Russett:
Well, what I'm suggesting is this.
Statistically, a public opinion poll asks 1000 people how they intend to vote. Say, 54% indicate candidate 'A'. The poll is accurate to within +/- 3% 19 times out of 20. Standard stuff. That 19/20 is 95% which corresponds to 2 SDs. So, the SD is 1.5%.
However, lets say they only asked 5 people...but they asked 5 people 200 times. Cumulatively, the results are the same. Individually...what would the SD be? I can see results ranging between 0%, through 20, 40, 60, and 80%, and a few 100% results. How meaningful is the SD of those values?
Remember as well, that the SD will increase with the larger size of the diamond classes. The SD/Avg will be lowest for the smallest class, and highest for the largest class.
All of which renders the use of statistics relatively useless here. After all, you will hopefully be mining diamonds larger than 99.9% of the ones recovered by caustic fusion. Theoretically...its that SD you need worry about, and that SD only. But..it's an unknown. :-)
Not all averages unaccompanied by SD or variance values lead to failure. Not professionally at least. An average accompanied by an SD that is meaningless is worse than useless...it's misleading.
Remember the adage: Lies, damned lies, and statistics. :-)
A drill hole of about 15 kg may have had 14 diamonds present in the 0.3 to 1.18 mm size class. On average. Individually, they may have varied from close to zero, to 30 or 50. Hard to say, and like you...I would like to know. But...it's relatively meaningless, since those diamonds will never be recovered in a mine in any case.
Using the next smaller size class...the SD/Avg ratio would be lower. Still meaningless. Using the next larger size class...I can imagine the SD/Avg ratio was extreme. Still, those diamonds are not recoverable for the most part.
What you're looking for is a minibulk sample at each drill hole. It doesn't exist.
But it would be nice to have.
Regards,
WillP |
