Factorio Forums

Posted: **Wed Sep 07, 2022 12:08 pm**

We all know that , if you place PM3 in mining drill, they produce less ore per second, and they cost more energy to do so that's like the definition of the productivity module the only good thing is that your patches last longer because they output slowly and you need more of them.

It's just terrible for UPS in all aspect, more of everything except the only useful thing : output.

It's only good AFTER the initial step in the production because it allows you to save up on the PREVIOUS step infrastructure size. Starting with prod modules in mining drill is an absolute non-sense if one is to talk about UPS. It is also explained by someone else in this thread viewtopic.php?t=53846 user mrvn must have forgotten.

If you were to beacon your mining setup, it's even worse to choose to put prod modules than speed modules or even efficiency as if you are to beacon all your miners you most likely have researched the mining productivity already and it will only keep growing as mining productivity is usually a goal and a measurement of progress, slowing down you mining drill by 30% to increase productivity by an insignificant amount compared to what mining prod research gives you overtime is non-sense.

It was not long explanation as it's pretty obvious : beaconned setup are better for UPS in general, but beaconned setup with prod modules are the worst kind of beaconned setup for mining drill.

If you decide to shift the goal to how much module you can produce instead of science it's up to your personnal choice how you want to have fun, if you divert 100% ressource to build module it's fine. It illustrate why it is absurd to say that module are build out of free ressources from modules, because one would still consume ressources even if not doing any science and instead being only focused on making more module and beacon, giving an efficiency of 0% toward the previous goal that was having a factory that make science for which you use modules.

The math part for the calculations of the "break even point" makes no sense to me, especially how you come up with the number of 11514 miner hours to break even and compare that "to PM3 modules in miners pay for themself in under 1 hours" therefore >11514 miners the mining bonus pays off quicker. The conclusion seem wrong, therefore i assume there is mistake in the part that is not explained properly in my eyes but maybe there are other mistake i missed, i haven't paid too much attention tbh.

Posted: **Wed Sep 07, 2022 1:43 pm**

One last reply because I actually have a correction to my numbers and them I'm sticking to the ignore.

mmmPI wrote: Wed Sep 07, 2022 12:08 pm It is also explained by someone else in this thread viewtopic.php?t=53846 user mrvn must have forgotten.

That thread is about a combined miner + smelter setup and uses beacons and speed modules in every miner and the argument was that mining bonuses avoids having to produce modules for every miner.

The comparison here is beacon + PM3 in miners vs. plain miners vs. mining bonus with plain miners. Neither case might be optimal but that are the three cases I compared.

mmmPI wrote: Wed Sep 07, 2022 12:08 pm The math part for the calculations of the "break even point" makes no sense to me, especially how you come up with the number of 11514 miner hours to break even and compare that "to PM3 modules in miners pay for themself in under 1 hours" therefore >11514 miners the mining bonus pays off quicker. The conclusion seem wrong, therefore i assume there is mistake in the part that is not explained properly in my eyes but maybe there are other mistake i missed, i haven't paid too much attention tbh.

So you didn't even look at it but I must be wrong. Your omniscience is showing. The numbers surprise me too but that is what math tells me.

The previous post showed that beacon + PM3 miners produce 1.98 times the ore as plain miners, some from the extra speed, some from PM3s. If you keep using 1 ore for science and divert the extra 0.98 ore to pay the resource debt for making the modules then it takes 1 hour to pay back all the resource used to making the modules.

The 11514 miner hours come about the same way. If you borrow all the resources to launch 9 rockets and research the 3 mining bonuses levels 4, 5 and 6 then it takes 1 miner 11514 hours to produce that make up the invested ore (10% bonus per level on 0.5 ore/s). Or 2 miners 5757 hours. Or to compare to the 1 hour the PM3 miners take to pay back the resource you need 11514 miners working 1 hour.

But I spot an error there now. It takes 1919 miner hours for one mining research level to pay off. Then the next level takes 2 twice that and the next 3 times that = 6 * 1919. But each level adds 10% bonus. So if you do all 3 levels together then the gained resources are 30% on 0.5 ore/s. So "only" 3838 miner hours, not 11514, a third. That's a bit more reasonable.

And one more thing I didn't mention: After that 1 hour is up to repay the resources then PM3 miners produce 1.98 times the base ore without any mining bonus. Researching mining bonus 1+2+3 is cheap and increases that to 2.28 times the base ore. While the science way miners with mining bonus level 6 produce only 1.6 times the base ore. So you would need a few more mining bonuses to be equal and they get more and more costly.

In the end the mining bonus (past level 3) is so expensive that it's cheaper to build a ton of modules unless you have huge numbers of miners. The speed modules have a huge impact because each beacon affects multiple miners and are the real reason for any gain.

