Once upon a time Dr Margaret Armstrong was entrusted with the review of opinions deemed at variance with Matheron’s new science of geostatistics. She was keen on talking about Matheron’s gift to mankind. She did so in De Geostatisticis! In N°14 she was beating around the bush about Freedom of Speech? From time to time she would dream up another and mail it to subscribers. I didn’t mind at all to be on her mailing list! On the contrary, it was a convenient way to keep track of thinkers who were into praising Matheron’s new science of geostatistics without thinking about counting degrees of freedom!
Let me explain what Marechal and Serra talked about at a Colloquium on geostatistics at The University of Kansas in 1970. They showed how to strip the variance off the distance-weighted average and call what was left a kriged estimate. Infinite sets of kriged estimates and zero kriging variances morphed into the heart and soul of Matheron’s novel science of geostatistics. Thank goodness, the Editor of De Geostatisticis didn’t really call it Matheron’s gift to mankind. On the contrary, it made me think of Mark Twain and his postulate that a mine is a hole in the ground with a liar on top. Nowadays it is still a hole in the ground but the liar on top is stringing Markov chains. Mining investors should be aware that stringing Markov chains does not give unbiased confidence limits for metal contents and grades of mineral inventories. The TSE and the OSC have not yet taken a close look at stochastic mine planning with Markov chains.
Margaret Armstrong had been awarded a Master’s degree in mathematical statistics at the University of Queensland. After migrating to France she was awarded a doctorate in geostatistics. What she does is called “stochastic modeling in the earth sciences”. Professor Georges Matheron had flunked his PhD in 1965 simply because didn’t know how to test for spatial dependence by applying Fisher’s F-test. Yet, it was Dr F P Agterberg who praised him as the Creator of Spatial Statistics upon his passing in 2000. It made me wonder when, where and why Professor Dr A G Journel had been awarded PhD. The more so since he only teaches those who do have a PhD in geostatistics!
Following is ad verbatim the text of Armstrong’s review of Precision Estimates for Ore Reserves. It was put together by Merks and Merks and praised by and published in Erzmetall. I’m proud that my son was the coauthor. He has a PhD in Computing Science, and was awarded the Graduate Dean’s Medal from the Faculty of Applied Science in 1986 and 1992. His interests include programming language design, software engineering and parallel algorithms.
Here’s what Dr Margaret Armstrong wrote about our approved and published paper:
“The text presents no equation, no development except for the trivial relative variance of a product of independent factors (p. 3-4). Variance of errors are mistaken for spatial variance, F-tests are called for forgetting the condition of independent samples, kriging variances are said to “violate the requirement of independence” when they were developed precisely to handle the case of (spatial) dependence, the central limit theorem is invoked for the trivial σ²/n variance of estimation of the mean of n independent sample values, the list could go on and on….. Definitive statements such as “Only the variances for ordered or randomized sets of measured grades generate reliable and realistic precision estimates….” are put without even a definition of what an ordered set is, less any proof. The individual has some understanding of sampling of particulate materials, and believes he can transfer such notions, as is, to problems involving spatial distributions.”