While you're all posting about retractions and misery, look at...

While you're all posting about retractions and misery, look at these two statements from today's release...





To me it sounds like the two groups were largely at odds rather than "working in conjunction". It would be interesting to read the consultant's report.

- agreed with some of the modelling assumptions - block size?, search ellipse dimensions and orientation?, octants to be filled?
- disagreed with the use of channel samples - should only have used drill holes
- disagreed with the modelled widths - I'll bet the consultants didn't say the lenses should be wider!

A block model is a rectangular prism of sufficient dimensions to contain the entire resource and filled with small blocks, in this case 10x10x4 metres and 5x5x2 metres in size.

As far as block size goes...in this case the minimum block width is 5 metres perpendicular to the lode. In today's release is Figure A2, a list of the best drill hole intersections. In that list there is only one intersection with a true width greater than 5 metres. 38 out of 49 intersections are less than two metres wide. So the drill holes are indicating a resource width narrower than the minimum block width which means that the resource volume will be overestimated.

The search ellipse determines how far from the centre of a block to look for assays, in this case 90 metres along strike (45m east and 45m west), 35 metres perpendicular to strike (17.5/17.5) and 24 metres on dip (12m up/12m down). Now these dudes have used high grade channel samples from the faces of the tunnels mined by the old timers and they say..."Short search radii in the Y direction have been used in an attempt to limit extrapolation and smearing of these high-grade copper values". Now the search radii in the Y direction are up to 17.5 metres each way. This means that a block 15 metres off the line of lode may find a high grade channel sample within its search ellipse and benefit from that grade even though the block is not part of the lode, just low grade rubbish or waste in the hanging wall or footwall.

The use of channel samples is also questionable because of their nature - poor QAQC especially in the100 year old ones but probably in the CCZ and CRAE ones too. They are nowhere near the quality of RC and diamond drill samples. Geostatisticians would argue that they are a dodgy dataset incompatible with the other sets and likely to introduce bias. Even different drill hole datasets, RAB, RC and diamond can show bias.

The modelled width starts with the wireframes which define the lodes, Figure 5 in the release. These are 3-D shapes which constrain the block model - the only model blocks which count are the ones which fall inside the wireframes. Now if your wireframe is say 30 metres wide (in the Y direction) then that is how wide your resource can be. It is filled to full width with those 5m and 10m wide block. Grade is then interpolated into those blocks, some get grade and some don't but give the large search in the Y direction, 17.5 metres each way it is quite conceivable that there will be areas where grade is interpolated across the full width, 30 metres. Remember that most of the true widths are less than 2 metres.

As for the true widths...Figure A2 implies true width = 0.62 x apparent width. Now the strike, dip and width of this lode are quite consistent. Drill hole orientations however range from 30 degrees off perpendicular (0.86 x apparent width) to parallel to the lode (CRC005) where a true width can't be calculated.

Anyways, five years on and 36 drill holes later you have the same contained metal and the same resource classification, inferred. I'd love to see what that consultant's report inferred!