Optimism bias refers to our tendency to expect better outcomes than what may happen. This can lead to underestimating project costs or timeframes, and overestimating benefits. To help avoid optimism bias in these areas you should use data, insights and lessons learnt from similar projects to get more grounded expected outcomes.
Go to the GOV.UK guidance for more information on optimism bias.
- Green Book supplementary guidance: optimism bias (UK Government)
Single-point probability analysis
Single-point probability analysis lets you work out a cost to accepting certain risks in your project. It is best used when you can estimate both the likelihood and consequence of the risk event well.
To work out an expected value for a significant risk, multiply the probability of the risk happening by the size of the consequence. The result provides the risk premium – the estimated cost of accepting the risk.
The disadvantage to this is that it does not account for variability in outcomes. This is especially true at the extremes when decision-makers may prefer not to accept the risk happening.
Quantitative risk analysis (QRA)
Quantitative risk analysis (QRA) is an evidence-based modelling technique. It gives risks numerical values based on quantifiable data, such as costs, logistics or completion time. This makes it easier to:
- assess the highest priority risks
- get a better understanding of the sources of risk to project outcomes
- get more accurate estimates of the likely costs or benefits
- highlight risks and their costs for stakeholders considering the business case.
A QRA approach is considered to be superior to an approach that relies on optimism bias or contingencies. You should use QRA as the first-best basis.
A list of approved risk analysis providers is available on the New Zealand Government Procurement website.
Monte Carlo analysis
Monte Carlo analysis is a form of QRA and is a technique used to understand the impact of risk and uncertainty based on a range of estimated values. The result approximates the full range of possible outcomes, and the likelihood of each.
The approach provides a systematic assessment of the combined effects of many sources of risk in key variables and can also allow for known correlations between these variables.
The Monte Carlo approach is more suited to investment proposals where:
- there are several key variables with significant uncertainties, correlated uncertainties, or both
- simpler approaches are not able to describe the variations in net benefits between proposal options well enough.
You may need expert advice to develop the Monte Carlo model and interpret the results.