To: russwinter who wrote (11215 ) 5/20/2003 4:37:34 PM From: E. Charters Respond to of 39344 Boredom may hold but it will not run. Scheherazade and Circe occupied Isles of enchantment. Which brings me to a question which I am sure is not related, but may also be not related peripherally. What is de Chauvenet's criterion, and is it wrongly applied in circumstances where Krieging would degenerate to arithmetic averaging due to sparseness of data? I am not completely sure it can be answered without some dissatisfaction that the answer can be utilized routinely to advantage in a layman's analysis of results. Of course, we are referring obliquely to that "red-herring" of news announcements, the cut-off grade (lowest possible grade to include as part of a calculation in a mined block), and how it is arrived at with variations above and below mean more than say 2 SD's. However I feel the fundamental principle is often applied correctly from experience. If it could be systematized mathematically and geometrically geologically, it could form a substantially revelatory technique that could throw light on ore viability from reduced sample sets in certain environments. What we need to know is how the statistic or related method applies rigorously in practical environments. This may yield to massive data mining and correlation. The conclusion cannot so far be arrived at, that it can or cannot be systematised, no matter how simple or opaque the task at first appears. We know this from the number of unsupported allegations of the correctness of cutting levels, and the inability of countless operations to arrive at a cut-off grade at all even after much advanced work. That a working dynamic cut-off grade is not routinely established in reports from hundreds of mines, demonstrates that the method is neither trivial, well established nor agreed upon. It is only partly this because of lack of knowledge of costs, as different costs can be assumed and dynamically assigned, otherwise why drill any rock at all with intent to mine? How important could it be? Well it could double or triple grade, and it could do the same to reserve or potential ore quantities. At least we are sure proper application of selection criteria and averaging methods would make a significant difference to ore estimations. The lack of ability to discern where and how to apply data acceptibility criteria may or may not prejudice operations, but it has to reduce somewhere the recovery of ore if improperly applied. Badly applied versions of data selection and averaging routines just cannot be totally unimportant where it is applied to data out of any seeming necessity. Did it affect Kerr Addison's bottom line? It did the Salmita's. The decision to "overrule" at least a certain once thought-correct applications of data averaging, contributed 100% to that mine's existence. Eldorado (U3O8) would NOT publish or even say what their cut-off grade was, but they would declare their average grade! Cut off was a trade secret. (1980) I am not sure I know why. Revealing it may have revealed something about not published recoveries of certain elements, or the fundamental economics of the mine pointing to fact that were not ordinarily divulged. On the subject of not published recoveries, we know that Inco has never publised the breakdown of PGE metals in their concentrate since 1926 by decision. They did ship the sponge overseas. Taxes, or security? EC<:-}