Comment on the Shiner–Davison–Landsberg Measure

The complexity measure from Shiner et al. [Physical Review E 59, 1999, 1459–1464] (henceforth abbreviated as SDL-measure) has recently been the subject of a fierce debate. We discuss the properties and shortcomings of this measure, from the point of view of our recently constructed fundamental, statistical mechanics-based measures of complexity Cs (γ, β) [Stoop et al., J. Stat. Phys. 114, 2004, 1127–1137]. We show explicitly, what the shortcomings of the SDL-measure are: It is over-universal, and the implemented temperature dependence is trivial. We also show how the original SDL-approach can be modified to rule out these points of critique. Results of this modification are shown for the logistic parabola.