That method that eliminates everything not in the Smith set, and then sequentially drops the smallest defeats, that probably gives the same results as Smith//Condorcet(EM). That method, SC, may be the best easily-computable method when there are many candidates. Mike