Software Engineering from the University of York, UK, in 2009. In 2013, he completed his PhD on multi-agent reinforcement learning at the University of York and visited Oregon State University funded by a Santander International Connections Award. His...

Read full biography »

Estimating the Cross Channel Effect of TV Advertising Campaigns for Mobile

Tracking the effects due to television in an automated manner is a central problem of television advertising. This issue is especially important in digital economy, like mobile games, where almost all the conversion funnel happen online with the exception of offline communication. Without quantification of television effects, marketers may fail to effectively use both online and offline channels. We propose a method to estimate the effect of each TV spot emission on a multichannel campaign. The probabilistic approach allows to measure the impact of overlapping communications in television isolating the effect from digital marketing actions. The method outperforms the traditional attribution method and direct response models. Our method allows a dynamic optimization of broadcast advertising campaign with a high granularity.

Santiago Murillo, Social Point

Santiago Murillo, Data Scientist at Social Point. He holds an Industrial Engineering Degree and a title of researcher in computer science from the Polytechnic University of Catalonia (UPC). He has more than 10 years’ experience working with technologic solutions and innovations in various industrial sectors. In the last years he...

Read full biography »

Santiago Murillo, Social Point

Santiago Murillo, Data Scientist at Social Point. He holds an Industrial Engineering Degree and a title of researcher in computer science from the Polytechnic University of Catalonia (UPC). He has more than 10 years’ experience working with technologic solutions and innovations in various industrial sectors. In the last years he...

Read full biography »

Causal Inference, Reinforcement Learning, and Continuous Optimization

Most modern analytical systems are descended, in perspective if not in actual lineage, from "business intelligence" and are, really, something that the software developers who used Crystal Reports to format data pulled from a FoxPro database, could feel a spiritual kinship to. That is, they are designed to sift through vast amounts of information and present summaries to humans, who then take action (or take notes for the next quarterly report).

The current passion for machine learning and predictive analytics has, for the most part, not managed to escape this. When you use supervised learning for churn prediction, the usual goal is to predict who is going to leave the system, and enable humans to take action. Similarly, a typical "A/B testing" scenario usually involves running a small number of simultaneous variations in a system for a short period of time, and then employing a significance test to determine a winner, which then becomes the default behavior of the system.

What if, instead, you want to continuously optimize a complex system? What if you are running a game (as a service) and want to customize the game to each player?

How would you do that? In this talk, we'll go through some of the lesser­well­known tools from the machine learning toolkit, and talk about how to design systems that continuously and constantly adapt to the player.

Bill Grosso, Scientific Revenue

Bill is the founder and CEO of Scientific Revenue, the leading provider of dynamic pricing for in-app purchases. Prior to starting Scientific Revenue, Bill ran a consulting company focused on business intelligence and predictive analytics, and was an entrepreneur in residence at Rustic Canyon Venture Partners. Prior to those positions,...

Read full biography »

Bill Grosso, Scientific Revenue

Bill is the founder and CEO of Scientific Revenue, the leading provider of dynamic pricing for in-app purchases. Prior to starting Scientific Revenue, Bill ran a consulting company focused on business intelligence and predictive analytics, and was an entrepreneur in residence at Rustic Canyon Venture Partners. Prior to those positions,...

Read full biography »

People Play, People Talk? Inferring Userbase Dynamics from Sparse Data

A deep dive into understanding organic downloads and lessons learned! Players not coming from user acquisition are key for marketability of a game in light of the struggle to keep lifetime value above effective cost per install. Unfortunately, data on these users is sparse. It is mostly unknown why they download an app. Did they hear about it through family and friends (word-of-mouth)? Did they discover it on the App/Play Store? Did they watch a TV ad? Do they know your brand and searched for it? This talk focuses on what is widely regarded as the most important source of organic downloads in the mobile app ecosystem: word-of-mouth. More precisely, it looks into the share of word-of-mouth in organic downloads and the quality of the resulting traffic. Finally, a framework of the drivers of word-of-mouth for mobile apps is sketched.

Julian Runge, Wooga

Julian works for Wooga, the Berlin based developer of mobile games like Jelly Splash, Pearl's Peril, and most recently Futurama - Game of Drones. After heading analytics and data science for Wooga for a couple of years, he most recently moved to a ‘researcher in residence’ role. In this new...