The one thing I can undeniably say about PM3s in miners is that they make the ore field last longer. If you plan to play longer than an extra hour and finding more ore fields is a problem then PM3s+beacons are a gain over plain miners. But really start putting PM3s in everything else first. Top to bottom on the production pyramid.

Posted: **Wed Sep 07, 2022 2:06 pm**

mrvn wrote: Wed Sep 07, 2022 1:43 pm That thread is about a combined miner + smelter setup and uses beacons and speed modules in every miner and the argument was that mining bonuses avoids having to produce modules for every miner.

The comparison here is beacon + PM3 in miners vs. plain miners vs. mining bonus with plain miners. Neither case might be optimal but that are the three cases I compared.

Mining bonuses avoid having to produce module for every miner.

That's my point since the beginning maybe you can't admit or see it.

mrvn wrote: Wed Sep 07, 2022 1:43 pm So you didn't even look at it but I must be wrong. Your omniscience is showing. The numbers surprise me too but that is what math tells me.

mrvn wrote: Wed Sep 07, 2022 10:47 am Launching 9 rockets sounds way less than building 30000 PM3, 20000 SM3 and 10000 beacons. Did I make a mistake in the math or is that just one of those cases where intuition is wrong? I too thought mining bonus was better, surely.

I think you just made mistake in the part that your are not detailing, the one you said "let's helmod this", you may have used wrong inputs in helmods, and even if not the settings used are not clear, one cannot waste time to understand what was your particular mond thinking at the time.

Just detail how you count the ore in those 9 rockets and in those 30000PM3 and 20000 SM3 and 10000 beacons for fun

Or explain WHY only 10 000 beacons vs 30 000 PM3. like you only use 1 beacon per miner ?

Those are so filled with error that it's impossible to understand how you came up with your number that is wrong.

mrvn wrote: Wed Sep 07, 2022 1:43 pm The previous post showed that beacon + PM3 miners produce 1.98 times the ore as plain miners, some from the extra speed, some from PM3s. If you keep using 1 ore for science and divert the extra 0.98 ore to pay the resource debt for making the modules then it takes 1 hour to pay back all the resource used to making the modules.

Why not compare with the obviously better choice of placing SPEED module in miners or no miners in a beacon setup ?

Your example makes no sense comparing beaconned with PM3 setup with regular minar only helps showing that beaconned setup are strong.

Compare that with speed module and you have a clear better option in terms of UPS.

If your argument is to make patch last longer, then just don't put beacons in them, simple just put PM3 in all mining drill.

Placing PM3 and speed beacon is no good at UPS nor at making patch last longer.

mrvn wrote: Wed Sep 07, 2022 1:43 pm Or to compare to the 1 hour the PM3 miners take to pay back the resource you need 11514 miners working 1 hour.

Not sure why you keep saying PM3 in miners pay back in 1 hour, that is not true, you should explain the math more otherwise it makes no sense. Most likely coming from your error of counting 1 beacon per miner.

mrvn wrote: Wed Sep 07, 2022 1:43 pm But I spot an error there now. It takes 1919 miner hours for one mining research level to pay off. Then the next level takes 2 twice that and the next 3 times that = 6 * 1919. But each level adds 10% bonus. So if you do all 3 levels together then the gained resources are 30% on 0.5 ore/s. So "only" 3838 miner hours, not 11514, a third. That's a bit more reasonable.

oh really ? why not 1243240 mining hours ? you didn't explain how you come up with the number, it's not easy to spot error when the calcul is not written, only the wrong conclusion.

There are were many error in the original post, since after one correction it still contains several.

mrvn wrote: Wed Sep 07, 2022 1:43 pm In the end the mining bonus (past level 3) is so expensive that it's cheaper to build a ton of modules unless you have huge numbers of miners. The speed modules have a huge impact because each beacon affects multiple miners and are the real reason for any gain.

It make not sense to build tons of PM3 instead of researching mining bonus past level 3. The only part of the math detailed are wrong in several points.

If you are going to build 30000 PM3 , you will either take incredibly long, or have enough miner to make it more sensible to reseach productivity bonus. Do not listen to someone who is attempting to proove himself right, just do the math yourself.

Comparing beaconned PM3 vs plain miners only help showing how good speed module in beacon are despite the sub-optimal choice of placing PM3. None of the math presented support this, even if there was no error in the numbers the comparaison between beaconed PM3 and beaconned efficiency would clearly show that efficiency is better let alone speed or even nothing,

The argument of using Productivity module to makes patch longer makes no sense with beacon that forces you to move miner in built ore patch and make patch dry quicker compared to no-beacon.

