Simple Risk Measurement¶

Simple Risk Measurement is written to help you measure complicated risks using a process that’s simple enough to work out on the back of a napkin and powerful enough to organize a rocket launch.

If you are an engineer motivated by the reduction of risk and are frustrated by how to measure your progress, you may find this documentation useful. Simple Risk Measurement can get you started towards a comprehensive and scientific approach to risk. It is designed to enhance subject matter experts who work with risk, especially those who mitigate complex risks on an ongoing basis.

See Risk to better understand this problem in engineering.

How does it work?¶

This approach is very simple to trivialize into a few steps.

1. Define a risk scenario: A well defined, undesirable future event that we want to measure.
2. Gather the evidence: We gather facts, reference classes, models, experiences, and opinions.
3. Estimate the outcomes: We estimate of the probabilities / impacts the scenario may have.
4. Make a decision: We make decisions with this information, and measure again if needed.
5. Keep Score: As outcomes occur, we check our work, improve our methods and reduce our uncertainty.

For example:

The Cubs win the 2025 World Series may be a future scenario you care about.

To better understand this scenario, you might gather evidence by reviewing historical baseball data, interviewing experts, or building statistical models. You would then forecast likelihood of outcomes like Yes or No, or estimate other values based on that information, and make decisions as a result.

With little effort, for example and entertainment purposes, this author prepares a 2.5% Yes and 97.5% No forecast. This starts from an assumption of 1/30 teams in the MLB (3%). The author observes that other teams more often win the World Series, a World Series might not happen at all, and the author holds an irrational and indefensible belief in the Curse of the Billy Goat subjectively reducing this forecast further towards 2.5%.

With greater evidence, time, and additional expert opinions, this documentation suggests methods to make this forecast increasingly rigorous, defending against cognitive bias (The author believes in curses? Really?) and introducing values that represent other knowledge. This can result in a more reliable value than an expert opinion that qualitatively states “They probably won’t win in 2020” which can not be leveraged by scientific measurement or as an engineering approach.

How does this help me?¶

You may be enthusiastic about some of the following risk problems:

Is a risk worth mitigating?
We can gather quantitative data that informs decisions to invest in a mitigation, (or not).
Has a specific risk gotten better or worse?
We can measure a risk over time and observe differences if it is being mitigated (or not).
Which mitigation option(s) should we choose?
We can measure many mitigation options and quantitatively compare their impact on undesired outcomes.
How do we measure many risks?
We can organize many risks that are being measured and use them to inform broader risks.
Have we spent enough time identifying new risks?
We can measure if further risk assessments are likely to reveal new risks, (or not).

What doesn’t this do?¶

This method is industry agnostic. This creates very helpful limitations that, hopefully, improves its accessibility.

This documentation does not prescribe:

1. The governance of your risks, or accountable parties.
2. The scenarios you should define for your risks.
3. The approaches you should to mitigate your risks.

By avoiding these subjects, this method hopes to measure any creative approaches to risk mitigation.