Use of Gaming Procedures in Evaluation of Management Experiments

摘要: Management often involves debate about whether to maintain some conservative, status quo policy or to experiment with some risky alternative that may produce higher long-term value. In comparing such alternatives, a key modelling problem is to predict the transient response pattern that will occur if an experimental approach is adopted but is not successful (stock collapses or fails to recover, enhancement makes no net conributution to catches, etc.), in terms of how long it will take to detect and react to the failure. Prediction of such transients involves modelling not only future data gathering and statistical analysis, but also how future managers will react to the information. Management gaming involving real managers may be the simplest and most efficient way to make such predictions. Application of gaming procedures to fisheries development and hatchery production examples indicates that such procedures may not only help in prediction, but also are useful for management training and clarification of objectives and priorities.