“…In fact, our estimation techniques significantly extend those for the case (ii) above in [17]; due to the increased complexity of the Quadratic Maximum Social Cost function (over Maximum Social Cost), far more involved estimations have been required in the present proof. Counterexamples to the FMNE Conjecture appeared (i) for the case of unrelated links in [17], and (ii) for the case of weighted users in [5]. In the context of selfish routing, the fully mixed Nash equilibrium and the FMNE Conjecture have attracted a lot of interest and attention; they both have been studied extensively in the last few years for a wide variety of theoretical models of selfish routing and Social Cost measures -see, e.g., [2,4,9,10,12,16,18].…”