For consequences the numbers are straightforward measures of the impact, such as the financial loss or the number of people killed in an accident.
For likelihood the number is usually a probability.
A probabilistic risk model will comprise a set of inputs and one or more outputs which are determined by these inputs. If a probability distribution is applied to each of the inputs the probability distribution of the output can be calculated. The main problem in doing this is to account for interrelationships between the inputs. This means the input quantities are not independent and the correlation between them must be modelled.
One approximate way to calculate risk models is called Monte Carlo. This involves exploring the possible futures a large number of times. Each version unfolds as a result of selecting randomly from each of the input distributions. The Monte Carlo method is simple, powerful and flexible. The Risk Agenda has its own risk modelling tool called Natural Monte Carlo. You can download it from our software page.