How do YOU use mining drills?

[Tertius]

Whether you use a higher or lower miner density depends on when you want to create new mining outposts. Your factory has some fixed ore demand. It doesn't care about miner density, it just wants to consume X ore per second.
So if you build a mine with low miner density, the mine output is somewhat low, so you need to build a bunch of outposts to get the throughput your factory requires. But after you built these outposts you can take a break, because they deplete only slowly.

On the other hand, if you build mines with high miner density, you need less outposts to achieve the same throughput. These mines deplete faster, so you need to create new outposts more often.

It's mostly personal preference. I prefer building mines with high miner density, because I need to build less outposts initially. This includes cleaning less area from biters and building less rail infrastructure. The factory is able to reach its possible output earlier, so I consider this more efficient.

On the other hand, it might be inefficient to build mines with miners overlapping their mining area. Tiles are not mined evenly.
If you observe a resource field half way through its mining time, you'll see overlapping areas are depleted first. The tiles are not mined equally. The miners are not talking to each other which tile they actually mine, so tiles on overlapping mining areas are mined twice while their central areas are mined only once. In the end, the overlapping areas become empty first, and the central tile of a miner is still full of resources, so mining becomes very slow in the last stages of such an outpost.

While mining, more and more miners become inactive on the outer parts of a resource field, so the number of miners decreases faster than with miners that don't overlap, so the general output decreases fast as well. Therefore, the whole lifetime of such an outpost is not much less than the lifetime of an outpost with non overlapping mining areas. The central most rich tiles are not mined faster. Such an outpost gets very slow in the last stage.

Usually, you don't mind that, because existing outposts just exist and don't require any active work to manage. However, I guess one need to create statistics if the fast ore throughput at the start of the outpost compensates the slow throughput near the end.

I thought about creating a mining outpost with high miner density where every single miner is circuit controlled. I set some throughput I want to achieve, perhaps 20-30% less than the maximum possible output. Now I enable only the miners with the most resources in their mining area, so the most rich tiles are mined first and most constantly. This way miners on the outer parts of the patch do not exhaust their tiles first, so the overall output of the mine stays constant and high.
However, while it seemed a perfect solution to make the most efficient use of a resource patch (it will keep up constant output almost up to the end, and because miners exhaust their tiles more evenly the mine is depleted very fast in general), I didn't make such an outpost. It's probably over-engineering.

[GregoriusT]

The miners are not talking to each other which tile they actually mine, so tiles on overlapping mining areas are mined twice while their central areas are mined only once. In the end, the overlapping areas become empty first, and the central tile of a miner is still full of resources, so mining becomes very slow in the last stages of such an outpost.

Uhm, Mining Drills do not get slower just because they only mine 1 tile as opposed to 25 tiles. They have a constant time. That is how i interpreted this part of your paragraph.

Yes outer drills vanish first and THAT reduces output for sure, but i wanted to just mention you might've misunderstood a minor element of the miner game mechanic.

[Tertius]

As far as I understand the game mechanics, a miner randomly chooses one tile for one mining pass, randomly from all available tiles in its area. Since for overlapping areas there are multiple miners, tiles in overlapping areas are chosen more often, so these deplete first.

If I speak of output in my previous post, I mean the accumulated output of all miners on a resource patch. This is simply higher the more miners are operating. One miner itself always has the same output, no matter if it has only one or multiple tiles to mine from.

[GregoriusT]

As far as I understand the game mechanics, a miner randomly chooses one tile for one mining pass, randomly from all available tiles in its area.

Yeah just wanted to clarify "all available tiles" refers to "tiles that actually do contain an ore", it does not include empty tiles. So a Miner with just one Ore Tile deadcenter in the middle of it is just as fast as a Miner with 25 Tiles of Ore. Just to clear up any potential misunderstandings for others.

[mmmPI]

However, I guess one need to create statistics if the fast ore throughput at the start of the outpost compensates the slow throughput near the end.

Short of actual stats i made mock up graph

overallthroughput of ore patch overtime.png (17.15 KiB) Viewed 2247 times

Supposedly the area under the curve is the same, but arguably it can change depending on when the research for productivity are done.

