Risk exposure
Ryan O. Murphy’s Homepagehttp://vlab.ethz.ch/ROM/DBGT/Homepage.html

    Is it better to start an endeavor by taking small risks, and then gradually increase risk taking over time? Or conversely is it better to “start big,” by taking large risks-- making aggressive moves-- and then decrease risk exposure over time? Or, as a final strategy, is it better to keep risk exposure at a constant level over the duration of an endeavor? These three questions all relate to how decision makers change their risk taking behavior in a dynamic environment where they can sequentially change their level of risk exposure. Regardless of how they do it, presumably decision makers adjust their risk exposure to maximize their expected utility, an assumption congruent with a central assumption of neoclassic economic models.

    Multiple real-world examples of such scenarios exist. A first example is found in sports, where a team's probability of winning is a major factor in a team's profitability.  Economic analysis using play-by-play data and stochastic modeling suggests that in the American National Football League, teams' decisions between kicking and trying for a first down show a significant bias which undermines teams' chances of winning (Romer, 2006). A second example is found at the blackjack tables in Las Vegas, where individuals must decide whether to play aggressively or conservatively as hands are played over time. In a recent working paper Carlin and Robinson (2009) find strong evidence of a significant omission bias as blackjack players, even those playing for very large stakes, incur large losses by playing too conservatively. Further, after playing aggressively, individuals show a rebound effect, playing even more conservatively thereafter.

Click here to play a sequential risk exposure game.