The most important spatial reference on a ⚽️pitch? Coordinate systems of data providers seem to suggest corner flags (@OptaPro@StatsBomb@Wyscout) or center mark (@TRACAB_). Since dimensions vary, this results in distortions and ambiguous coordinates. #soccer#analytics
In case you missed the presentation “The Origins and History of the Social Networks Perspective”, I left the workshop video on my personal webpage! (section: Short Workshops): https://t.co/gPZFgWF6ct #sna#socialnetworks#networkscience
@Jiang_Siyu98, MSc in Psychology: Learning Sciences from @LMU_Muenchen, joins as a doctoral student for a project on social networks and learning, supervised by @c_stadtfeld . Welcome to @sn_ethz ! 💐✍️😃
1/n Members of the @sn_ethz are ready for the european social network conference #EUSN2023 in Ljubljana! #sna#networkscience#socialnetworks We are actively participating as presenters and organisers of an event:
Preprint https://t.co/BRMryjceEQ submitted just in time. Algorithm extends any match outcome model to @FIFAWWC winning probabilities in split second, w/o simulation.
cc: @ivanasmokovic@schochastics, @RoyalStatSports prediction competition
A: no simulation at all. Given any match outcome model, tournament winning probabilities can be computed exactly - and in the time of fewer than 500 simulation runs. (2/4)
While the preprint is being polished, here are the results for a simple model based on the @FIFAcom World Ranking.
Joint work with Gordana Marmulla and @ivanasmokovic. All hoping for a different outcome. (4/4).
The key observation is that team trajectories are dependent only after mixing. Single team trajectories can therefore be computed from joint probabilities of pairs of teams advancing together. (3/4)