The exact shape between the 2 curve questionned marked would be the result of the aggregate data of individual miner and the ratio of density of ore between the edge of the patch and the inside. When do 1 miner stop working due to depletion. when does the 2nd one do also stop.

Possibly those curves have a flatter start if the ore patch is big enough so that not all mining drill work at all time.

The lower start is meant to represent less dense mining drill placement, which would start with lower yield, but i exagerated how they would last longer at a higher output per unit of time considering each individual miner has more underneath it for the same total amount of ore in the patch.

I made also a graph for individual mining drill , according to my conjectures :

individualoutput.png (9.44 KiB) Viewed 2247 times

Either it never deplete and has the same per second output all game ( this is again not factoring the productivity research curves should all trend upward by steps that are depending on player choice to do the research or not ).

It can be interrupted, because of output full, no power, anything. ( moving it to another location would look the same on the graph).

It can also be never actually properly connected once built.

In here the amount under the curve has no reason to really be the same.

[Tertius]

Short of actual stats i made mock up graph

Yes, that's exactly visualizing what I think about ore mine depletion. Of course the area covered by the graph is always the same, as it represents the amount of ore mined, and this is a fixed amount - the size of the ore patch. The ideal graph is a rectangle. Constant same output that drops to zero if the mine is depleted. The next to last graph of yours shows this.

To estimate the lifetime of an outpost, I consider the longest lasting tile. It's the (imaginary) tile with the most resources of the resource patch. It's somewhere in the center. The minimum lifetime of the outpost is the time required to mine this tile with a constantly operating miner.
If this tile is in an overlapping area of two miners its lifetime is halved, so it cannot be considered the longest lasting tile. So the longest lasting tile is actually the one with the most resources that is located in a non overlapping miner area.

This tile is mined more often, so depletes faster, the more of the other tiles available to the miner are exhausted.
So there is actually a difference between a layout with overlapping mining areas and a layout with non overlapping areas. Since overlapping areas get exhausted sooner, the miner has less tiles to mine from, so our tile with the most resources is actually mined more often. I got that wrong previously.

What remains is the question how the whole output develops. The output lowers with the amount of miners completely exhausting their area. How fast does it lower with which mine setup. I guess only a simulation could shed more light on that question.

[Nosferatu]

Earlier in this thread I said I also space out my miners.
That was inaccurate.
My blueprint has the center miners more dense because of ore density...

[GregoriusT]

Earlier in this thread I said I also space out my miners.
That was inaccurate.
My blueprint has the center miners more dense because of ore density...

This one is an interesting point, if you put the effort of spreading out the further miners while densely clustering the center Miners, you can ensure a consistent output at your Mines. This is a good hybrid solution, but only if you expect the Ore Patch in particular to actually run out during your playtime.

The 10M+ Patches later in the game dont really have this Issue normally, especially once you get to those, you can already beat all Biters til 90% Evo with just Blue and Military Science. Yellow Science is where you can just Iron Man dat shiz and laser em all to death by walking at them. Not to mention you already have invested a lot of time into things, so the later Patches are also starting to be depleted at a later point in time as opposed to right away at the beginning.

[mmmPI]

Yes, that's exactly visualizing what I think about ore mine depletion. Of course the area covered by the graph is always the same, as it represents the amount of ore mined, and this is a fixed amount - the size of the ore patch. The ideal graph is a rectangle. Constant same output that drops to zero if the mine is depleted. The next to last graph of yours shows this.

I tried to draw what you said, but i see now my mistake in the graph, what you say/said is more precise. It's written twice throughput, but as you say the top graph should be said to represent the quantity of ore left in a idealized situation where mining drill are operationnal 100% of time. But then on the second graph the line shouldn't be flat if it was also the quantity of ore left, in the second row it has to be the throughput. And then it's misleading to have used twice the same word and the same color for graphs.

