Model III:
Description: This model is an unsuccessful attempt at developing a useful adaptive version of the previous model. Two algorithms were tested, one in which the threshold was modified after each parking event based on the normalized ratio of regret to time spent driving. This model always moved quickly to a uniform threshold value of 0, notable as the worst possible threshold from a utility maximization perspective. The second model was an attempt to use a cognitively simple update procedure based on the same two variables such that if the regret was non zero the threshold was updated to be 10% more risk seeking. If the driving time exceeded the maximum distance of the parking row the threshold was updated to be 10% less risk seeking. This was normalized primitively by disallowing a threshold that exceeds 1. This dynamic drives the threshold for all agents to 1; another non-optimal result.
Space: Linear, same as Models I and II
Time: Synchronous/Asynchronous: same as Models I and II
Agents: Heterogeneous variation based on threshold.
Decision Structure: See Figure 1.
Parameters tested for robustness:
Criteria for judging outcomes:
Model
Outcomes:
No outcomes are reported for this model as it is believed to be incomplete and only partially functional.