I thought I understood prospecting but what is up with this. The % went up but the reading went from high to decent.

[Camron Thackeray]

See pictures

I have read that we shouldn't spend too much time looking at the per mil reading and focus on the descriptive word like Hight, or Ultra high. I know that different ores have different rarities and they are relatively rare or common to themselves not overall. BUT why do these two readings fairly close together go up in per mil but go from a High to a Decent? Which one is better for me to mine down at? 

[MKMoose]

1 hour ago, Camron Thackeray said:

I have read that we shouldn't spend too much time looking at the per mil reading and focus on the descriptive word like Hight, or Ultra high. I know that different ores have different rarities and they are relatively rare or common to themselves not overall. BUT why do these two readings fairly close together go up in per mil but go from a High to a Decent?

The descriptive word is influenced by two factors:

• the ore map value - an under-the-hood map which decides how common ores are, which can be interprered as proportion of maximum possible commonness,
• the proportion of ore-bearing rock within the allowed Y ranges - how many blocks in the block column at the reading's coordinates within the allowed Y range the ore can generate in (this is why ores which spawn only in sedimentary rocks tend to have low readings, as sedimentary rocks are generally a thin top layer and the igneous stone below can't host the ore); this introduces a small random factor due to how it's calculated.

For example, if the ore map decides that at the reading's location a sedimentary-only ore like galena ore has 60% of maximum commonness, and there is a 40-thick sedimentary layer with surface at Y = 120, given that galena is allowed to spawn at any Y level, gives you a 0.6 * 40 / 120 = 0.2 total factor, which ends up being "poor".

For an ore like borax which spawns at Y levels of 44 through 110 only the blocks at those relevant levels will be considered, giving a 0.6 * 30 / 66 = ~0.27 total factor, translating to "decent".

 

The permille value is influenced by the two above factors, as well as the following:

• the deposit's absolute average quantity of ore blocks - a rough estimate for the average quantity of ore blocks per chunk at maximum deposit frequency - this introduces a small random factor on each world load,
• the surface Y level, used as the denominator to estimate the proportion of ore blocks in the area.

For example, copper has an absolute average quantity of ~4330 if I recall correctly, so for a total factor (ore map multiplied by rock proportion) of 0.4 and surface Y level of 120 it has a permille value of 0.4 * 4330 / (120 * 32 ^ 2) = ~14.1‰ (the 120 * 32 ^ 2 part is the equivalent chunk volume at the surface Y level).

 

So, to answer your question: the permille value can be lower in spite of the descriptor getting higher when the surface level at the reading coordinates is higher, so there's more rock for the amount of ore. Sometimes the small random component which is reinitialized on each world load can also influence this.

I would actually love to know how close I am in my estimate, since your Y level readout on the map seems dysfunctional: the "high" reading is probably Y = ~170-180, and the "decent" is probably Y = ~130-140. There are some possibilities which could throw it off somewhat that I can't account for.

Feel free to ask for any clarification, since I'm probably omitting some details or assuming something is intuitive when it might not be. And keep in mind that the system can be a bit inconsistent or buggy in a few places.

 

1 hour ago, Camron Thackeray said:

Which one is better for me to mine down at? 

As a general rule, the better word tends to be better to mine at. There are some caveats, e.g. it's almost always preferable to avoid mining in the mountains for most purposes, but mining at the better word is a sufficient simplification for almost all purposes. In your case, the difference is quite minimal, so it's pretty much "whichever you like".

 

Side note: there's generally no need to take density readings any more frequently than every ~100 blocks. The ore map doesn't change fast enough for a smaller movement to make a meaningful difference in the vast majority of cases.

Edited September 29 by MKMoose