To estimate the lifetime of an outpost, I consider the longest lasting tile. It's the (imaginary) tile with the most resources of the resource patch. It's somewhere in the center. The minimum lifetime of the outpost is the time required to mine this tile with a constantly operating miner.

I think you correct on the principles but in game there would be other phenomenon at play. On large ore patches, mining drill closer to the train station sometimes can't output because it's full. When a train arrive, immediately the mining drill from the back start working to replenish the supply, and they would also stop last their activity. Causing the back of the outpost to deplete faster. I found often time the last mining drill are not those on the absolute center/with more ore, but rather a little offset toward the "front" of the ore patch, where mining drill are only active when chests aren't full and all other mining drill aren't enough to fill the belts, the effect piles up if there are more mines than the belts can serve too; As those backing up first are the same located near the train station. What you described i think is the ideal case if the output belts are never fully compressed so that all mining drill works all the time. In such case i would agree with your reasonning about the time of the oupost being the time to mine the tile with the most ore.

What remains is the question how the whole output develops. The output lowers with the amount of miners completely exhausting their area. How fast does it lower with which mine setup. I guess only a simulation could shed more light on that question.

I agree with the need for a simulation for more answers on the question , given all the possible perturbation on the system to account for, it's hard to find at a glance which one will be the last mining drill. It can be made even more chaotic considering a player can drastically reduce the rate of depletion of an outpost when it reaches 20 % ore left or something, in early game, when you don't have many outpost yet and adding a new one divide the rate of the others. And you monitor more carefully the level of the 2 3 patches you have and you go "ohh i need a new iron patch" before the only one is depleted. But then with a second ore patch, the first one suddenly sees 50% less traffic. Causing only some mines to be active all the times, but not others modifying the rate of depletion of mining drills relative to each others. That's again probably over thnking it.

Worse is i don't even apply this knowledge/reasonning always in my game as you could see on the previous picture me add mining drill around the ore patch to catch the little bits of ore and being inefficient in the process as it increases the density of mining drill where the density of ore is the lowest. In a way to answer the original question of the topic : i spread mining drill everywhere EXCEPT where it would make the more sense

[mrvn]

The amount of aliens you have to kill and mining outposts you have to build for N ore is also the same.

But not the ore per second.

Ore per second per outpost is the metric to use, since it gets you the most full belts with the fewest outposts.

Ore in the ground is ore you aren't using.

Ore per second is determined by the number of miners. Doesn't matter if the belts are full or not.

In fact I would say a full belt will usually give you less ore per second because some miner will not be able to place all its ore on the belt. A miner that is blocked is just wasted resources, hence why you should not overlap miners when the ore field is large enough to fill the belt without it.

[mrvn]

I generally know if i'm going for a 20 hours 100hours or 250+hours game, that's roughly how i see them. And i will plan accordingly while looking at the quantity of the outpost.

Interesting, so at the end of your games, are your first Outposts running out or is more than half of the original Ore still left?

If at the end of the game your first ore patches aren't used up then you had too many resources per patch imho. What's the fun of a game where you don't have to expand your mining? It's a difficult balance to achieve though. I don't want to be always building new mining outposts but also I don't want ore fields to last forever.

Also if at the end of your game you have lots of mining outposts that aren't kind of on the last leg then you didn't estimate needed resources very well. If the end is in sight and you only need 1 million more iron ore while you have 2 million left in outposts then why would you build more mining outposts?

[mrvn]

Now all of this might mainly be because when I look at the Map Preview I look for a cluster of Iron, Copper and Coal in a corner of the Map, and then run straight to that Patch Cluster, destroying as many Trupens Rocks as possible, and build a Base in Biter infested Territory right away. This nets me with "Starting" Ore Patches that are in the 3 to 5 millions, while at the same time getting me closer to good Outposts (also Uranium and Oil are likely closer to there).

Yes, that breaks the intent of the map generator. You are supposed to start in the starting area with limited resource and then expand. With 3-5 million and vanilla you can finish the game multiple times just from that. No extra mining outposts needed at all.

[Illiander42]

In fact I would say a full belt will usually give you less ore per second

A full belt is 15/30/45 ore per second, regardless of how you load it.

