Atomistic models for friction suffer from the severe length- and time-scale restrictions of molecular dynamics. Even when they yield good qualitative results, it is difficult to draw meaningful quantitative conclusions from them. In this presentation, a novel approach to quantify the scratching work and the energy associated with plastic activity is used. The approach is combined with a statistical criterion to determine the significance of simulation box size, microstructure and sliding rate effects on the friction coefficient. These two methods are applied to a large parametric molecular dynamics study of single-asperity aluminum nano-scratch on mono-crystalline and poly-crystalline substrates. The results show that the simulation size effects are a considerable obstacle to understanding the atomistic origins of friction -- using present-day computing hardware -- as they have a strong influence on the core mechanisms of sliding friction, therefore motivating the development of 3D multi-scale methods for a hybrid nano- and micro-scale description of plasticity.