Sometimes simulations are done with random rankings or random ratings, and sometimes that's better, but usually a spatial simulation is considered to be the most lifelike one. Randomly position candidates & voters in an issue space, of one or more dimensions (But of course if there are more than 3, it won't make any intuitive sense spatially). For best realism for U.S. political elections, have the issue dimensions strongly correlated. By the way, having more than 3 issue dimensions in a spatial simulation doesn't discourage me in the way that it would in Hoffman's method, because in Hoffman's method one is actually working with the geometry, to integrate a "volume" of many dimensions, between the many hyperplanes that are its boundaries. But it's still nice to keep to only 3 spatial dimensions, so that it makes intuitive sense as something that we could build a model of, or draw a perspective drawing of. Some simulations assume that candidates tend to be center-seeking, so that the candidates are more tightly distribued about the mean than the voters are. Mike Ossipoff _________________________________________________________________ Get your FREE download of MSN Explorer at http://explorer.msn.com