Review of "Making Good Decisions Without Predictions"

Making Good Decisions Without Predictions: Robust Decision Making for Planning Under Deep Uncertainty

Under the condition of unprecedented, transformative, and surprising change, the quantitative models and methods may become counterproductive or insufficient.

Predictions are not always true.

Robust Decision Making (RDM), informs good decisions without requiring confidence in and agreement on predictions. The basic idea of RDM is that, instead of reliying upon models and data, hundreds to thousand of scenarios are generated using models to describe how plans prforms in these plausible futures. Then, future conditions in which the plans perform well is distinguished using visualizationa and statistical analysis.

The traditional predict-then-act framework may leads to gridlock or overconfidence, when the predictions are in accurate or controversial.

Analytics - the discovery and communication of meaningful patterns in quantitative information.

Deep uncertainty - when the parties to a decision do not know or agree on the best model for relating actions to consequences or the model to represent the future or the likelihood of the future events.

RDM put forward the backward analysis, that is beginning with a proposed decision or plan and testing the plan against many different plausible futures.

RDM combines two approaches, scenrios and probablistic analysis. The steps involved in the RDM analysis are, decision structuring, case generation, scenario discovery, and then trad-off analysis.

Robsutness and flexibility can provide the best response to deeply uncertain future conditions.