Noting that mathematical modeling is a relatively new phenomenon in higher education and that much can be learned from the misdirections and mistakes that characterize modeling in general, the author describes major criticisms of modeling and suggests improvements, particularly in communication between modelers and potential model users. (JT)