Very interesting article, I've noticed the tendency for people to give more relevance to the wrong data points in all sorts of problems I've come across in my life, school and work experience..precisely as described in this writing. It is very interesting how people applying the wrong algorithm of choices to their lives, end up making the wrong choices. It would be interesting if the researchers could apply their "win score" algorithm to other complex relationships and interactions to make predictive projections. Imagine running the algorithm in a hypothetical but possible complex problem space and then letting the predicted win score for a given course of actions in that problem space determine a real world response to the problem should it ever materialize. In computer science research, scientists have been doing something very similar to solve complex problems for years called "genetic algorithms", basically they use stochastic processes on autonomous programs to allow the interactions of these programs over time to "evolve" efficient solutions to a given problem. These algorithms have been put to good use by some investors in the field of quantitative analysis, the brokers that use such tools to buy and sell stocks are called "quants" and have made consistent success using their tools. It would be interesting to see these ideas and algorithms applied to the social domain.