FWIW To put the Rover issue into a real world perspective here are three associated issues from exploration and mining with a view to introduce some to concepts that some who may be unfamiliar with them and for review, additions, corrections by those who are. Apologies if my sense of humour is warped and a little on the nose
Disclaimer: I am not a resource geo and the following is general info on a scatalogical perspective of digging holes in the ground - DYOR
Cutoff grades: Using the example Klaus mentions in his recent interview (http://hotcopper.com.au/posts/18453118/single) of the 15m at 250 g/t Au this would be termed an uncut assay.
To arrive at a cutoff grade accurately requires a statistical analysis of a relatively large number of samples to be meaningful. IMO this is not possible for DM atm.
Lacking this rigorous method for a cutoff grade a simple expedient, commonly used, is to choose 30 or 35 g/t (Roughly one ounce).
Thus it is not uncommon to see announcement where intersections are reported as both uncut and cut (using XXg/t)
e.g. for the example Klaus gave it could appear like this (if my math is correct) 15m @ 255.6g/t UNCUT; 15M @ 10.6g/t CUT (35g/t). Or as reported 24/6/16:
All are correct. I would expect individual 1m samples to be assayed and reported which will give a variation of the above.
How results such are these are perceived by the market can be (is) quite different to how resource geos do. They take these numbers (plus those from another 10Km, 50Km, 100Km etc of drilling (among other things)) and see if they can be turned into a mineable resource.
With the advent of powerful computers it is now common to see such a resource quoted at multiple cutoff grades. These cutoff grades, while similar to the individual intersection example above, take the process a stage further and consider the entire deposit. The final cutoff grade for a mine is a complex decision - way beyond me and not necessary atm.
Inherent in deciding what cut off to use is that it is an artificial and statistical tool. This leads to a very important aspect of mining Reconciliation.
Put very simply:
For a period of time the mine geos estimate how much gold is going to be mined from X amount of tons.
After mining the actual gold can be weighed and compared to the estimate.
Negative reconciliation = Bendigo Mining version 1 ?2005 - suggests (among other possibilities) poor grade control and or a highly variable gold distribution (High coefficient of variation).
Neutral - a satisfactory result which suggests a fairly uniform deposit or very good grade control data or both.
Positive reconciliation - (Some examples: Sunrise Dam, Troy at Sandstone, ?Telfer) Happy faces all round but similar reasons as suggested for a negative recon.
In the case of positive reconciliation Rover has been very generous and the statistical distribution of doings is very real. However, it is dangerous to assume the level of positive reconciliation and factor it into a resource calculation.
An example, that needs more research as it is just a vague memory, is where such an assumption was made for Telfer mark 2 after it had been shut down for a couple of years. Apparently the re-opening didn't work out as well as planned and this was one of the reasons given.
The lack of reconciliation is related to the third aspect: Coefficient of variation (CV).
It tries to put a number on Rover's back yard activities.
Evenly spaced around the yard would give a low CV (less than ?5) and would likely give a neutral reconciliation.
Very irregular large accumulations would give a high CV and potential for negative reconciliation.
A variation of irregular large but statistically real doings combined with relatively even spaced ones (not quite certain how to stretch the analogy) gives the sweet smell of potential positive reconciliation. = Smiles for management and shareholders alike.
I know which type of reconciliation I would like but am aware others are possible given the current level of drilling
(20min delay)