The Game of Software Process Improvement: Some Reflections on Players, Strategies and Payoff

When starting the software process improvement (SPI) journey, there are many SPI standards to select from. Selecting good SPI standards can be a technical problem, from a software engineering point of view, but it can also be a political problem, some standards fitting more with internal political agendas than others. As it is well-known that SPI without management commitment can have disastrous effects on SPI, so can also be the consequence of selecting standards that are technically unfit. The dilemma on how to select SPI standards provides a picture of SPI as a political game played out between managers, software engineers and SPI people. Starting with SPI from the viewpoint of control theory, the paper identifies different conflict situations within the control theory framework, and suggests using game theory and drama theory for finding optimal control strategies. Drama theory is further explored through a SPI case study that illustrates how SPI standards stabilize in spite of conflicts and social disaster. The contribution of the paper consists of introducing the concept of ‘evolutionary drama theory’ (derived from evolutionary game theory, EGT) as a tool for describing and analysing how an artefact like a SPI standard evolves towards equilibrium (evolutionary stable strategy, ESS) by looking at repeated dramas where equilibriums may not necessarily be found or, if found, may not necessarily fit with the ESS.