How are productivity modules for miners calculated?

[BitterSwede]

Couldn't find any info on the wiki for it, so I'm asking here.
It's not super important for me to know, just something I'm curious about.

Electric miners are subject to both modules and the research Mining Productivity. MP gives 2% productivity per level and modules give up to 30% productivity (3 module slots).

But is the productivity from modules added on top of the productivity from research, or are both added together first and then applied?

For example, if you have MP at level 10, you'd get 20% bonus from that. With three module Mk3 you'd get 30%.
Will the end result be 150% productivity (both added together plus base, so 100 + 20 + 30) or 156% productivity (30% bonus to 120% productivity, so (100 + 20) x 1.3 )?

[Lady Naween]

Add them together, you can check by looking at the info on your miner and slotting a module in. I have a base in my current world of +142. Adding one module (level 3) brings it to 152. 142+10 (which is what a level 3 gives) = 152

[BitterSwede]

Цитата допису Linda Hartlen:

Add them together, you can check by looking at the info on your miner and slotting a module in. I have a base in my current world of +142. Adding one module (level 3) brings it to 152. 142+10 (which is what a level 3 gives) = 152

Thank you

[The Chaotic Coder]

Just a side note, you may be interested to know... When prod. mods and research are applied to miners working on Uranium, they increase the ore yield without increasing the amount of sulphuric acid used. The base ratio is 10 units of acid for each ore chunk, but with 150% prod. bonus you get 1.5 ore chunks per 10 acid (or 3 ore per 20 acid)

[Warlord]

Цитата допису jchardin64:

Just a side note, you may be interested to know... When prod. mods and research are applied to miners working on Uranium, they increase the ore yield without increasing the amount of sulphuric acid used. The base ratio is 10 units of acid for each ore chunk, but with 150% prod. bonus you get 1.5 ore chunks per 10 acid (or 3 ore per 20 acid)

Well, of course. Productivity is simply a bar that fills as the normal production of the device runs. Once filled, the productivity bar generates one free output without consuming any input. Basically, uranium miners are an assembly machine that takes in sulfur and spits out ore, so the same thing applies.

[The Chaotic Coder]

Цитата допису Warlord:

Well, of course. Productivity is simply a bar that fills as the normal production of the device runs. Once filled, the productivity bar generates one free output without consuming any input. Basically, uranium miners are an assembly machine that takes in sulfur and spits out ore, so the same thing applies.

Exactly. I just thought I'd mention it in case anyone else here was unaware.

I'd be interested to know what's the highest mining productivity research anyone has gotten. It's infinite, so basically the question is "how high did you go before you got bored and decided to start a new game?"

[Warlord]

Цитата допису jchardin64:

Цитата допису Warlord:

Well, of course. Productivity is simply a bar that fills as the normal production of the device runs. Once filled, the productivity bar generates one free output without consuming any input. Basically, uranium miners are an assembly machine that takes in sulfur and spits out ore, so the same thing applies.

Exactly. I just thought I'd mention it in case anyone else here was unaware.

I'd be interested to know what's the highest mining productivity research anyone has gotten. It's infinite, so basically the question is "how high did you go before you got bored and decided to start a new game?"

Hah, well, my last game where I hadn't made a big mega-base dedicated to research, I got up to rank 50, or 100% more ore per cycle. Got bored, but it was still cheaper research than any of the others I used (robot speed, laser dmg, etc). Couldn't bring myself to raise it any more, I liked seeing the purple bar fill as fast as the red one in sync.

[BitterSwede]

That brings an interesting question to mind. The productivity bar must have a limit somehow. It's filled at a certain speed during operation based on productivity bonus.
So is there a moment where the filling would be near instant, at which point any further research into mining productivity won't actually yield any bonus anymore?

I need to make a modded chest that deletes stuff in it and then cheat in the research step by step until the productivity tab shows no difference anymore.

Also, since stone mines faster, would that limit (if it exists) be a lot lower for stone? Or does productivity bonus only work by mining power, and not mining speed?

Factorio Mythbusters is a GO!

[BitterSwede]

Up to level 948, a single miner produces 645-ish iron ore per minute.
Increased from 600-ish at level 875.

Above level 600 it seems like you need about 50 levels to increase productivity by 1 ore per second, increasing fast to about 75 levels per item per second above 800.

Finger is getting tired, but I'm pushing on!

Mining Productivity 1111 pushed that single electric miner to produce 60 ore per second. It's a landmark, but it is not the limit, it would seem. If done legit, it would've taken years to research, unless you could somehow provide thousands of each science pack per second.

