My prospecting pick is not working right

[tulamort]

OK  so after a lot of practical research on how to improve my use of the prospecting tool, I finnally begin to understand how the tool actually works (if you have precise data it would be huge though, like what zone does it search, how many blocks, what is the precise meaning of "rare", "poor" etc...).

So I set myself to a proper, clean session of tryhard prospection. After too hours of not being able to get more than "very poor" on a hematite concentration, I began to get crazy, and finally I turn my solo game in creativ mode and go mad pickaxing every corner of the floor. Only to find out that my hematite was literraly right 30 blocks bellow my main researching point.

So I go and prospect the very touching block to this huge hematite gisement, and... "very poor". I try to stay cool but I think my mind is forever broken.

Is this a bug? Tell me this is a bug, please.

 

[Streetwind]

The prospecting pick in density search mode (mode 1) will report you the spawn chance of the ore in this area. Not what actually is there, but rather how likely it is that something could be there.

Your area has a very poor chance for hematite.

However, even very low chances occasionally become reality. The hematite deposit succeeded on a very unlikely spawn roll and pop, there it was. You then managed to also luck into discovering it. (This actually works unusually well with iron, due to the sheer size of iron ore deposits, which makes them prone to accidental discovery.)

When you switch your prospecting pick to node search (mode 2) by pressing F, it will no longer report spawn chances, but rather report on actual ore blocks nearby. However, the range on this mode is only six bocks in all directions. It is meant to allow you to detect ore veins you might be just barely missing, after you have already identified a suitable area for digging using density search.

[tulamort]

25 minutes ago, Streetwind said:

The prospecting pick in density search mode (mode 1) will report you the spawn chance of the ore in this area. Not what actually is there, but rather how likely it is that something could be there.

 

Oh, ok well I definetely didn't at all understood the functionnement of this. I have to say, now it seems to really be a frustrating mechanic... I mean wow, am I suppose to wander for hours to find a high spawning chance place while it could actually be free from any ore?? I don't really get how I should have fun with this.. Don't mind me, I'am a bit pissed off right now, I will cool and try again.

[Streetwind]

It definitely is the one mechanic in the game with the highest learning curve. Beyond just understanding how the tool itself works, you need to develop a methodology that is conducive to consistent results. And yeah, there's always an element of randomness. Less so in the spawn chances themselves (I've never not found ore in an "ultra high" abundance area - in fact, once I found five copper deposits in a single vertical shaft), and more in the way that what you can find at all depends on the rock layers available to you in your location.

There is no single location that can host all ores. You will always have to travel, sometimes considerable distances, to find everything you need. This is intentional. If that sounds like something that would greatly annoy you, then this might not be the game for you.

Edited October 14, 2023 by Streetwind

[tulamort]

Well Yes, I actually have read your method on reddit, and I just retry to apply it on hematite.

 

Found a reading for it

tracked it to ultra high.

dug 4 pits to bottom layer searching in node mode every 12 blocks going down. Each pit spaced of 20 blocks, all in a ultrahigh concentration.

Found like a ton of copper and not any freaking iron...

gaddam...

Also, what is the meaning of the purthousant numbers? (mine was around 6.8 at the best)

Thanks for your answers mate.

[Bumber]

I'm pretty sure iron ores show very low concentration just because they're a very small volume of the chunks taken vertically. The flat shape of the deposits make them easier to find, and you can follow them horizontally through multiple chunks, which is really what matters.

6 hours ago, tulamort said:

Also, what is the meaning of the purthousant numbers? (mine was around 6.8 at the best)

Permille.

6.8‰ = 0.68%

[Streetwind]

14 hours ago, tulamort said:

I just retry to apply it on hematite.

As far as difficulty tiers of prospecting go, tier 1 would be copper, which can be hosted by almost any rock and generates absurd numbers of deposits in higher abundance areas. It's enough to just go off prospecting anywhere until a halfway decent reading shows up, and digging there.

Tier 2 would be ores like cassiterite, which require knowledge of and dealing with rock layers, but still generate in largely the same way. You need to know that there are areas where you are more likely to find good readings than in others, and you want to bias more towards really good readings rather than just decent ones; but once you have that really good reading, you can still just dig and expect to find ore.

The iron ores would then be tier 3, and adds another layer of difficulty: knowledge of the quirks of the ore itself.

See, iron is not like the ores before it. Earlier ores all generate small to medium size deposits, but multiple of them per chunk column. Copper will try as often as 25 times per chunk column to succeed on a spawn chance roll and place a deposit. Iron ores are the polar opposite, generating massive deposits that span multiple chunk columns, but only attempting less than one try per chunk column. Magnetite being the worst at 0.3 tries per chunk column. Magnetite could have a 99% chance to succeed its spawn roll in an "ultra high" area, but before it can even try, it would have to beat another dice roll in order to be allowed to try to spawn, and that roll has a flat 70% failure rate. That compounds with the fact that even if everything goes right and the ore tries to spawn, the random location selected during the attempt might not be able to carry it (i.e. it selected a water block in an underground lake, or the wrong rock layer, or something out of bounds for its min and max y-level, or most of it got deleted by a cave system that generated afterwards, and so on.)