Posted: **Wed Sep 07, 2022 2:44 pm**

mrvn wrote: Wed Sep 07, 2022 1:43 pm
mmmPI wrote: Wed Sep 07, 2022 12:08 pm The math part for the calculations of the "break even point" makes no sense to me, especially how you come up with the number of 11514 miner hours to break even and compare that "to PM3 modules in miners pay for themself in under 1 hours" therefore >11514 miners the mining bonus pays off quicker. The conclusion seem wrong, therefore i assume there is mistake in the part that is not explained properly in my eyes but maybe there are other mistake i missed, i haven't paid too much attention tbh.
So you didn't even look at it but I must be wrong. Your omniscience is showing. The numbers surprise me too but that is what math tells me.
mrvn wrote: Wed Sep 07, 2022 1:43 pm
But I spot an error there now. It takes 1919 miner hours for one mining research level to pay off. Then the next level takes 2 twice that and the next 3 times that = 6 * 1919. But each level adds 10% bonus. So if you do all 3 levels together then the gained resources are 30% on 0.5 ore/s. So "only" 3838 miner hours, not 11514, a third. That's a bit more reasonable.

Seriously ? You agree with me that you made a mistake in a part that you detailed only afterward in an edit ?

All i said was CORRECT, not only you did make a mistake in a part you didn't explained, but there was also a mistake in a part that was written, namely you said launching 9 rocket was more expensive that building 30000 PM3 and 20000 SM 3 and 10000 beacon.

The conclusion SEEM wrong and when looking a bit more, it makes to sense to even use those numbers in the first place.

You couldn't do the most basic verification yourself as it was confirming your ridiculous argument that it's better to build PM3 modules for each miners.

Don't be surprised people pay not much attention to those kind of post, only when they conclusion could be misleading for other player i would care to say your math are wrong and your conclusion is wrong and your attitude is dishonnest.

Posted: **Wed Sep 07, 2022 3:13 pm**

https://kirkmcdonald.github.io/calc.htm ... leLMq4wg8=

This is as far as i can tell the cost of 9000 science in raw ore, per minute.

Around 2 million iron+copper combined

https://kirkmcdonald.github.io/calc.htm ... hYldjUAQA=

This the cost of 2000 PM3 module.

Modules are already costing slightly more iron+copper combined than 2 million. and more in coal too.

One must be very dishonnest to claim that math says launching 9 rocket is more expensive than 30 000 PM3 + 20 000 SM3 + 10 000 beacon, making obvious math mistake doesn't excuse not verificating one bit the conclusion when according to one's word the conclusion seem wrong.

Posted: **Wed Sep 07, 2022 4:15 pm**

mmmPI wrote: Wed Sep 07, 2022 2:06 pm
mrvn wrote: Wed Sep 07, 2022 1:43 pm That thread is about a combined miner + smelter setup and uses beacons and speed modules in every miner and the argument was that mining bonuses avoids having to produce modules for every miner.

The comparison here is beacon + PM3 in miners vs. plain miners vs. mining bonus with plain miners. Neither case might be optimal but that are the three cases I compared.

Mining bonuses avoid having to produce module for every miner.
That's my point since the beginning maybe you can't admit or see it.

That's not a point but a claim. One the math seem to prove to be only valid for a certain number of miner upwards, a large number. And you don't have to shout, that doesn't make you any more right.

mmmPI wrote: Wed Sep 07, 2022 2:06 pm
mrvn wrote: Wed Sep 07, 2022 1:43 pm So you didn't even look at it but I must be wrong. Your omniscience is showing. The numbers surprise me too but that is what math tells me.

mrvn wrote: Wed Sep 07, 2022 10:47 am Launching 9 rockets sounds way less than building 30000 PM3, 20000 SM3 and 10000 beacons. Did I make a mistake in the math or is that just one of those cases where intuition is wrong? I too thought mining bonus was better, surely.
I think you just made mistake in the part that your are not detailing, the one you said "let's helmod this", you may have used wrong inputs in helmods, and even if not the settings used are not clear, one cannot waste time to understand what was your particular mond thinking at the time.

Just detail how you count the ore in those 9 rockets and in those 30000PM3 and 20000 SM3 and 10000 beacons for fun

Or explain WHY only 10 000 beacons vs 30 000 PM3. like you only use 1 beacon per miner ?

Those are so filled with error that it's impossible to understand how you came up with your number that is wrong.

Here are the helmod production lines if you want to check them yourself:

Miners + modules:

Mining Bonus:

And I did post a picture of the miner/beacon setup:

: miners-4.png (37.75 KiB) Viewed 3844 times

Same layout as with the dense miners except 2 miners diagonally replaced by beacons. 2 miners, 2 beacons. Yep. 1 beacon per miner.

And before you say anything I'm ignoring extra beacons or loss of speed at the outside. It's close enough for all but tiny ore fields.

