Dynamic Modelling

System Dynamics uses computer simulation to achieve better understanding of social and corporate systems. It draws on organisational studies, behavioural decision theory and engineering to provide a theoretical and empirical base for structuring relationships in complex systems. We use it because it is a highly effective and rigorous method of analysing and understanding complex systems, especially social systems. It goes beyond the use of spreadsheets, process maps and other static analytical tools to provide a deep insight into behaviour and how it changes over time. It helps to identify the levers which will have the greatest effect on performance. Dynamics Modelling is based on a small number of key concepts:
  • The use of an icon based language. The user builds models using purpose built software to create a map of the stocks and flows in the problem area which is then populated with equations to describe the relationships between different model elements and data to describe the starting state of some of them.
  • Dynamic behaviour results from structure - the way that material and information flows around the organisational system. Organisation components are usually highly connected and interact accordingly. To change observed behaviour, you have to first understand and then alter the structure. Being based on structure and not data, models can cope with approximate, sparse and missing data and still provide useful insights.sample model
  • The effect of feedback loops which may be positive (self reinforcing) or negative (self correcting). Models may contain many loops of both types interconnected with time delays, non-linearities and accumulations. The dynamics of all systems result from the interactions of these loops. Results can influence causes and delays can lead to instability and oscillation. Small perturbations are often amplified by the feedback system, creating recognisable patterns such as start - stop motorway traffic or boom and bust cycles.
  • NonLinear behaviour means that effect is rarely proportional to its cause; small changes can have big impacts and vice versa.
  • Capturing the problem as it unfolds over time, explaining current behaviour in terms of past actions, decisions and learning which underpins the organisation’s ability to adapt to changing circumstances. The fundamental cause of a problem will often be separated from its effects by time, leading the unwary to treat recent symptoms while missing the disease.
  • The representation of soft and intangible variables such as human behaviour, a major factor when considering organisational performance, in combination with the more concrete system elements; and particularly the mathematical equations which describe the relationships between the different elements.
  • Simulation to improve understanding about what is actually happening, match the observed behaviour to historical data and to test drive hypotheses and possible solutions in a safe environment.
  • Group model creation: people are unlikely to make decisions for which they will held to account based on someone else’s view of reality and so the model creator needs to use group facilitation techniques to involve decision makers in creating the model, resulting in a sense of ownership and commitment with the model being the repository of the group’s thinking about the problem in focus. Click here to find out more about the facilitation of large Groups. At a certain level of maturity, models can be packaged as learning environments and used to communicate about the problem to a wide audience with confidence.
The Oil Producers’ Model by Morecroft et al demonstrates the use of equations to define relationships between model elements and it also shows how a model can be assembled and used so as to improve understanding of a market sector (in this case the oil industry) in order to help with decision making and strategy setting. To find out more, Morecroft’s book ‘Strategic Modelling and Business Dynamics’ is highly recommended.

Our Approach To Dynamic Modelling

Translating a multifaceted real life problem into a working, useful simulation model is not easy and requires a structured approach.

Legal | Privacy