The panel on the right is a block model: the pit is discretised into a grid of blocks, each holding one estimated ore grade. Grade is never measured everywhere — only at a handful of sparse drill holes — so every block's grade is estimated from those samples by inverse-distance weighting (IDW):
grade(block) = Σ( grade_i / dist_i^p ) / Σ( 1 / dist_i^p )
Closer drill holes get more influence than distant ones, and the power exponent p controls how sharply that influence falls off: low p (≈1) blends samples smoothly over a wide area; high p (≈5–6) makes the estimate hug the nearest hole almost exactly, producing "bullseye" patterns around each sample.
- Cutoff grade — the minimum grade that pays for mining, hauling and processing a block. Blocks at or above it are classified ore (sent to the mill); below it, waste (sent to the dump). Moving this slider directly changes the tonnage and average grade reported below — the classic mining trade-off: a lower cutoff mines more tonnes at a lower average grade, a higher cutoff mines fewer, richer tonnes.
- Drill holes — more samples shrink the average distance to the nearest hole, so the estimate tracks the real (simulated) deposit more closely — watch the "Est. vs. true match" accuracy stat rise as you add holes.
- True deposit overlay — this demo also carries a hidden underlying grade field the drill holes were sampled from (with assay noise). Toggling it on lets you see how well the IDW estimate reconstructed the real ore body, which is the whole point of exploration drilling in real geometallurgy.