Cultivating superior forecasting
In recent years, work by various groups including the Good Judgement Project has established some important new results about forecasting by experts in difficult domains such as geopolitics and financial markets. Such forecasting generally is done very badly, but contrary to what hardened skeptics assert, we now know it is possible to develop skill; that some people can perform remarkably well; and certain factors contribute to high performance. As part of work for a large Australian superannuation fund, Tim has distilled recent research into seven guidelines for the development of a "superforecasting" culture within an organisation. Of particular note is the use of scoring rules for evaluating forecasts, which can provide greater insight into how well individuals or groups perform, and whether good outcomes are due to skill or luck.
Was the head of MI5 a Soviet mole? Mapping the case
On 10 April a workshop will be held by the Institute for World Politics in in Washington DC on the longstanding question as to whether Roger Hollis, Director General of MI5 from 1956 to 1965, was a Soviet mole like Kim Philby. The workshop will be based on a report prepared by Paul last year, which powerfully demonstrated how using argument mapping can display a complex set of arguments with exceptional clarity and rigour. Unable to travel to Washington to lead the workshop, Paul prepared a set of videos outlining the case made in the report, the reasons why it matters and the purpose of the forum.
Paul Monk in the first of four videos laying out the case against Roger Hollis
What do the experts think? Combining elicitation and rationale capture
Even in this age of "big data," there are important problems for which our only option is to ask experts for their best guesses. How exactly should such guesses be obtained and combined - and defended? This is the challenge we confronted in work on the estimation of future aircraft losses. Working with scientists from DSTO and the Centre for Excellence in Biological Risk Assessment, we developed a new approach combining standard expert elicitation techniques, which focus on numerical estimates, with visual mapping of the rationales (causal explanations and evidence) for those estimates. The approach may have application to many other challenges of expert judgement.
Tip of the Day - be more Anti-Social Why the modern world is bad for your brain is a compelling account of how certain aspects of modern life impair our thinking and decision making abilities. A key lesson is that one of the easiest ways to improve thinking is to improve the conditions under which we try to think - in particular by removing distractions. In this regard Tim confesses to using Anti-Social, which can help you selectively restrict the internet's peculiar ability to degrade focus and clarity.