RE:An expert's viewThat was an interesting release. Not sure it was advised since they haven't determined what the issue is and it's rather technical.
VG causes issues when sampling. If you were a core logger, subconcious or not, you likely lean to sampling the better material. The CRM didn't show any differences between the two labs which hints at samplig bias.
Would be interesting to see the assay from the first split and see if that one was consisently higher. That should shed some light on a potential sampling bias as well.
There are variables in the fusion process in the oven that could cause some of the gold to report to the slag or lost to the cupel in the later step. Unlikely, but needs to be checked.
The real weakness in the dataset is the difference between precision and accuracy. Accuracy is how close the two samples are from the two different labs in this case. There appears to be some differences, but it's a little difficult to determine from the diagram. Precision is the absolute value in the rock. That would be hard to determine once the core is split, but requesting the rejects from both labs, combining and doing empiring assaying would be a good start. Maybe difficult, but a strong Aqua regia leach with an oxidant for a boost might help getting the all the gold in solution as another method.
I think the PR is too technical for most, but there are missing facts that make it confusing.
Just because one lab provides consistent lower values doesn't mean it's not precise.