I have concluded that if a limit exist, the amount of resources it would take to get there is far beyond what you would get back from it. Any theoretical limit is far beyond what my finger is capable of reaching when clicking that research button in cheat mode at a rate of 3 clicks per second (I thank my years as a bass player for that speed) without aching too much.

I'm calling it at 1500, where the lonesome miner produces 70 ore per second, at which point I do not think there's anything in the game that can remove the ore quick enough. That's a good enough limit for me.

To conclude, I would have to say no. I do not think there is a limit where productivity gets so fast that resource collection is instant, handing you infinite ore for free, and where any additional research would have near zero result. As it would seem, the higher you get, the longer between an actual increase in productivity, so it would seem like the near opposite might exist: A limit where there's an unfeasible amount of research between actual statistical increases.
In short, 2% of 100 is quite a lot, when compared to 2% of 3000.

[The Chaotic Coder]

Interesting research.

1 blue belt = 72 items per second (both lanes) or 36 on one side. If the miner were going directly to a provider chest, however, then the number of robots that could be picking up from it simultaneously is only limited by how many roboports you could pack around the ore patch to give them places to recharge. (and the speed of robots also increases with its own infinite research)

Of course, at some point it's no longer about getting more ore per second from each drill so much as making a finite ore patch last longer. Even if the drill can only run for 1/4 of a second at a time without its output jamming, the prod. bonus is still extending the number of ores it will produce before it runs out.

[BitterSwede]

Цитата допису jchardin64:

Interesting research.

1 blue belt = 72 items per second (both lanes) or 36 on one side. If the miner were going directly to a provider chest, however, then the number of robots that could be picking up from it simultaneously is only limited by how many roboports you could pack around the ore patch to give them places to recharge. (and the speed of robots also increases with its own infinite research)

Of course, at some point it's no longer about getting more ore per second from each drill so much as making a finite ore patch last longer. Even if the drill can only run for 1/4 of a second at a time without its output jamming, the prod. bonus is still extending the number of ores it will produce before it runs out.

As I said, these numbers are for a single miner. A miner at base speed (which is what I tested at) gives 0.525 ore/s, so 69.475 ore per second at level 1500 is for free. So if you can get to that research before you reach pension age, you'll get the ore patch to last roughly 140 times longer.

[Warlord]

Цитата допису BitterSwede:

Also, since stone mines faster, would that limit (if it exists) be a lot lower for stone? Or does productivity bonus only work by mining power, and not mining speed?

Stone mines faster? I never noticed that.

Anyway, if it does, productivity literally fills up based on the progress bar of the "recipe" runs. Take for example (if it could be done) productivity on a mk3 module assembler that takes 60 seconds to run. Over the time that the bar fills, it advances the pink productivity bar based on how much productivity is. 40% productivity will fill the pink bar 40% over the course of the red bar filling. Ugh, that didn't make much sense, did it....?

Anyway. So increasing the productivity only effects how much the bar fills PER recipe. A slower recipe or a fast recipe fills the bar the same amount per recipe, so a (possibly) faster stone miner won't provide any more or less stone than a miner on an iron patch.

As for your research about miner productivity... it should be a (mostly) linear increase. Another rank in the research shouldn't provide more free ore than the first rank in the research did.

Lastly, I find it interesting that miner research is linear in cost. Unlike all the rest, the cost goes up a fixed 1K more research than the last rank cost. Since the bonus is linear, that's fine. But why is the exponential cost increase for the rest so extreme? Double the cost of the last tier is absurd for what is also a linear increase in bonus.

[Hedning]

Цитата допису Warlord:

Stone mines faster? I never noticed that.

Lastly, I find it interesting that miner research is linear in cost. Unlike all the rest, the cost goes up a fixed 1K more research than the last rank cost. Since the bonus is linear, that's fine. But why is the exponential cost increase for the rest so extreme? Double the cost of the last tier is absurd for what is also a linear increase in bonus.

Hover over the stone and the other ores. You'll notice they have different "hardness".
As for the linear increase on mining productivity I think that is simply because it doesn't get absurd like the others. It just helps you spend less time on moving outposts. It doesn't make robots turn invisible as they aproach the speed of light.

[BitterSwede]

Цитата допису Hedning:

Цитата допису Warlord:

Stone mines faster? I never noticed that.

Lastly, I find it interesting that miner research is linear in cost. Unlike all the rest, the cost goes up a fixed 1K more research than the last rank cost. Since the bonus is linear, that's fine. But why is the exponential cost increase for the rest so extreme? Double the cost of the last tier is absurd for what is also a linear increase in bonus.