The reason for the beacons was to maintain at least the ore output the plain miners had. I didn't want to resource starve the smelters. It also uses the same space so no need to find extra ore fields. It's a drop in replacement giving you more ore. Probably not the best setup or the cheapest but certainly one of the easiest to convert an existing (the dense miner setup) setup into.

mmmPI wrote: Wed Sep 07, 2022 2:06 pm
mrvn wrote: Wed Sep 07, 2022 1:43 pm The previous post showed that beacon + PM3 miners produce 1.98 times the ore as plain miners, some from the extra speed, some from PM3s. If you keep using 1 ore for science and divert the extra 0.98 ore to pay the resource debt for making the modules then it takes 1 hour to pay back all the resource used to making the modules.
Why not compare with the obviously better choice of placing SPEED module in miners or no miners in a beacon setup ?

Your example makes no sense comparing beaconned with PM3 setup with regular minar only helps showing that beaconned setup are strong.

Compare that with speed module and you have a clear better option in terms of UPS.

If your argument is to make patch last longer, then just don't put beacons in them, simple just put PM3 in all mining drill.

Placing PM3 and speed beacon is no good at UPS nor at making patch last longer.

mrvn wrote: Wed Sep 07, 2022 1:43 pm Or to compare to the 1 hour the PM3 miners take to pay back the resource you need 11514 miners working 1 hour.
Not sure why you keep saying PM3 in miners pay back in 1 hour, that is not true, you should explain the math more otherwise it makes no sense. Most likely coming from your error of counting 1 beacon per miner.

mrvn wrote: Wed Sep 07, 2022 1:43 pm But I spot an error there now. It takes 1919 miner hours for one mining research level to pay off. Then the next level takes 2 twice that and the next 3 times that = 6 * 1919. But each level adds 10% bonus. So if you do all 3 levels together then the gained resources are 30% on 0.5 ore/s. So "only" 3838 miner hours, not 11514, a third. That's a bit more reasonable.
oh really ? why not 1243240 mining hours ? you didn't explain how you come up with the number, it's not easy to spot error when the calcul is not written, only the wrong conclusion.

There are were many error in the original post, since after one correction it still contains several.

For each ore it's "miner hours = amount of ore needed / ore produce above base value by one miner" and then added together.

In the PM3 case the surplus is made up of the extra speed and productivity = 0.98 of the base value of 0.5 ore/s. In the research case it's made up of just the productivity increase = 30% of 0.5 ore/s.

mmmPI wrote: Wed Sep 07, 2022 2:06 pm
mrvn wrote: Wed Sep 07, 2022 1:43 pm In the end the mining bonus (past level 3) is so expensive that it's cheaper to build a ton of modules unless you have huge numbers of miners. The speed modules have a huge impact because each beacon affects multiple miners and are the real reason for any gain.
It make not sense to build tons of PM3 instead of researching mining bonus past level 3. The only part of the math detailed are wrong in several points.

You haven't pointed out any errors. We aren't all as as godlike as you.

mmmPI wrote: Wed Sep 07, 2022 2:06 pm If you are going to build 30000 PM3 , you will either take incredibly long, or have enough miner to make it more sensible to reseach productivity bonus. Do not listen to someone who is attempting to proove himself right, just do the math yourself.

Again you take a true fact and a unsubtantiated claim and make up a story and add a personal attack and implying you did better.

True: building 30000 PM3 modules will take incredible long.
Claim: "have enough miner to make it more sensible to reseach productivity bonus".
Attack: Do not listen to mrvn.
Implication: just do the math yourself ==> I did it and proofed him wrong.

The time needed to build the PM3s isn't limited by the miners. If you followed the argument you would have seen that the start is you have some miners with beacons and modules, which you build on loan. The math then answers the question how long it takes for the increase in ore production to repay that lone. And it's all ore, no oil is used as oil production is too variable.

The math actually shows that below a certain number of miners it is cheaper to build PM3, speed modules and beacons in the stated ratios while above a certain number doing mining research is cheaper. And the number of miners the math comes up with is 3838 miners for recovering the spend resources in one hour. After that the ore balance goes into the green. And that's 3838 miners in total across iron, copper, coal and stone.

Unfortunately the margin in this forum seem to be too small to add your proof. Or even to point out any errors I might have made. It's not impossible but if I did then point them out or shut up.

mmmPI wrote: Wed Sep 07, 2022 2:06 pm Comparing beaconned PM3 vs plain miners only help showing how good speed module in beacon are despite the sub-optimal choice of placing PM3. None of the math presented support this, even if there was no error in the numbers the comparaison between beaconed PM3 and beaconned efficiency would clearly show that efficiency is better let alone speed or even nothing,

The argument of using Productivity module to makes patch longer makes no sense with beacon that forces you to move miner in built ore patch and make patch dry quicker compared to no-beacon.

That might be true. None of the math was supposed to show that one way is the best way. Or to find the best or even just a better way.

The thing the math was supposed to show was:

1) Given each specific setup with the cost in ore shown for said setup how long does it take to produce a surplus equal to the cost to build/research the setup. I was expecting days or weeks but the result turned out to be 1h for PM3.

