Computational physics

A simulation samples from the wrong distribution, causing some physical states to be overrepresented. What happens if the researcher simply runs the same sampling procedure for vastly more draws?