A miner that is blocked is just wasted resources, hence why you should not overlap miners when the ore field is large enough to fill the belt without it.

How about you grab a second belt from the patch instead? Starter patches on default settings can support 3-5 yellow belts of ore.

Also if at the end of your game you have lots of mining outposts that aren't kind of on the last leg then you didn't estimate needed resources very well. If the end is in sight and you only need 1 million more iron ore while you have 2 million left in outposts then why would you build more mining outposts?

Because you'll get to the end faster if you have 10 outposts each giving you 4 red belts each, rather than 1 outpost giving you 2 red belts.

I would much rather have 50 tiles of 500 ore each than 1 tile of 25000 ore.

Ore per second entering your factory is the important number. Ore under your miners is almost irrelivant once you have construction bots.

[mrvn]

In fact I would say a full belt will usually give you less ore per second

A full belt is 15/30/45 ore per second, regardless of how you load it.

But a miner gives you e.g 1.1 ore per second. So 13 miners will be 14.3 ore per second, not a full belt. And 14 miners would be 15.4 ore per second and you loose 0.4 ore per second for the sake of filling the belt.

A miner that is blocked is just wasted resources, hence why you should not overlap miners when the ore field is large enough to fill the belt without it.

How about you grab a second belt from the patch instead? Starter patches on default settings can support 3-5 yellow belts of ore.

DUH. Obviously you don't just snake one belt through the whole ore patch. But I tend to build miners as simple rows or columns and ore patches are often large enough that a single row of miners across the patch will more than fill the belt. In those cases overlapping the miners just means you have miners that are idle. Wasted resources.

Note: I still frequently build more miners than the belt can carry for two reasons:

1) I'm lazy. I build and forget. I will get all the ore eventually.

2) Ore density isn't uniform across a patch. If you aren't splitting the ore patch so belts go from the middle to the outside then miners at the start of the belt will exhaust quickly leaving fewer miners working.

Maybe one should start with mining just the border of an ore patch. Use up the low density parts and get the patch into a more rectangular shape. And then build a nicely balances mine that fills the belts uniformly. Naaaahhhh, see point 1.

Also if at the end of your game you have lots of mining outposts that aren't kind of on the last leg then you didn't estimate needed resources very well. If the end is in sight and you only need 1 million more iron ore while you have 2 million left in outposts then why would you build more mining outposts?

Because you'll get to the end faster if you have 10 outposts each giving you 4 red belts each, rather than 1 outpost giving you 2 red belts.

I would much rather have 50 tiles of 500 ore each than 1 tile of 25000 ore.

Ore per second entering your factory is the important number. Ore under your miners is almost irrelivant once you have construction bots.

You do know that you can finish the game (vanilla) with just 2 yellow belts of iron and copper ore each in under 6 hours, right? Actually those 2 belts give you enough ore to do all the research and start a rocket with satelite (not even needed and takes halve the research) in 3 hours. The other 3 hours is for building the factory.

So if you NEED 10 outposts with 4 red belts each then I'm guessing you are playing an infinite game.

[GregoriusT]

2) Ore density isn't uniform across a patch. If you aren't splitting the ore patch so belts go from the middle to the outside then miners at the start of the belt will exhaust quickly leaving fewer miners working.

This is actually a great Idea, how did I not think of that one before, that way the Miners wont end up depleting the border of the Patches first, and the overall Ore per Second would be a lot more evened out without sacrificing total throughput at all!

It is even better if you happen to use one of them "Infinite Ore Patch in the Center of normal Ore Patches" Mods.

Not to mention it would provide an even larger amount of Ore Belt lanes in the Extreme Mining Productivity Postgame!

Though that also means i would need to make my Mining Blueprint rotationally symmetric or have an alternative Version with the Power Pole and the Landmine in swapped positions. Or click to remove the Landmines that are free floating on the Ore Patch.

[mrvn]

2) Ore density isn't uniform across a patch. If you aren't splitting the ore patch so belts go from the middle to the outside then miners at the start of the belt will exhaust quickly leaving fewer miners working.

