Journal on Policy and Complex Systems
Table 4 . Average aggregate K per type in three homogeneous constant environments
Result 1 : Flexibility Threshold Effects
It is not surprising that flexibility is not always useful , but it is not obvious when
it is and is not . The results of this model demonstrate two specific reasons why flexibility may not always be good , and it points to two circumstances under which this is so . The first reason flexibility may not always be useful has to do directly with utility , and the second reason is indirect — it can lead to other outcomes in a system that may undesirable , including inequality and loss of diversity .
In terms of the direct effects of flexibility on utility , we consider the role of the flexibility threshold ( θ ). It turns out that at high levels of ( θ ) combined with r > 0.5 , long periods of switching between types means that agents do not stay at a particular type long enough to earn points . That is , it takes a few periods of clustering agents attempting to cluster before they get close enough to one another to really earn high scores ( generally in the 0.8 – 1 range per time step ). Similarly , spreading type agents also need a few time steps to move around before they have spread themselves out to a point where their utility could grow over time . In other words , if the agents are too “ picky ” and they switch back and forth all the time , the entire period during which they are switching they aren ’ t earning as many points as they might if they would just pick one type and stay there . What ’ s more , since these agents have high thresholds , it takes them that many more time steps of switching in order to reach those high thresholds , after which they finally converge on a consistent distribution of types .
With respect to indirect effects , flexibility can mean that agents switch type frequently early on in a run , and then are locked into a certain ratio of types once the satisfying threshold θ is triggered . Interestingly , this is robust to any threshold . The agents which do well early on , and thus trigger their “ stay at this type ” threshold earlier than other agents end up losing out tremendously if the system ends up being one that favors agents of one type over another . This is a kind of unexpected first succeeder disadvantage and has some real world examples in industries where being the pioneer is not necessarily the most advantageous position in the end ( e . g ., social media examples where Friendster and MySpace were successful early on but then could not adapt when Facebook joined the mix later ). In addition , the lower the threshold , the more agents there are who can capitalize on this new “ knowledge ” of which strategy is best , which exacerbates inequality in number ,
128