“…This 'dual' character of the CDF suggests a power law with a cutoff, since the power function does not fit the overall distribution, but only the most-connected nodes. The above visual observations are supported statistically by curve estimations based on OLS and the relevant log-log transformations (Faloutsos, Faloutsos, and Faloutsos 1999;Gorman and Kulkarni 2004;Patuelli et al 2007;Schintler et al 2004;Tranos 2011;Reggiani, Bucci, and Russo 2010). The results of the OLS are presented in Table 3, where three different forms are tested: exponential, power, and power with cutoff (Tanner function), accordingly: ,…”