> Why do we need the science of complex systems? If there are dependencies in the systems, then statistics don’t work. Standard calculus can’t describe things properly when there are abrupt large scale changes that involve changes in what many individuals are doing.