This is actually a great Idea, how did I not think of that one before, that way the Miners wont end up depleting the border of the Patches first, and the overall Ore per Second would be a lot more evened out without sacrificing total throughput at all!

It is even better if you happen to use one of them "Infinite Ore Patch in the Center of normal Ore Patches" Mods.

Not to mention it would provide an even larger amount of Ore Belt lanes in the Extreme Mining Productivity Postgame!

Though that also means i would need to make my Mining Blueprint rotationally symmetric or have an alternative Version with the Power Pole and the Landmine in swapped positions. Or click to remove the Landmines that are free floating on the Ore Patch.

Wouldn't you just take your big mining blueprint and then reverse the belt directions for half of it? Everything else can stay the same.

Note: With yellow, red, blue belts going from the middle to the side you can mine quite huge ore patches in maximum time. Probably bigger than you normally find in the game. But if you are limited to yellow belts then splitting the patch in half helps a lot.

[mmmPI]

Ore under your miners is almost irrelivant once you have construction bots.

Where it can be relevant is regarding the research for mining productivity cost, if you have 20 times the price of the research worth of ore covered by your mining drills, then after the research is done, you are left with 95% of the initial quantity and you get a 10% bonus of extra ore added to this effectively "increasing" the amount of ore that will be generated by the mining drill you have already placed.

To compare it to a situation where you'd only have 2 times the price of the research worth of ore being covered, it would reduce by 50% your "expected available ore left" to do the research, and then getting an extra 10% on this remaining quantity would not amount to effectively "more ore" after the research.

I would much rather have 50 tiles of 500 ore each than 1 tile of 25000 ore.

If you multiply that by 100 though, it changes things to me, with 5000 tiles of 500 ores one has a big ore patch that require many mining drill to be stamped only to last less than an hour , whereas with 100 tiles of 25000 ore, you can place way fewer mining drill, but they will last you the time to do some other things.

But a miner gives you e.g 1.1 ore per second. So 13 miners will be 14.3 ore per second, not a full belt. And 14 miners would be 15.4 ore per second and you loose 0.4 ore per second for the sake of filling the belt.

To me it doesn't change the fact that the full belt give you more ore per second than the non full belt. If you only use 13 miners the output of the factory is a non-full belt of 14.3 ore per second. If you use 14 miners , you have 15 ore per second. You have 0.7 more ore per second. Full belt 15 ore per second, non full belt 14.3 ore per second.

You do know that you can finish the game (vanilla) with just 2 yellow belts of iron and copper ore each in under 6 hours, right?

It seem to me that it is not the optimal choice regarding your precedent reasonning, one would lose less ore per second if you were to use a red belt instead of 2 yellow belts no ?

So if you NEED 10 outposts with 4 red belts each then I'm guessing you are playing an infinite game.

It's quite comon that i have games with more than 10 outposts. Most my games past 100 hours i would say. They are not infinite but sometimes unfinished

[mrvn]

But a miner gives you e.g 1.1 ore per second. So 13 miners will be 14.3 ore per second, not a full belt. And 14 miners would be 15.4 ore per second and you loose 0.4 ore per second for the sake of filling the belt.

To me it doesn't change the fact that the full belt give you more ore per second than the non full belt. If you only use 13 miners the output of the factory is a non-full belt of 14.3 ore per second. If you use 14 miners , you have 15 ore per second. You have 0.7 more ore per second. Full belt 15 ore per second, non full belt 14.3 ore per second.

If you only look at that one belt then sure. But on a larger scale that 14th miner won't be just left out. It will be placed somewhere else. So you get one belt with 14.3 ore per second and on some other belt you get an extra 1.1 ore per second for a total of 15.4 ore per second.

Now if you have belts going straight from the miners to smelters this seems stupid. The furnaces will use 14 ore per second or 15 ore per second. With 14.3 ore per second you either end up with ore backing up or a furnace not running all the time. So you have to balance ore from multiple belts get everything balanced to nice integers (or close enough).

