The Parker Challenge is Open

https://ausimm.eventsair.com/13409-mineral-resource-estimation-conference-2023/parkerchallenge/Site/Register#msdynttrid=susk_tJOUZaX_Oc1FDERqm3Slvt074E0wDHg0lVFV20

As scientists and engineers we like to think we rely on hard data and facts. We like to think that our analysis is objective and the conclusions we draw from our data are both accurate and precise. We also like to think that someone else, presented with the same data, would come to the same conclusion, or at least approximately the same.

But simply because we like to think something does not make it true!

What if I told you there was a hidden factor lying behind our industry??A factor that is not discussed, not reported and rarely tested? What If I told you that this factor may well play a pivotal role in he success or failure of some projects, it may affect how we perceive resource projects and the risk of those projects?

What if I told you that resource risk is a function of this hidden factor??What if I told you that resource classification was a function of this hidden factor?

It is all true. I’m not trying to be sensational or to trick you. There is something we simply don’t know about resource estimation. It’s incredible to think that in 2022, more than 30 years after the first release of the JORC Code and more than 60 years after the so-called Poseidon nickel boom and bust, we have never sought to investigate how much our own judgements affect the classification and reporting of mineral resources.

No, we simply assume that we are all similar and that we all would produce approximately the same estimate from the same starting point. How naive.

Let’s think about this objectively for a moment.

While the population of Competent Persons is relatively small it is nonetheless a large number of people. Our ages and experience are wide varied. We have learned our trades in different ways; from the self-taught to the formally educated. We have learned in different countries, commodities and by working in different organisations with sometimes unrecognisably different cultures and approaches to risk.

And we all have different personal risk appetites.

How can we possibly think these individual differences are all simply set aside when we are estimating and classifying mineral resources? ?

That is almost exactly the premise taken by the JORC Code and other reporting codes. It’s not their fault and it’s not a surprise. We are all blind to the impact of our individual differences. We simply do not know how much those differences affect things.

Why? Because we are blind to the amount of human judgement involved in creating these magical estimates. We forget that our expertise is used to judge:

?- the data, the amount, quality and location of samples and other data.

?- the way we group or domain the data into ‘like’ populations so we can make statistical inferences from that data.

?- the parameters we decide to use during estimation

?- even the estimation method itself is a judgement!

We also use judgement when we are classifying estimates. Is my understanding of ‘Measured’ the same as yours? The various Codes are not prescriptive. It’s an expert opinion.?

Let me quote some examples from outside of the resource industry. These are all referenced in Kahneman et al., book “Noise” (pub. Williams Collins 2021).?

Example 1. The legal system. As early as the 1970’s Marvin Frankel, a Lawyer and Judge pointed to the differences in sentences given for similar crimes. Two men, found guilty of small (<$60) crimes were sentenced to 30 DAYS and 15 YEARS respectively. Their sentence was a consequence of their judge not of their crime.

Example 2. The legal system. A study of 6 million decisions by French judges over 12 years showed that defendants were given more leniency when sentenced on their birthday.?A review of 207,000 immigration cases over 4 years identified a correlation between the temperature and the likelihood of being granted asylum.

Example 3. The insurance industry. A study by Kahneman et al found that the median cost of an insurance policy could vary by more than 55% depending entirely on the person doing the underwriting. Look at that number again… 55% variance in the MEDIAN. More than half the differences were above 55%. And let’s not forget underwriters are experts in the field of judging risk!

Example 4. What about a more scientific endeavour? Starting in the mid-early 2000’s Itiel Dror, a cognitive neuroscience researcher from University College London started investigating the variability of fingerprint identification. Yes, fingerprints. Those ‘infallible’ marks of identity. Guess what? It turns out they may not be quite as infallible as we like to believe. Dror’s research involved recruiting a group of fingerprint experts who agreed to examine the same fingerprint case (in a blind test) at some random time in a 5 year period. The same person, looking at the same fingerprint… how different could their judgements be? Just for a small twist the experts were given contextual information (a standard practice). Four out of five changed their opinion… yup 80% change. Their judgement was affected by unrecognised cognitive biases. If they were told “the police think this is their guy” they were more likely to agree that the fingerprints matched. Told “we don’t think this is the culprit” they were more likely to say the prints were different.

WOW

So here we are, Competent Persons one and all or someone who relies on the judgment of a Competent Person and we are entirely ignorant of our own biases and their impact. What’s worse, we know we all have slightly differing views but nonetheless we think that calling something “Measured” or “Indicated” is entirely reasonable - without understanding how our individual difference might impact on those judgements.

Well folks, now it’s time to start trying to understand this person-to-person variability in resource estimation. Now is the time for the Parker Challenge.

Give the same data, the same information to anyone willing to participate. See how different the results can be. Just how much variability do we create through our own judgements and biases? What’s more, there’s a chance to be rewarded for your efforts - a $55,000 prize sponsored by Rio Tinto. Not bad, contribute to the science of uncertainty and maybe walk away with a handy little bonus!

So, follow the link, enter the competition. I encourage you all to get involved. As someone on the organising committee I am looking for nothing more than for everyone to make my life very very busy in a few months when we are assessing the entries.?

Let's start to peel back some of the unknown unknowns. Help us shed some light on how our differences affect the quality of resource estimation!