2) How many miners would it take for the mining bonus to produce a surplus equal to the cost to research said bonus in the same timeframe the first setup (PM3 + beacons) take, i.e. one hour.

You can also graph out the result:

: cost-graph.png (4.17 KiB) Viewed 3844 times

That's what I've shown. Below 3838 miners the PM3 modules repay the lone quicker while above the mining research (level 4+5+6) is quicker.

As a side I also noted that the PM3 setup produces 2.28 times base value while the mining research only produces 1.6 times base value. But that's the lack of speed modules in the research setup (which as you say just makes the ore get used up faster).

Posted: **Wed Sep 07, 2022 5:52 pm**

I note that you are not able to ignore me.

I note that

Mining bonuses avoid having to produce module for every miner.

could be amended with your conclusion :

only valid for a certain number of miner upwards, a large number.

But i'd rather say ALWAYS VALID past a certain number. maybe you understand this way why contrary to your sayings i'm highlighting it as it's general and not only true for one specific suboptimal setup deisgned to proove a point.

Here are the helmod production lines if you want to check them yourself:

I'm sure many people are interested thank you.

And I did post a picture of the miner/beacon setup:

This is not even a complete setup, it's so small i thought it was a mistake. I had not realized that was the setup you used for your conclusion lol as you put the other setup with 12 beacon with a much larger picture

And before you say anything I'm ignoring extra beacons or loss of speed at the outside. It's close enough for all but tiny ore fields.

That's good because then i'm free to add that you also ignore the yield difference between outside and inside of an ore patch and the fact that its size means some of the miners won't be functionning 100% of time either because they deplete faster, or because a blue belt can't output everything, but this is only true for large patch of ore, you choose yourself why your math do not apply in game .

The reason for the beacons was to maintain at least the ore output the plain miners had. I didn't want to resource starve the smelters. It also uses the same space so no need to find extra ore fields. It's a drop in replacement giving you more ore. Probably not the best setup or the cheapest but certainly one of the easiest to convert an existing (the dense miner setup) setup into.

Oh this means in total you go and shuffle the miner 3 time , once when you built it, then when you add the beacons, and finally to finish the left over pieces . That's even worse than i thought.

This setup you used as example to advise anyone to not research past mining prod 3 and instead build tons of modules ?

That's terribly wrong.

For each ore it's "miner hours = amount of ore needed / ore produce above base value by one miner" and then added together.

In the PM3 case the surplus is made up of the extra speed and productivity = 0.98 of the base value of 0.5 ore/s. In the research case it's made up of just the productivity increase = 30% of 0.5 ore/s.

This is not correct way to represent how much "free ore" the productivity research increase will give you on existing miner as it will makes them work less often for the same capped output in real game.

You haven't pointed out any errors. We aren't all as as godlike as you.

i don't see the relation between the first part that is your subjective interpretation and makes me wonder your definition of the word "ignore" that you claim you were doing already 3 times , and the second part that seem to imply i'm not like other human, which is also your subjective interpretation and makes me wonder your definition of godlike.

If you are going to build 30000 PM3 , you will either take incredibly long, or have enough miner to make it more sensible to reseach productivity bonus. Do not listen to someone who is attempting to proove himself right, just do the math yourself.

Again you take a true fact and a unsubtantiated claim and make up a story and add a personal attack and implying you did better.

I was saying that for myself, i should have been more careful in my wording and realize that you could think " someone who is attempting to proove himself right" was for you and a personnal attack at the same time.

I meant that doind the math for oneself is always better than listening to someone else, especially someone who claims several times putting someone else on ignore and not ignoring that person, that is not something that indicate the person is to be trusted since words and actions do not match.

The time needed to build the PM3s isn't limited by the miners. If you followed the argument you would have seen that the start is you have some miners with beacons and modules, which you build on loan. The math then answers the question how long it takes for the increase in ore production to repay that lone. And it's all ore, no oil is used as oil production is too variable.

I'm sorry i didn't realize you were trying to proove your point that there exist conditions in which it is possibe to misinterpret result in a way that could allow you to continue claiming that it makes sense to place PM3 in mining drill and to beacon it.

Too bad that in order to jump to a conclusion you make gross approximation and take a ridiculous setup that is terribly impractical in game and doesn't allow the math to hold true for the general conclusion you give.

That might be true. None of the math was supposed to show that one way is the best way. Or to find the best or even just a better way.

The thing the math was supposed to show was:

finding a way to continue arguing about placing PM3 in mining drill despite it being the worst place if they are placed, and not placing them being better in all things people are interested in apart from those that want to only build outpost to make more module for more outpost and no science.

That's what I've shown. Below 3838 miners the PM3 modules repay the lone quicker while above the mining research (level 4+5+6) is quicker.

