Sunday, 11 October 2015

Resilience as concurrent interplays of opponents: preliminary ideas and call for collaborations

In my previous work “A behavioural model for the discussion of resilience, elasticity, and antifragility” I discussed resilience as the emerging result of the dynamic interplay between the behaviors exercised by a system and those of the environment it is set to operate in. Then, in my latest work “On ambients as systemic exoskeletons: Crosscutting optimizers and antifragility enablers” I introduced a “horizontal” and a “vertical classification” of systems. The idea in this second paper is that any system may be considered at the same time a “whole”, a set of “parts”, and a nested compositional hierarchy of functional nodes.

By applying the idea of the first paper to this second idea of a system’s horizontal and vertical classifications, I thought of resilience / antifragility as the emergent result of concurrent behavioral interplays taking place across the involved systems’ [ wholes / sets of parts / hierarchies of nodes ].

Let me clarify this through two examples:

  1. A bullet is shot and passes through the body of a living being. Such a traumatic event directly affects a number of organs and systems of that being. Interdependence among organs and systems is likely to lead to cascading effects that may in turn lead to severe injuries or the loss of life.
  2. A hurricane hitting a region. Catastrophic events such as this one typically ripple across the involved hierarchies of nodes triggering the concurrent reactions of multiple crisis management organizations.

In both cases, resilience may be modeled as the result of the effects of an external event on a system’s horizontal and vertical organization. The external event manifests itself at all systems and networks levels and activates a response that is both individual and social.

Now, I believe a good way to model such response could be in terms of Game Theory. The idea is to consider:

  • GT players as the involved whole / parts / nodes (let me call them “opponents”).
  • Behavior classes as the strategic choices available to the opponents.
  • GT strategies as the behaviors planned and enacted by the opponents.
  • Moreover, I propose to assign energy budgets specific of the involved opponents. Said energy budgets serve as global constraints shared by all of the nodes of the GT players across their horizontal and vertical organizations.

  • Finally, I propose to associate GT payoffs to the possible behavioral responses, with costs (in terms of consumed energy budget resources) proportional to the complexity of the chosen behavior.
      (↦ A simple purposeful behavior only requires sensory / actuator parts, thus calls for less energy than a complex proactive behavior, which requires analytical / planning parts; in turn, that behavior would cost less energy than a complex antifragile behavior, which requires in addition complex knowledge/wisdom management parts.)

The idea I would like to develop is that to define nested compositional hierarchies of payoff matrices: sort of interconnected and mutually influencing payoff “spreadsheets”. Overall resilience may be expressed, maybe in some analytical form, as the solution of those payoff spreadsheets taking into account the outcomes of all the concurrent interplays and the consequences of those outcomes on physically or logically neighboring parts / nodes.

      (↦ Think once more of the case of the bullet passing through a living body and of the cascading effects of the many concurrent “confrontations” between the involved organs and the bullet.)

The payoff spreadsheets would also model the local and global costs in terms of the shared energy budgets. Adopting the most expensive behaviors at all levels could result in too high an energy consumption, while being able to “keep control”, and proportion one’s response hierarchies, may translate in the ability to survive longer.  Two very eloquent examples of this fact were provided by Dr. Jeff Washburn here).

I would be happy to discuss the basic ideas and possible “implementations” of the just sketched model. In particular I would welcome discussing with experts in game theory and evolutionary game theory willing to co-write a paper on the above idea for ANTIFRAGILE’16.