Data Supplement

Data Supplement

Files in this Data Supplement:

  • 10% informed in a group of 200 people (Footage courtesy of Quarks & Co/WDR). - The 20 informed individuals were wearing turquoise baseball caps and were trying to reach target 4 (out of shot on the centre right). The group quickly became stretched out towards the target as the informed individuals positioned themselves at the edge of the group closest to their target. After 75seconds the whole group reached the target of the informed individuals without splitting.
  • Conflict 20 versus 10 informed (Footage courtesy of ?Quarks &Co?/WDR). - The 20 informed individuals (wearing purple baseball caps) quickly reached their target (number 5, slightly to the left at the top of the shot) and within 60 seconds had taken approximately half of the group with them. The group then proceeded to become stretched out across the arena as the 10 informed individuals (wearing yellow baseball caps) attempted to reach their target (number 11, directly opposite the target of the 20). After approximately 110 seconds they had also reached their target and managed to take at least 40% of the group with them. A bridge of people remained between the 2 targets with a constant oscillation of people between the 2 targets.
  • Side view of torus with 200 people in Cologne, Germany (Footage courtesy of ?Quarks&Co?/WDR). - Interestingly, lane formation in a torus as observed in our experiment has never been described from animal groups. In fish and bats it is unlikely to arise because of the turbulence that opposing lanes would produce which would increase energy expenditure. However, lane formation (though not in a torus context) is well described in ants (Couzin & Franks 2003; Dussutour et al. 2004, 2005) where just like in humans (Helbing & Molnar 1995; Helbing et al 2001) it is unlikely to increase locomotion costs.
  • Aerial view of torus with 100 people in Freiburg, Germany (Footage courtesy of ?Die grosse Show der Naturwunder?/SWR).