Read full biography »

Julian Runge, Wooga

Julian works for Wooga, the Berlin based developer of mobile games like Jelly Splash, Pearl's Peril, and most recently Futurama - Game of Drones. After heading analytics and data science for Wooga for a couple of years, he most recently moved to a ‘researcher in residence’ role. In this new...

Read full biography »

Using a Genetic Algorithm in 10 vs. 10 Matchmaking in TOTAL WAR: ARENA

Based on work on Total War: Arena, Christian Rothe will talk about his experience implementing a genetic algorithm to search for good solutions for 10 vs. 10 multiplayer matchmaking based on multidimensional optimization criteria. The talk will cover modelling the problem space to the GA approach and touch non­obvious pitfalls encountered during implementation and performance optimization.

Christian Rothe, Creative Assembly

Christian is currently working at Creative Assembly on the upcoming Total War: Arena title. His main interest lies with highly scalable online systems, but he also enjoys all kinds of other challenges (which brings him to nucl.ai). Prior to joining Creative Assembly, Christian worked as a freelancer on a variety...

Read full biography »

Christian Rothe, Creative Assembly

Christian is currently working at Creative Assembly on the upcoming Total War: Arena title. His main interest lies with highly scalable online systems, but he also enjoys all kinds of other challenges (which brings him to nucl.ai). Prior to joining Creative Assembly, Christian worked as a freelancer on a variety...

Read full biography »

Are (All) Players Worth the Trouble? Modelling Acquisition and Population Dynamics [Panel]

Heather Stark, Harman International

Heather is a games analyst who has worked on over a dozen games, as a consultant and in-house, across a wide range of genres and platforms. She is an active member of several London tech meetups and has a background as a consultant, tech analyst, and HCI researcher. Recently, she...

Read full biography »

Heather Stark, Harman International

Heather is a games analyst who has worked on over a dozen games, as a consultant and in-house, across a wide range of genres and platforms. She is an active member of several London tech meetups and has a background as a consultant, tech analyst, and HCI researcher. Recently, she...

Read full biography »

Common Techniques for Profiling and Modeling Individual Users [Extended Q&A]

Alex Zook, Blizzard Entertainment

Alex Zook is a Senior Data Scientist at Blizzard Entertainment and PhD candidate at the Georgia Institute of Technology. As a Blizzard data scientist he's work on problems including identifying gameplay styles of World of Warcraft players, analyzing map design in Overwatch, and refining matchmaking in Hearthstone and Heroes of...

Read full biography »

Alex Zook, Blizzard Entertainment

Alex Zook is a Senior Data Scientist at Blizzard Entertainment and PhD candidate at the Georgia Institute of Technology. As a Blizzard data scientist he's work on problems including identifying gameplay styles of World of Warcraft players, analyzing map design in Overwatch, and refining matchmaking in Hearthstone and Heroes of...

Read full biography »

Future Directions for Analytics and Machine Learning [Roundtable]

Sam Devlin, Digital Creativity Labs

Dr Sam Devlin (http://www-users.cs.york.ac.uk/~devlin/) received an MEng degree in Computer Systems and Software Engineering from the University of York, UK, in 2009. In 2013, he completed his PhD on multi-agent reinforcement learning at the University of York and visited Oregon State University funded by a Santander International Connections Award. His...

Read full biography »

Sam Devlin, Digital Creativity Labs

Dr Sam Devlin (http://www-users.cs.york.ac.uk/~devlin/) received an MEng degree in Computer Systems and Software Engineering from the University of York, UK, in 2009. In 2013, he completed his PhD on multi-agent reinforcement learning at the University of York and visited Oregon State University funded by a Santander International Connections Award. His...

Read full biography »

Speakers & Organizers

Alex Zook

Pierre Gutierrez

Sam Devlin

Santiago Murillo

Bill Grosso

Bill is the founder and CEO of Scientific Revenue, the leading provider of dynamic pricing for in-app purchases. Prior to starting Scientific Revenue, Bill ran a consulting company focused on business intelligence and predictive analytics, and was an entrepreneur in residence at Rustic Canyon Venture Partners. Prior to those positions, he held senior positions in engineering and product at various companies, where he was well known for building high-performance organizations which shipped innovative products on time. He has also written two books on software design, has worked as a research scientist in artificial intelligence, and is the author or co-author of over 40 academic journal articles.