That's only 1 setup, the worst, compared with nothing is that right ?

Given such particular bad setup, then it makes mining research inneficient ?

Congrats for finding it.

please don't give general conclusion about researching productivity is not as good as placing PM3, this is only the case in a setup that is poorly designed.

As a side I also noted that the PM3 setup produces 2.28 times base value while the mining research only produces 1.6 times base value. But that's the lack of speed modules in the research setup (which as you say just makes the ore get used up faster).

not sure what this 2.28 and 1.6 means to me it makes no sense i'm sorry, i would say the mining research gives you an amout of free ore that depend on the ore covered by your mining drill ,and this can be wayyyyyyy much more than anything, assume the loan is of 120343023823523523592359 miners. 10% extra ore on what is covered is much more than if there was only 1000 miners.

Not sure how you math the benefit of the research in such general way.

Though given that it's compared with a terrible setup, no conclusion could be held as general. Definitly in most cases and commonly used design of miners that do not sacrifice the productivity research effect by slowing down their mining drill with PM3 :

Mining productivity research bonus avoid having to produce module for every miner

Posted: **Wed Sep 07, 2022 6:28 pm**

https://kirkmcdonald.github.io/calc.htm ... VwX72/+Qc=

https://kirkmcdonald.github.io/calc.htm ... qNsakEAA==

Cost of 9000 science in raw ore is LESS than Cost of 2000 module 3 in raw ore.

One could argue that the second link would use 30% less ore from the ore patch. That is correct.

Cost of 9000 of all science packs in raw ore is LESS than Cost of 3000 module 3 in raw ore. no matter the 3O% less ore consumed the difference is to high with 3000 module.

But the claim from mrvn is wilder lol :

Launching 9 rockets sounds way less than building 30000 PM3, 20000 SM3 and 10000 beacons. Did I make a mistake in the math or is that just one of those cases where intuition is wrong? I too thought mining bonus was better, surely.

Posted: **Wed Sep 07, 2022 7:15 pm**

mmmPI wrote: Wed Sep 07, 2022 5:52 pm I note that you are not able to ignore me.

I tried it for one last discussion, this one, but it keeps being a fail.

PS: The ore numbers in your links are meaningless as you used the grey assemblers and didn't add productivity modules, for example iron gear wheels can be produced with 4x Prod3 mdoules for far less ore. Also comparing 9000 science packs to 2500 science packs makes it impossible to see if helmod and kirkmcdonald come to the same numbers.

Posted: **Wed Sep 07, 2022 7:37 pm**

That's a weird way of trying one last discussion to say 3 times you put the person on ignore after posting many times things containing mistake and asking if there were any and blaming that person for not pointing them even though that person, me , did pointed some mistakes and asked for more details not knowing if i should even try to show you why or where there could be things that make your intuition better than your math on that last argument.

If you want to understand your mistake it's roughly because the modules are giving free ore toward infinity the more you have them active. What you did purposedly or not was to make a setup that make use of that property and then substract an artifically long time of free productivity ore (that you start counting from the start as if it possible to have 3000 miner with PM3 but 0 mining productivity research) to the cost of building the module.

This doesn't represent what happens in game because the conditions are never met and because the math calcul is not correct as the substraction of the "free ore" is already canceled out if you consider that you start to measure with mining prod 3 vs mining drill with 3 prod module. Getting to mining prod 3 is cheaper than building the number of PM3 supposedly required to have the "free ore" generated via module amount for a significant enough number to makes you think that it's cheaper to build them than to do the first 3 level of research.

Then you made simplifications to how the productivity research is applied in real condition in game and it further biais the representaion of the cost it has compared to the cost of what you considered the initially loaned modules which in game would pay for the 3 level of research easily, and would start giving you 30% ore way before you earn the benefit of any single PM3 free ore.

All of this i felt had no other purpose than insisting on you idea to put PM3 in mining drill, which i feel is not representative on how PM3 are used by players in general or what accurate math would show, those are best used not in mining drill, and in mining it's also better not to use them and use speed instead or efficiency, or nothing, and just do the research as it's sort the goal of the game and how people measure their factory to compare and improve, and not the number of module although that could be a trend

Edit: That doesn't mean using PM3 in mining drill or pumpjacks is inherently "bad", it could be a choice, i'm playing currently a game with a mod that i know in the late game stages allows me to make coal from wood and grow tree and I'm spawning modules from an infinity chest in a creative-mode style. In this case i can totally decide to place PM3 in my coal mining drill. The goal would be to cap the ouput of the ore patch so that it last longer allowing me more time before needing the substitution, and it would be easier to replace with a sustainable source later as the ouput was capped. If i were to place speed module, i would oversize the factory processing the coal and then have a hard time doing the substitution with sustainable source that is slower and have bigger footprint than the concentrated carbon that is coal.

