Reconstructing a forecast: what hindsight can distort
In Zurich, decision reviews are a regular feature of work in financial services, insurance, tech and the European headquarters of international groups. The managerial question is specific: how can a team learn from an outcome without reconstructing what it previously thought it knew in light of that outcome?
What the literature establishes
Baruch Fischhoff (1975, Journal of Experimental Psychology: Human Perception and Performance) demonstrated experimentally that knowing an event's outcome changes retrospective estimates of its probability. Participants told the outcome judge it more likely than those who do not know it, without fully recognising the influence of this information on their judgement. A forecast reconstructed after the event is therefore not a neutral substitute for one recorded beforehand.
Why the past becomes easier to read
In their review, Neal J. Roese and Kathleen D. Vohs (2012, Perspectives on Psychological Science) distinguish between memory distortion, perceptions of inevitability and perceptions of foreseeability. Hindsight bias is therefore not simply a memory failure: the search for a coherent explanation can also make an outcome appear more inevitable than it previously seemed. In a project review, explaining why an outcome occurred does not demonstrate that it was predictable.
The premature conclusion
It would be tempting to conclude that retrospective reviews are useless or that every explanation offered after an event is suspect. These studies instead suggest separating two tasks: understanding the mechanisms behind an event and establishing what could have been anticipated with the information available. A new explanation can be correct without having been accessible when the decision was made. (our executive and employee training programmes)
What the evidence does not establish
The classic experiment examines judgements in controlled settings; it does not directly measure the quality of project reviews in Zurich businesses. The review describes several manifestations of the bias without establishing that a documentation procedure would eliminate them in every organisation. A written forecast provides a reference point, but its existence guarantees neither its initial quality nor effective learning.
A practical check in Zurich
As part of SHR's decision-making programme, a Zurich team could test a protocol across a defined series of comparable decisions, such as service launches or IT migrations. Before each decision, it would record a verifiable outcome, its time horizon and its estimated probability; during the review, once the outcome was known but before reopening the record, each participant would reconstruct the team's original estimate. The measure would be the difference, in percentage points, between each reconstructed estimate and the archived value, distinguishing events that occurred from those that did not. This record would make possible retrospective reconstruction visible, without being presented as proof that the protocol already improves decisions. To go further: explore the Decision-making training in Zurich in the canton of Zurich, or browse our executive and employee training programmes in Switzerland.
In pictures: Decision-making in Zurich
- decision-making
- Zurich
- research
- training zurich