Abstract: Whether making a run in behind the defence or being unmarked at the far post during a cross, scoring in football is often about finding space in dangerous areas on the pitch. In this talk I will discuss how we can highlight key moments during a match or identify talented attacking players by looking at off-ball scoring opportunities.

• Interpreting Deep Sports Analytics: Valuing Actions and Players in the NHL
  Guiliang Liu, Wang Zhu and Oliver Schulte
• Player pairs valuation in ice hockey
  Dennis Ljung, Niklas Carlsson and Patrick Lambrix
• Model Trees for Identifying Exceptional Players in the NHL and NBA Drafts
  Yejia Liu, Oliver Schulte and Chao Li

• Measuring football players' on-the-ball contributions from passes during games
  Lotte Bransen and Jan Van Haaren
• Forecasting the FIFA World Cup – Combining result- and goal-based team ability parameters
  Pieter Robberechts and Jesse Davis
• Predicting Pass Receiver In Football Using Distance Based Features
  Yann Dauxais and Clément Gautrais
• Football Pass Prediction using Player Locations
  Philippe Fournier-Viger, Tianbiao Liu and Jerry Chun-Wei Lin
• Deep Learning from Spatial Relations for Soccer Pass Prediction
  Ondřej Hubáček, Gustav Šourek and Filip Železný
• Predicting the Receivers of Football Passes
  Heng Li and Zhiying Zhang

• Evaluating NFL plays: Expected points adjusted for schedule
  Konstantinos Pelechrinis, Wayne Winston, Jeff Sagarin and Vic Cabot
• Distinguishing between roles of football players in play-by-play match event data
  Bart Aalbers and Jan Van Haaren
• Player valuation in European football
  Edward Nsolo, Patrick Lambrix and Carlsson Niklas
• Ranking the Teams in European Football Leagues With Agony
  Stefan Neumann, Julian Ritter and Kailash Budhathoki
• Characterizing Soccer Players' Playing Style from Match Event Streams
  Aron Geerts, Tom Decroos and Jesse Davis
• * Mining Soccer Data: Subgroup discovery of tactics from spatio-temporal data
  Laurentius Meerhoff, Arie-Willem de Leeuw, Floris Goes and Arno Knobbe

Abstract: Basketball is currently second only to baseball when it comes to integrating data in team operations, decision making and game preparation. While data and statistical analysis have been always part of basketball operations, the availability of detailed player tracking data as well as, additional contextual meta-data, have pushed the envelope further. In this talk, we will start by presenting a novel approach to an old problem, namely, 5-men lineup analysis and evaluation. In particular, we will show how (i) win probability models can be used to evaluate lineups, while (ii) we will also present a generic network-based approach for evaluating team units by learning latent node representations. Then we will present a tensor decomposition approach for learning latent patterns on context-rich basketball raw and meta- data. This can facilitate a variety of operations including pattern analysis, but also faster video retrieval, flexible data queries, as well as, potentially synthetic data generation.

• Real-time Power Performance Prediction in Tour de France
  Yasuyuki Kataoka and Peter Gray
• Automatic classification of strike techniques using limb trajectory data
  Kasper M.W. Soekarjo, Dominic Orth, Elke Warmerdam and John van der Kamp
• * Pacing Strategies and Athletic Performance in Long-Distance Running
  Arie-Willem de Leeuw, Rens Meerhoff and Arno Knobbe