We've looked at many different algorithms with varying run times.
Some problems, like topological sort and stable marriage, admitted low-overhead linear-time (in the size of the input) algorithms. Others, like solving a system of linear equations and sorting, took more time.
Many of the problems also had simple exponential-time algorithms, but we generally felt that these were not the ones we wanted.
These kind of observations have led theoreticians to make a distinction between polynomial-time and exponential-time algorithms, with exponential meaning ``bad''.
