Percolation on strings and the cover-up of the c=1 disaster

We study percolation on the worldsheets of string theory for c=0, 1/2, 1 and 2. For c<1 we find that critical exponents measured from simulations agree quite well with the theoretical values. For c=1 we show how log corrections determined from the exact solution reconcile numerical results with the KPZ predictions. We extend this analysis to the large c regime and estimate how finite-size effects will effectively raise the ground-state energy, masking the presence of the tachyon for moderate values of c>1. It thus appears likely that simulations for c=2, 3 .. on numerically accessible lattices will fail to capture even the qualitative behavior of the continuum limit. (orig.)