Posted: **Wed Sep 07, 2022 8:19 pm**

mrvn wrote: Wed Sep 07, 2022 7:15 pm
mmmPI wrote: Wed Sep 07, 2022 5:52 pm I note that you are not able to ignore me.
PS: The ore numbers in your links are meaningless as you used the grey assemblers and didn't add productivity modules, for example iron gear wheels can be produced with 4x Prod3 mdoules for far less ore. Also comparing 9000 science packs to 2500 science packs makes it impossible to see if helmod and kirkmcdonald come to the same numbers.

I'm sorry you missed the only link out of four where i didn't make a mistake .

I'm reposting the correct one and the updated one where the full productivity module is added onto the cost for the 9000 science.

https://kirkmcdonald.github.io/calc.htm ... 2ReET5BQ==

This one i already posted , but i repost, it's the cost of 2000 modules, which is way higher
https://kirkmcdonald.github.io/calc.htm ... qNsakEAA==

I took 9000 science because you said yourself it was cheaper to launch 9 rocket than to build 30 000 PM3 modules. and 20 000 speed 3 module and 10 000 beacon.

It's actually cheaper to launch 9 rockets and make all the 9000 of all science pack, rather than to build 2000 PM3. I'm not sure why you used yourself the 9 rocket as comparaison if your helmod is set for 2500 science and that doesn't allow you to compare.

There is a comparaison : https://kirkmcdonald.github.io/calc.htm ... jq2yeJR5Q/

Cost of 1750 science packs of all sort : More than what's required for reaching mining prod 3 research.

https://kirkmcdonald.github.io/calc.htm ... HRKo2xqQQA

Cost of 200 productivity modules 3

That would mean that ressources needed to reach the level 3 for mining productivity as if it was costing from LVL 1 all typed of sciences pack, is LESS compared to the expensive task of building 200 production 3 modules, or enough to accomodate 40 mining drill, if you consider they are all beaconned with 1 beacon and speed and prod module are the same cost so it's 5 modules per mining drill and the beacon is for free.

If you have more than 40 mining drill, just do the research or something

Posted: **Wed Sep 07, 2022 10:03 pm**

I wouldn't put prod modules in miners at all. Rather some efficiency modules 1's to reduce power usage and pollution.
The mining productivity research is so powerful already, which adds productivity for no speed penalty.
Even if a patch last some little time more, it's not worth IMO

I mean, that leads to more meta discussion, what is the goal of the factory. If you just want to saturate all possible modules positions, you can add them also to miners, which gives a minor bonus. But it has quite little impact.

Posted: **Wed Sep 07, 2022 10:45 pm**

jodokus31 wrote: Wed Sep 07, 2022 10:03 pm I wouldn't put prod modules in miners at all. Rather some efficiency modules 1's to reduce power usage and pollution.
The mining productivity research is so powerful already, which adds productivity for no speed penalty.
Even if a patch last some little time more, it's not worth IMO

I mean, that leads to more meta discussion, what is the goal of the factory. If you just want to saturate all possible modules positions, you can add them also to miners, which gives a minor bonus. But it has quite little impact.

Totally, it's the least useful place to add modules and, if at all, should be the last place you put modules. That was never the question, at least to me.

In my mind the question was: How long would it actually take to pay off adding the modules? Because everyone claims it has so little effect that it will never pay off or take weeks of play time. Unless I made a mistake in the math it's just an hour, not days or weeks.

Posted: **Wed Sep 07, 2022 10:57 pm**

Obviously there is an error as the time it takes for a PM3 unfortunatly placed in a mining drill to pay for itself depend on the speed of the mining drill, which itself depend on the amount of beacon surrounding it.

This means 1 hour is just a random result based on 1 setup which is not even possible to verify as there is no way to know how mrvn decided to count the price in ore of 1 module despite being asked several time.

Furthermore , the reduction in speed of the PM3 module impact negatively the mining drill output, losing some of "free ore" that would have come with the existing mining productivity bonus in favor of some less "free ore" from the module which also is depending on enough parameter than saying 1 hour makes little sense.

How much time is required for a module to pay for itself is based on amount of extra ressources generated from consumed ressources per time. This means that the mining drill producing 0.5 ore per second is one of the worst contender to be accomodated with production module. For example making an electronic circuit require 3 copper cable and 1 iron plate and has base speed of 0.5 second too, this means each module placed in an assembly machines producing green circuit produces an extra 10% based on 1.5 copper + 1 iron per cycle.

An assembling machine 3 has a base crafting speed of 1.25 which means the receipe that is 1 second will be made 25% faster than a mining drill takes time to produce just 1 ore. The receipe itself representing 250% more material on which the productivity module is effective and being only 0.5 second in the case of electronic circuit.

