We employ an axiomatic approach. Assume the existence of a 2×2 payoff matrix for students corresponding to the teacher’s binary stance? either sheep or wolf. Let α denote the proportion of students who prefer chatting, and 1? α the proportion who dislike it. The student group adopts a strategy that maximizes their aggregate utility, and linear programming reveals that when α is around 0. 5 or lower, the wolf stance is more effective in suppressing in-class chatter. However, if the teacher is sheep stance and is able to suppress strategy x? ? , then even as α increases, the sheep can silence the classroom more effectively. This result is nontrivial. Moreover, assuming that students’ preferences toward chatting switch according to a Poisson probability process, it is proved that the steady-state value of α depends on λ/μ, where λ, μ are the parameters defining the Poisson density of this switching behavior. According to the author’s observational data, λ/μ was approximately 21.
Tomonori Koyama (Sun,) studied this question.