Engineers use probability to consider different risks when we are preparing a project. But probability is not a measure of risk. There is an interesting debate in Linkedin about this issue: We can consider that risk is a subjective perception, not as a property of a system. The probability of default is an indicator of the risk, and an attempt to put a subjective thing into a mathematical model. Fukushima’s catastrophe could give us an example of this point. Surely, the probability of a big tsunami (one every x years) is the same after the last event, but the perception of the danger of nuclear energy is higher. Considering risk as a subjective perception, we could think that nuclear risk is higher today, although the probability of an earthquake (mathematical risk) is the same.
As engineers know, statistics only lets us to assure that the mean of the benefits will be over the mean of the faults when the number of events is large, but it cannot assure when certain event can arise. This assumption is very useful to run an insurance business, for instance, but could not be sufficient to guarantee the control of a hazardous system when an unexpected event arises. Then, we should use additional variables to improve the safety of a dangerous facility.
Unlike risk, complexity is a property of a system and like other properties of a system can be measured. When an unexpected event arises complexity increases. Then, complexity could give us a pre-alarm of a fault in our system, and can be used to improve the safety.
Additionally, Fukushima’s Tsunami will produce disturbances in the world economy because of the well-known butterfly effect, but measuring the complexity of our business can help us to analyze if it is being affected by this effect before we cannot control it.