The value of simplicity

Would you trust a strategy with three clearly defined conditions more than one with fifteen filters? At what point does adding confirmation become unnecessary complexity?