The buying decisions of consumers towards movies, music CD’s, video DVD’s and other forms of purchasable entertainment can be modeled after an information network. A simple network model can be used to show how movie critics, for example, can make, or break, a movie. A person, P’s, binary decision of whether or not to watch a movie can be modeled as an appreciation score with a cutoff threshold t_a. Suppose P has a network of friends and critics that provide advice in the form of a simple “thumbs up, thumbs down” system (thumbs up z=1, thumbs down z=0). Furthermore, each friend has some reliability, established before the model is ran, labeled “r_i” and P has made his own evaluation of the movie by watching the trailer. P’s own evaluation can be called z₀ and P also has a reliability score for his own evaluation r_0. P creates an appreciation score A by summing up the products of the advice, z_i, and the reliability r_i. A=SUM(z_i*r_i, for all i) + r₀*z₀. If A is above t­_a­ then P will go and see the movie, otherwise P will not see the movie. This model only becomes interesting if r₀ < SUM(r_i, for all i), or if the person weighs the net advice of others as more significant than his own information. If a special type of person is introduced into the network, a critic, then the dynamics of the network change significantly. Firstly, a critic does not listen to anyone else’s advice, and generally has many outward edges with very high reliability ratings. In this sense, the critic can exert vast influence over other movie-goer’s decisions. Assume for now that most critics are in agreement that a given movie is worth seeing (z=1). After the first iteration of this game, the first few people to watch the movie are those with very high opinions of the critics, and those with very high expectations and high self-reliability ratings. Also, people with low thresholds t_a, will see the movie as well, since it does not take very much convincing to get them to go. It can be shown that such a setup exhibits information cascades because a decision is based not only on internal information, but information given by other trusted individuals. After several rounds, a movie can become a “must see” movie, where regardless of what a person thinks about a movie, he must go see it because of the overwhelming advice of his/her peers.

Some interesting aspects of this model are the values given to each parameter. T_a, for example, can depend on how much money or free time a given person has. The more money and time a person has, the lower their threshold will be to see a movie, and the less convincing it will take to get them to go and see the movie. Thresholds can change with time as well because it may take more enthusiasm to want to see a movie on a Wednesday night versus a weekend night. The reliability scores can be any number, where the greater the number, the greater a given person’s faith is in another person’s advice. One interesting aspect of the model that is difficult to explain is the establishment of the reliability ratings. It should be interesting to see how, using a feedback system, reliability ratings can be established. For example, if a person’s opinion of a movie afterwards was in agreement with a critic’s review, his/her reliability rating increases, and likewise decreases for a disagreement. Some implications of this model can explain why movies are shown to critics before the general public, and why some movies are released to limited audiences in select locations at first. Also, time varying thresholds can explain why large budget movies are released near the weekend. It should be interesting to see if the given process leads to Bose Einstein distributions like those found experimentally here.