Hover over the stone and the other ores. You'll notice they have different "hardness".
As for the linear increase on mining productivity I think that is simply because it doesn't get absurd like the others. It just helps you spend less time on moving outposts. It doesn't make robots turn invisible as they aproach the speed of light.

Logistic bots, now with wormhole technology... But they still can't move further than the red line!

[piccolo255]

Цитата допису BitterSwede:

That brings an interesting question to mind. The productivity bar must have a limit somehow. It's filled at a certain speed during operation based on productivity bonus.
So is there a moment where the filling would be near instant, at which point any further research into mining productivity won't actually yield any bonus anymore?

I need to make a modded chest that deletes stuff in it and then cheat in the research step by step until the productivity tab shows no difference anymore.

Also, since stone mines faster, would that limit (if it exists) be a lot lower for stone? Or does productivity bonus only work by mining power, and not mining speed?

Factorio Mythbusters is a GO!

I remember reading answers for these, I think it was on the official forums by bobbingabout (author of Bob's mods)... IIRC, the conclusion at the time was that you can get at most one extra per tick, so 60 extra ores or whatever per second if running at 60UPS.

If the productivity is large enough that you shoul be producing more then one item per tick, any extra productivity over that one would get stuck in the buffer variable, and would eventually crash the game or something due to overflow. I don't know if the bug was fixed, and if it was how, so the behavior might be different now.

This can come up very easily in modded games, since you can fill up Mk5 miners with 8x raw productivity modules 8 (+40% each), and cover them in dozens and dozens of pure speed modules 8 (+160% each) set in Mk3 beacons (which have a large cover area and 100% transmission efficiency). Yeah, it's absurd :)

[bob152637485]

I realize this is an old thread, and that now as we are about to enter 0.18 a lot has changed(simplified mining speeds/hardness, as well as belt speeds), but the idea intrigued me. Reading the earlier posts, this reminded me of infinite series problems I did back in my second semester of calculus, specifically, does the series converge or diverge?

Then it dawned on me that solving this problem for a miner depositing directly on a belt is actually quite simple. Assuming that you have no modules in or around your miner, at 0 levels of mining productivity, it will spit out 0.5 ore per second for most ores, and 0.25 for uranium.

Every 100% additional bonus you get from mining productivity research raises the output per second to an additional 0.5 (0.25 for uranium) ore per second. This equates to every 10 levels of research made. A miner only puts on one side of a belt, so assuming you are using blue belts, that means the maximum throughput would be 22.5 ore/second. This means that at +4400% mining productivity (or +8900% for uranium), you would be fully saturating one side of a blue belt, and can not push out ore any faster. This means that at level 440 mining productivity, or level 890 for uranium, any more levels will not increase your ore/second on a belt based system. More levels can help, but only to the effect of making your ore veins last longer.

Now, an analysis of now all speed modules 3 and all productivity modules 3 would affect these numbers. Again, we aren't going to assume any use of beacons, since using beacons in a mining setup is not normally the most ideal, due to the center square of the beacon never getting mined.

Using 3 prod 3 modules, you save 3 levels of mining productivity, bringing normal ore down to level 437 and uranium down to 887. BUT let's now account for the massive speed penalty! the total speed penalty is -45%. This means that you need to divide the levels needed by the new speed of 0.55 the original speed. 437/0.55 is about 795, and 887/0.55 is about 1613! That is almost double the original levels needed for only a tiny amount longer that the ore vein will last.

Using 3 speed 3 modules, there is no productivity penalty, keeping the mining productivity levels at the original numbers of 440 and 890 for uranium. The speed bonus, though, is a whopping +150%! This means the new speed of the miner is 2.5 times that of the original speed. If we divide the original levels needed by our new speed, 440/2.5 is 176 and 890/2.5 is 356! That is less than half of the levels we originally needed!

With any of the three setups I examined, you only need 2 miners to fully saturate a blue belt, which is pretty cool in my opinion!

Regarding a robot based setup, there isn't really such a thing as maximum throughput, so I can't really do the math in this post. That said, I tried to document my steps to solve the levels needed for maximum throughput on a blue belt (22.5 items a second per side), so I would say, pick a number for the about of throughput you are looking for on a single miner for your specific robot configuration(you could have one or two miners per chest, and theoretically have infinite throughput if you have the needed robots!), and follow the same steps I did to see what the ideal level of research for you would be.

Hopefully this brings the thread back to life, because I am really looking forward to the feedback you all have to give on this analysis!