Ergo, prospecting for iron is more difficult; it is not enough to find the one "ultra high" hotspot, because the ore might simply have not been allowed to even try spawning there. You need to examine the entire surrounding area. On the upside, due to the sheer size of iron deposits, you only need to make new shafts every 50 blocks, instead of every 20 like is recommended for other ores. And once you do find an iron deposit, chances are you won't need to look for another for a long, long time.

 

14 hours ago, tulamort said:

Also, what is the meaning of the purthousant numbers? (mine was around 6.8 at the best)

It's closer to debug output than anything useful. In a few situations you could let it guide you while trying to figure out in which direction a spawn chance increases, but in general I recommend to people to just ignore the number and focus on the wording of the prospecting result.

Specifically, the permille number represents the share of blocks in the block column in which you broke the first of your three prospecting samples that generated this result, which might on average be replaced by this ore, given the interpolated spawn chances in the surrounding area and the specific config settings of this ore, and taking into account the geological makeup of the block column.

Due to the way the math functions under the hood, and the inputs it pays attention to, this number can vary greatly between ores... and thus can be extremely misleading. A 0.2‰ of cassiterite will almost guarantee you hitting a deposit because it comes along with an "ultra high", whereas a 30.0‰ of halite can have you dig twenty shafts and find nothing because it's just "poor" or even "very poor", and halite has super quirky spawn mechanics to boot. Hence, trying to derive meaning from this number, especially as a newcomer, is a recipe for frustration and wasted time.

 

Edited October 15, 2023 by Streetwind

[Maelstrom]

In my last world, I found two hematite veins with a vertical separation of five blocks.  I assume that one vein spawned in a chunk column and the other in an adjacent column and they happened to overlap?

[Streetwind]

That would be my interpretation too.

In fact, since chunks have borders, it would even be possible for the centers of deposits to only be one block apart on the x-axis, if both spawned directly on their respective chunk border

[Bumber]

19 hours ago, Maelstrom said:

In my last world, I found two hematite veins with a vertical separation of five blocks.  I assume that one vein spawned in a chunk column and the other in an adjacent column and they happened to overlap?

I've been looking at the prospecting code, and I'm not clear on what happens if you prospect starting on an overlapping x/z coord versus not.

[Maelstrom]

The first block you break with the pro-pick determines the reading.  Prospect a second time with a block adjacent to the first prospecting attempt will yield almost identical results but may be a bit different.  The y coord doesn't matter, to my understanding.

[Bumber]

On 10/18/2023 at 11:19 AM, Maelstrom said:

The first block you break with the pro-pick determines the reading.  Prospect a second time with a block adjacent to the first prospecting attempt will yield almost identical results but may be a bit different.  The y coord doesn't matter, to my understanding.

I meant specifically in your case of overlapping deposits versus a non-overlapping x/z.

I looked through the available source code to determine the precise meanings as asked in the OP. Node search is simple:

Trace: < 10 blocks
Small: 10 to 19
Medium: 20 to 39
Large: 40 to 79
Very Large: 80 to 159
Huge: 160+

Prospecting is much more complicated. It works on something called totalFactor:

No reading: < 0.002 totalFactor
Miniscule: 0.002 to 0.025
Very Poor: 0.025 to 0.133
Poor: 0.133 to 0.267
Decent: 0.267 to 0.4
High: 0.4 to 0.533
Very High: 0.533 to 0.667
Ultra High: 0.667+

totalFactor seems to be the regional ore distribution factor at the x/z coord (oreMapFactor, value 0.0 to 1.0,) times the proportion of specific ore blocks to regular blocks (rockFactor, again should be 0.0 to 1.0,) in the blockColumn (as it was when freshly generated) at the x/z coord. This means to get Ultra High, you'd need the max possible oreMapFactor at the prospected coord, as well as have over 2/3rds of that column's blocks consist of that type of ore. Not sure that seems right.

The numerical reading given (in permille) after the quantity description is determined by taking the totalFactor and multiplying it by 1000 times the expected proportion of blocks that belong to an ore deposit started in the prospected chunk versus the total number of blocks in the chunk. (ppt = totalFactor * 1000 * absAvgQuantity / qchunkblocks; where absAvgQuantity is the number of blocks in an average of 100 generated ore disks via RNG times triesPerChunk using the ore's json definition, and qchunkblocks is the generated terrain height at the x/z coord times the chunksize (i.e., 32) squared.)

Edited October 19, 2023 by Bumber