But if you have mining outpost with trains that problem goes away. Regardless of what fractions you have at some point the train is full and the smelter will get a nice 15/30/45 ore per second on it's belt when unloading the train. You just need enough mining outposts so the sum of all miner ore is >= the consumption by the furnaces.

And then 14 rows of 13 miners will produce 200.2 ore per second, allowing for a furnace setup with 200 ore per second consumption. On the other hand 13 rows of 14 miners will only produce 195 ore per second.

You do know that you can finish the game (vanilla) with just 2 yellow belts of iron and copper ore each in under 6 hours, right?

It seem to me that it is not the optimal choice regarding your precedent reasonning, one would lose less ore per second if you were to use a red belt instead of 2 yellow belts no ?

So if you NEED 10 outposts with 4 red belts each then I'm guessing you are playing an infinite game.

It's quite comon that i have games with more than 10 outposts. Most my games past 100 hours i would say. They are not infinite but sometimes unfinished

Sure, a red belt means the 2 fractions left on the yellow belt get added and might be enough for one extra miner to fit onto the belt. But red belts costs more too. If you can cover the whole ore patch (or a row across the patch anyway) with 13 miners instead of 14 (or more) then why add the cost of the extra miner and the red belts just so the ore patch is exhausted earlier and you have to move all the miners.

Let me be explicit: The argument is not that 13 miners are better than 14. But that 13 miners and putting the 14th miner somewhere else is better.

PS: Imho full belts are overrated. With different recipe times and item counts and later modules it's often impossible to get a perfect 15/30/45 items per second. Maybe in a mega factory the time saved by full belts becomes significant. Personally I don't mind some gaps in belts but an expensive assembler with modules that isn't working makes me cringe. Which is different from the start of the game where I don't mind a stone furnace not working due to the output being backed up. They are dirt cheap and a nice buffer in case I need 100 iron plates.

[Illiander42]

But a miner gives you e.g 1.1 ore per second. So 13 miners will be 14.3 ore per second, not a full belt. And 14 miners would be 15.4 ore per second and you loose 0.4 ore per second for the sake of filling the belt.

That's why you stick a compressor on the outputs of the miners. To turn those almost-full belts into full belts.

DUH. Obviously you don't just snake one belt through the whole ore patch. But I tend to build miners as simple rows or columns and ore patches are often large enough that a single row of miners across the patch will more than fill the belt. In those cases overlapping the miners just means you have miners that are idle.

I find it's rare for a patch to need more than red belts to not have miners stalled. Maybe I just lay out my mines better than you?

You do know that you can finish the game (vanilla) with just 2 yellow belts of iron and copper ore each in under 6 hours, right? Actually those 2 belts give you enough ore to do all the research and start a rocket with satelite (not even needed and takes halve the research) in 3 hours. The other 3 hours is for building the factory.

Why would I want a rocket every three HOURS? That's really, really slow.

[GregoriusT]

Wouldn't you just take your big mining blueprint and then reverse the belt directions for half of it? Everything else can stay the same.

Normally yes however my Mining Blueprint contains a Half Halo of Landmines, which will clash with the 180° rotate.

PS: Imho full belts are overrated. With different recipe times and item counts and later modules it's often impossible to get a perfect 15/30/45 items per second. Maybe in a mega factory the time saved by full belts becomes significant. Personally I don't mind some gaps in belts but an expensive assembler with modules that isn't working makes me cringe. Which is different from the start of the game where I don't mind a stone furnace not working due to the output being backed up. They are dirt cheap and a nice buffer in case I need 100 iron plates.

"Full Belts" indeed do not matter, they are just a metric by which a Player can communicate to another Player how much Ore they are getting. In all practical applications you should always aim to have some space left on your Mining Belts, so that later Mining Productivity Science wont fill it up instantly.

Not to mention if you even have just one Mining Productivity Science, you cant sync Miners with Belts. That only works with the 30 Miners no Productivity Setup.

One Miner one Steel/Electric Furnace works best with 3 Mining Productivity Levels as far as I know. (my source is a Warptorio lets play, might be inaccurate as heck)