Tag Archives: reference class forecasting

Another reflection of what it is to be Risk Savvy, in the context of RCF

This blog is a first look at the psychological aspects of Reference Class Forecasting and how this relates to Project Management. I link this blog to several papers and contemporary academic debates that sit central to the direction project management betterment is being directed toward. These initial source flags simply highlight the contemporary nature of current debate which in some quarters may be represented as definitive truth.

This is prompted by a line in Gerd Gigerenzer’s 2014 book Risk Savvy, and a passing comment I am yet to better source. This suggestion that his perspectives differ significantly from Daniel Kahneman and Amos Tversky. Given the central theme Kahneman and Tversky play in the papers introducing Reference Class Forecasting to Project Management, these two perspectives may guide my own research better in whether one perspective can inform or must necessarily dispute the other.

Project Management and reference class forecasting – RCF

Whilst explaining some rudimentary mistakes in representing risk, Gigerenzer states the following, “left on their own people intuitively fill in a reference class to make sense for them” (pp3).

From a Project Management perspective the contemporary discussion on cost estimating is often framed around the concept of “reference class forecasting”. The Infrastructure and Projects Association (IPA) advocate this approach {click here and refer to slide 28}. Oxford Said have supported RCF and developed it into a meaningful betterment of government estimates of project cost, examples here are projects in Scotland and Hong Kong. RCF also has 21st Century and mainstream backing in psychology.

However, government advise has not been ubiquitous in its support. Note the reference here to a paper presented to a House of Commons select committee enquiry in 2019, sourced from the open records of an equivalent representative body in Newfoundland, Canada during the recent Muskrat Falls enquiry.

I remain undecided either way. I have had the privelege of attending several lectures by the Oxford Said Business School. One specifically outlined how RCF is being applied. The Gigerenzer perspective, and the RCF counter-narratives flagged here, present reason to keep asking what it is that drives our decisions. Is RCF sufficiently robust to enable defensive decision-making to be countered? Or are these two accounts compatible? Particularly if this reflects separate sets of variables and influences beyond optimism bias.

In this regard I see Gigerenzer presenting different dynamics to those of Daniel Kahneman and Amos Tversky, and the entire set of risks I believe RCF are intended to address. Both may therefore be correct, but neither complete. I will perhaps understand this better once a more complete review of the literature is undertaken.

v | b | t

Per my last blog, it is the Gigerenzer case that seems more compatible with what I am leading with, as possible root-cause. I am of the view that many of our project failings are not directly resulting from the estimates of cost, but more the divided motivations of employer and contractor that thereafter emerge. The human behaviour element, being the unaccounted for reality of colloquial decision-making motivations. This is my reason to think the Gigerenzer view to be at least as valid as the estimating bias being countered by RCF.