Conference Speaker

Santiago Murillo

Data Scientist

Santiago Murillo, Data Scientist at Social Point. He holds an Industrial Engineering Degree and a title of researcher in computer science from the Polytechnic University of Catalonia (UPC). He has more than 10 years’ experience working with technologic solutions and innovations in various industrial sectors. In the last years he has led the development of machine learning tools for marketing technology field and also a number of R&D projects supported by the Spanish Ministry of Science and Innovation, which focused in developing technology for data mining and artificial intelligent applied in digital advertising industry.

Conference Speaker

Research Associate

Dr Sam Devlin (http://www-users.cs.york.ac.uk/~devlin/) received an MEng degree in Computer Systems and Software Engineering from the University of York, UK, in 2009. In 2013, he completed his PhD on multi-agent reinforcement learning at the University of York and visited Oregon State University funded by a Santander International Connections Award. His research interests are focussed on artificial intelligence, data mining and machine learning. He was a Research Associate on the EPSRC funded New Economic Models and Opportunities for digital Games (NEMOG) project from 2013-2015, working on data mining for collective game intelligence. He now holds a full academic role as a transitional fellow in the Digital Creativity Labs.

Conference Speaker

Pierre Gutierrez

Senior Data Scientist

Pierre Gutierrez is a senior data scientist at Dataiku. As a data science expert and consultant, Pierre has worked in diverse sectors such as e­business, retail, banking, but also gaming. He has experience in various topics such as fraud detection, bot detection, recommender systems, or churn prediction.

Conference Speaker

Senior Data Scientist

Alex Zook is a Senior Data Scientist at Blizzard Entertainment and PhD candidate at the Georgia Institute of Technology. As a Blizzard data scientist he's work on problems including identifying gameplay styles of World of Warcraft players, analyzing map design in Overwatch, and refining matchmaking in Hearthstone and Heroes of the Storm.

Conference Speaker

Julian Runge

Christian Rothe

Heather Stark

Alex J. Champandard

Alex J. Champandard

Conference Co-Ordinator & Co-Founder

Alex is the mastermind behind AiGameDev.com, the largest online hub for artificial intelligence in games. He has worked in the industry as a senior AI programmer for many years, most notably for Rockstar Games. He regularly consults with leading studios in Europe, most recently on the multiplayer bots for Killzone 2-3 at Guerrilla Games. Alex authored the book "AI Game Development" and often speaks about his research. He's associate editor for the IEEE Transactions on AI in Games, and serves on the program committee for the AIIDE and CIG conferences.

User Research Specialist

Heather is a games analyst who has worked on over a dozen games, as a consultant and in-house, across a wide range of genres and platforms. She is an active member of several London tech meetups and has a background as a consultant, tech analyst, and HCI researcher. Recently, she has been an early stage advisor for two analytics-related startups. She has been forgetting R syntax for over 30 years, dating from back when R was S. When she finds shiny things she shares them on @hastark.

Track Chair and Co-Organizer

Christian Rothe

Data Scientist

Christian is currently working at Creative Assembly on the upcoming Total War: Arena title. His main interest lies with highly scalable online systems, but he also enjoys all kinds of other challenges (which brings him to nucl.ai). Prior to joining Creative Assembly, Christian worked as a freelancer on a variety of game and non-game projects, co-founded a startup and worked at SAP.

Conference Speaker

Julian Runge

Researcher in Residence

Julian works for Wooga, the Berlin based developer of mobile games like Jelly Splash, Pearl's Peril, and most recently Futurama - Game of Drones. After heading analytics and data science for Wooga for a couple of years, he most recently moved to a ‘researcher in residence’ role. In this new capacity he focuses on the (rare, but sweet) projects that yield both tremendous business impact and meaningful scientific contributions. He also is a Ph.D. candidate at Humboldt University's school of business and economics in Berlin. He recently spent three months at Stanford University’s Graduate School of Business as a visiting researcher, deep diving on purchase decisions in the free-to-play and freemium business model. An economist by training, he developed a strong interest in machine learning in recent years, especially in its application to game design and marketing. His research in the field has been recognized by a best paper nomination at CIG 2014 and a best paper award at AIIDE 2015.

Conference Speaker

Browse Conference Tracks

Analytics & Data Science

Agent Behavior & Coordination

Character Animation

Cognition, Bots & Understanding

Dynamic Dialog & Storytelling

Generative Systems & Design

Human Interaction & Recognition

Procedural Content Generation

Real-time Decisions