In many navigation and search problems navigational information can come from
direct sources, such as visual or olfactory cues, or indirect sources, such as
the movement directions of group neighbours or idiothetic movement
(dead-reckoning). It is not immediately obvious how best to combine these
direct and indirect sources of navigational information when determining how
and where to move. I will demonstrate using theoretical arguments that in
scenarios with realistic levels of error, the best navigational strategy for
individual and collective movement is to counterintuitively rely more on
indirect navigational information than direct navigational cues. I will
discuss the potential difficulties in validating these predictions using real
movement data, and relate this to more general problems inherent with path
analysis of animal movement.