Conference panels
Conference panels
Panel: The Parker Challenge Resuts
How much judgement and decision noise is there in resource estimation?
Have you ever wondered how many decisions go into a resource estimate? The models and estimates that underpin the minerals industry are constructions, based on multiple judgements, by people. Which means that our models and estimates are a reflection of our unique skills, knowledge, expertise and flaws. It is this aspect of resource estimation that the Parker Challenge is examining. How do the differences we bring to our professional practice affect the results we produce? How different can an estimate be when the only difference is the person responsible for the estimate?
Historically there has been a focus on improving estimation accuracy. New technology and algorithms, faster computers, increased automation. All aimed at making a better model and doing it faster. But the missing piece, is developing an understanding of how estimation precision varies between people.
Much of our perception of resource estimation risk is based on heuristics or studies where the estimation is completed by one person or one team. This approach restricts our view of the true risk. Cognitive biases such as anchoring and confirmation reduce our ability to see viable alternative models or to understand just how much of an impact a completely different perspective can bring.
And that’s a problem. How can we begin to understand and quantify resource risk and the implications for resource classification without understanding the full range of possible estimation outcomes? It’s like we have been hobbled before the race has begun.
The Parker Challenge is the first step towards understanding the impact of decisions and the interplay of mental models with resource models. One data set. Multiple estimates by anonymous and completely independent people. A range of different levels of education, experience and expertise. Where does the human judgement noise sit compared to other risks and what are the implications for Competency and Resource Classification?
What do you think the person-to-person precision of a Measured, Indicated or Inferred resource should be? And what is it in reality?
Let’s find out!