Cascade risk
A failure to imagine what can go wrong
may lead to catastrophe
"
"
Through our work we have come to realise that risk is:
Difficult to do right so that is a rewarding, value-add activity
Abstract and hard to understand
Boring and painful with complicated and slow assessments
Subject to common biases
Difficult to apply at scale to manage hidden hazards
This is due to some common pitfalls:
An unclear context as risk is abstract
(the risk to what?)
Not recognizing that risk is a creative endeavor
(and not creatively identifying hidden hazards)
We perceive risks to be linear
(but are often a power law)
A cascade of multiple hazards are unpredictable
(but may lead to catastrophe)
Biases
(we are biased, knowledge is biased, experience brings more bias)
Probabilities are generally counter-intuitive to human nature
(we are often fooled by randomness...and uncertainty)
Difficulty in getting knowledge and experience back into the risk management system
Leveraging biased knowledge can cause more harm than good
(particularly probabilistic knowledge that gets misinterpreted and causes blowups)
A reluctance to acknowledge that hidden hazards may remain
(and that risk controls require ongoing management)
Many risk tools and it is unclear when we need to do formal analysis
(and when more informal risk-based decision making is enough)
Inability to provide a helicopter view of risks to people in charge
(and they make decisions without understanding them)
Not providing a cohesive, repeatable framework across an organisation
(and the exceptions mean people do what they want)
Not understanding that the result of a risk assessment becomes less relevant over time (environments are changing, it never covered all of the risks, may have been biased and there is new knowledge)
Not encouraging open reporting to generate real-world data about detectable risks
(near misses are free, but catastrophes are not)
Not letting go of old risk assessments
(Leveraging old imagination will not reveal new hidden hazards)
Trying to quantify risks from subjective scores
Through the cascade commitments we strive to:
1.
Remember that a failure to imagine what can go wrong may result in catastrophe.
2.
Prevent catastrophes to people, planet & possessions, in that order.
3.
Avoid being abstract, make the aspect and objective clear.
4.
Recognise that a cascade of events that are complex and unpredictable can be catastrophic, that impact is non-linear, and instead use non-linearities to our advantage.
5.
Recognise also the non-intuitive nature of probabilities that cause us to be fooled by numbers and randomness.
6.
Be diligent against biases that limit group collaboration, reduce imagination and generate counter-intuitive aspects of risk, randomness and probabilities.
8.
Never stop as our efforts are always incomplete to identify unknown unknowns in hazards and knowledge. Leverage the creativity of the right mix of people and experience, and quickly respond to changes and discard the obsolete as new knowledge is generated and hidden hazards are revealed, as all efforts are point in time under uncertainty.
9.
Provide guidance on risk controls through signals monitoring and reporting to ensure problems are danger signals for something is wrong not a prediction that everything is fine. Not to blame but to leverage human variability as an asset to leverage 'near misses' as free lessons for early detection.
10.
Be vigilant to prevent unbounded downsides by using non-linearities to our advantage and to seek non-linear 'immunizations' in uncertainty through tinkering
11.
Be effective through recursive (fractal) loops to keep effort commensurate with risk and prevent catastrophes at all levels of complex systems. A complex system will have a different context at different levels of detail with a different audience, experience and knowledge required. Our endeavours need to ensure we are value-add, through lean iteration and scalability.
12.
Embrace uncertainty to build resilience. Favor experimentation over stories, but storytelling can be useful to convey the right message. Leverage stories to immunize others against catastrophe, uncertainty and our own biases. Stories have the power to distort the facts, so stories backed by empirical evidence can be independently checked and encourage the reader to come to their own conclusion.