Uncertainty is the state where a decision-maker cannot accurately or precisely) predict the outcome of an event. Uncertainty is categorized into aleatory (random) and epistemic (probabilistic) processes. The first type is due to variability, which is an intrinsic property of natural phenomena or processes. The variability cannot be reduced unless the phenomena or process is changed. This variability cannot be reduced by collecting more data. Only by changing the process can the variability be reduced. Epistemic uncertainty is due to lack of knowledge.
Epistemic uncertainty is reducible, through the collection of data or acquiring knowledge. [1]
In order to manage in the presence of uncertainty effectively, the decision maker needs to know the possible outcomes of uncertain events and processes, and assess the likelihood of these outcomes. Theories of uncertainty, including probability theory, evidence theory, and imprecise probability are available for quantifying uncertainty.
In all cases, the assessment of these uncertainties and their impacts on the decision-makers ability to make a risk-informed decision requires estimates to be made.
No risk-informed decision can be made in the presence of uncertainty without estimating.
#NoEstimates advocates conjectures you can.
No matter how many times, #NoEstimates claim this can be done, it is simply not possible in any principle-based sense to make a credible decision in the presence of uncertainty without estimating.
[1] Design Decisions under Uncertainty with Limited Information, Efstratios Nikolaidis and Zissimos Mounrelatos, CRC Press, 18 Feb 2011.
