Smearing and the nugget effect: when one rich metre carries the hole

Smearing is when a single very high-grade sample is averaged across a long interval, making the whole interval look better than most of the rock in it. In gold it often goes hand in hand with the nugget effect, where coarse gold is so unevenly spread that one sample can be rich while the sample next to it is almost barren.

How one metre carries an interval

Drill samples are usually assayed in 1m lengths (sometimes shorter in diamond core, or 2-4m composites in early RC drilling). The reported interval is the length-weighted average of those samples. Averages are easily dominated by an outlier.

Take a hypothetical headline: 20m at 2.35g/t gold. That is 47 gram-metres, comfortably "good" on the usual gram-metre scale. Now look at the assays inside it:

The 1m at 40g/t contributes 85% of this interval's gram-metres. The other 19 metres average 0.37g/t, which on its own would be below the cut-off most open pits would use. Nothing here is false: the arithmetic is correct and the assays are real. But "20m at 2.35g/t" invites you to imagine 20 metres of 2.35g/t rock, and that rock is not there.

The nugget effect

In some gold deposits, much of the gold sits in coarse particles rather than being finely disseminated. A 3kg sample of crushed rock might catch two visible nuggets and assay 60g/t, while a duplicate from the same metre returns 4g/t. Geologists call this high variability the nugget effect.

It matters to investors in three ways:

• Repeatability. Re-assaying the same pulp, or assaying the other half of the core, can give a very different grade. Good announcements report repeat or duplicate results for high-grade samples.
• Bias in both directions. A nuggety deposit can be undersampled (missing gold that is really there) as easily as oversampled. A single 40g/t metre does not prove the deposit is rich, and a low assay does not prove it is barren.
• Assay method. Standard 30-50g fire assays can struggle with coarse gold. Companies in nuggety deposits often use screen fire assays, bulk leach (such as LeachWELL) or PhotonAssay on larger samples. JORC Table 1 Section 1 requires the assay technique to be described, so check which was used.

Cut-off grades and internal dilution

How a company groups samples into an interval is a choice, and it changes the headline. Two settings matter:

• Lower cut-off grade: the minimum grade a sample needs to start or extend an interval, for example 0.3g/t or 1g/t.
• Maximum internal dilution: how many consecutive metres below cut-off can sit inside an interval before it is split, for example "up to 4m of internal dilution".

Generous settings join scattered higher-grade samples into one long interval. Strict settings break the same hole into several short intervals. Using the example above with a 0.5g/t cut-off and 2m maximum internal dilution, the hole might report as something like 3m at 13.8g/t plus two or three short low-grade intervals, rather than one 20m interval. Neither version is wrong, but they paint different pictures.

JORC Table 1 Section 2 (data aggregation methods) says weighted averaging techniques, any maximum or minimum grade truncations and cut-off grades are usually material and should be stated. It also says that where aggregated intercepts include short lengths of high-grade results and longer lengths of low-grade results, the procedure should be stated and some typical examples shown in detail. If the cut-off and dilution rules are missing, that is a disclosure gap worth noting on its own. DrillScore records these gaps on the red flags page.

Reading "including" intervals

Announcements commonly report an outer interval and a richer inner one: "48m at 1.6g/t from 102m, including 6m at 9.8g/t from 131m". The inner interval is part of the outer one, not additional to it. To see what the rest of the rock averages:

Worked through:

• Total: 48 × 1.6 = 76.8 gram-metres
• Included: 6 × 9.8 = 58.8 gram-metres
• Remaining: (76.8 - 58.8) ÷ (48 - 6) = 18 ÷ 42 = 0.43g/t

So the hole is better described as a 6m high-grade zone inside 42m of low-grade halo. That may be geologically very encouraging (a high-grade shoot is often exactly what explorers hope for), but it is a different proposition from 48m of 1.6g/t rock. Be careful never to add the "including" figure to the outer interval; that double counts. Watch for headlines that quote only the inner interval, too; cherry-picking and balanced reporting covers that pattern.

Sometimes there is an "including" inside an "including": "48m at 1.6g/t, including 6m at 9.8g/t, including 1m at 41g/t". Work from the inside out. Here the single metre supplies 41 of the 76.8 gram-metres, more than half of the whole interval.

Top-cuts

When estimating resources, geologists usually apply a top-cut (also called a top cap): any assay above a chosen grade is reduced to that grade, so a handful of outliers cannot dominate the estimate. Top-cuts are often chosen statistically, from the point where the grade distribution breaks down.

Applying a 10g/t top-cut to the original example changes it from 47 gram-metres to 17 gram-metres:

• Rich metre capped: 1m at 10g/t = 10 gram-metres
• Other 19 metres unchanged: 7 gram-metres
• Top-cut interval: 20m at about 0.85g/t

Drilling announcements usually report uncut grades, which is acceptable as long as the company says so. Some report both cut and uncut figures, which is helpful. When you see an extraordinary interval, ask what it would look like with a top-cut similar to the one the company uses (or would likely use) in its resource. If the company already has a resource, its top-cut is normally disclosed in JORC Table 1 Section 3.

Signs an interval is being smeared

And signs of careful reporting:

How DrillScore handles it

Where an announcement provides an "including" interval, DrillScore works out the share of the interval's gram-metres that comes from the inner zone and flags intervals that rely heavily on a short high-grade section. That is a factual description of the reported numbers, not a judgement on the company. See the methodology for how this feeds the rating, and the gold drill results page to see it applied. As always, this is general information, not a recommendation about any stock.

The short version

• A single rich metre can supply most of a long interval's gram-metres; in our example 1m at 40g/t made up 85%.
• Back out "including" intervals with (total gram-metres - included) ÷ (total width - included width).
• Cut-off grade and internal dilution rules change the headline, and JORC expects them to be stated.
• Top-cuts show how an interval would look without its outliers.
• Nuggety deposits need repeat assays and suitable assay methods before any single grade is taken at face value.