Several recently proposed semi-automatic and fully-automatic coarse-graining schemes for polymer simulations are discussed. All these techniques derive effective potentials for multi-atom units or super-atoms from atomistic simulations. These include techniques relying on sin,le chain simulations in vacuum and self-consistent optimizations from the melt like the simplex method and the inverted Boltzmann method. The focus is on matching the polymer structure on different scales. Several ways to obtain a time-scale for dynamic mapping are discussed additionally. Finally, similarities to other simulation areas where automatic optimization are applied as well are pointed out. (C) 2004 Elsevier Ltd. All rights reserved.