If you factor in the speed reduction it doesn't help, you have -60% for the assembly and -45% for the mining drill. this means 1.25x0.4=0.5 crafting speed for the assembly vs 0.5x0.55= 0.275 ore per second for the mining drill on which the PM3 module will be effective. This means in 1 second the assembly does 1 cycle, meaning consuming 3 copper cable and 1 iron plate, or 2.5 ore on which you gain 40% so 1 extra ore for the ressources consumed in 1 second.

While the mining drill in 1 second will consume 0.275 ore from the ore patch , on which you gain 30% extra for a total of 0.3575 ore, so an extra 0.0825 extra ore for the ressources consumed in 1 second.

This means it appear that the benefit for placing 4 modules in an assembly machine producing electronic circuit is more than 12 times the one that you get from placing 3 productivity module in a mining drill. The benefit per module being calculated doing (1/4)/ (0.0825/3) = 0.25/0.0275 = 9 times. roughly.

0.0825 ore per second is the benefit or 3 PM3 in a mining drill without speed beacon that would increase it or productivity bonus research lost due to slower speed that would need to be substracted.

if you count 4000 ore for a module, it will take (4000x3)/0.0825 second to break even, or also 145K second, or 40 hours for 3 modules to pay for themselves.

edit: not counting the wasted ressources in additionnal energy production
nor factoring in the cost of the potential beacons and speed modules that one would use depending on which setup is used in combinaison with the PM3 in the mining drill to try having them pay for themselves faster because to me it make little sense to use both beacon and speed module, like braking and accelerating at the same time with a car.

Posted: **Wed Sep 07, 2022 11:12 pm**

mrvn wrote: Wed Sep 07, 2022 10:45 pm Totally, it's the least useful place to add modules and, if at all, should be the last place you put modules. That was never the question, at least to me.

mrvn wrote: Mon Sep 05, 2022 12:24 pm People keep talking about the cost of prod modules. But consider this:

Put in miners they produces 20% more ore (and oil although speed is probably better there). Imagine taking those 20% and using it exclusively to produce the infrastructure to make more modules and the modules themself. That will grow exponentially. Put prod modules in electric furnaces for 20% more plates for a total of 44% gain. The exponential growth has gotten more than twice as steep. Add it to copper wire and steal and many others and the gain just keeps growing.

Other than the initial investment to get the process started it wouldn't cost you anything, the modules produce themself. Well, except for the time for you to set it up and expand.

It was not clear from your post thanks for clarifying what you meant again.

"people keep talking", "everyone claims" , "the strawman did it"

Posted: **Thu Sep 08, 2022 12:00 am**

Ahh, found an error in the initial math for the 2 miners + 2 beacons setup. It's not 2 hours but 4 hours 16 minutes to break even from just the 30% free ore. So ~2 hours with all the ore gained from PM3 + beacons above the original output from 4 miners. I wonder if anyone else can spot it.

That brings the cross over point for mining research down to to 1919 miners I think. But I really should redo the calculations with mining research 1+2+3 included. I doubt anyone will skip those.

Posted: **Thu Sep 08, 2022 12:56 am**

Let's try to move the discussion away from Kindergarden back to something fact based...

According to https://kirkmcdonald.github.io/calc.htm ... tASIURsgoA a single productivity module 3 requires 1200 coal + copper + iron, when using prod 3 along all processing steps (which, given everyone seems to be in agreement that miners should be the last thing to get modules, seems a sensible requirement).

To repay itself, a single prod 3 module thus needs to produce 1200 extra ore, which with 10% productivity, will happen after 1200 / 10% = 12000 mining cycles. (Or 12000 / 0,5/s = 6h 40min at speed 1.) (When using multiple prod modules, the modules will still repay themselves after 12000 mining cycles since cost increase and productivity increase will cancel out.)

Since prod modules have a speed penalty, to make meaningful comparisons to the mining productivity research, we need to bring the speed back up to 1.
Possible arrangements are e.g.
- 2 prod 3 + 1 speed 2 without beacons: 2614,57 / 20% / 0,5/s ~ 7h 16min https://kirkmcdonald.github.io/calc.htm ... FWgk0BAA==
- 3 prod 3 in the miner + 3 speed 2 in 2 beacons, assuming you can hit 8 miners with each pair of beacons: 29683,745 / 8 / 30% / 0,5/s ~ 6h 52min https://kirkmcdonald.github.io/calc.htm ... mo65GEAA==

Mining productivity 4 research takes 281498,667 ore to research https://kirkmcdonald.github.io/calc.htm ... OZxDztCw==. In a single miner, it'd take 281498,667 / 10% / 0,5s ~ 1564h ~ 65d to repay itself. When comparing to the 8 miner + 2 beacons setup above, the brake even point is at ~227,6 miners. Taking mining productivity 4 through 6 instead, the brake even point doubles to ~455,2 miners.

Posted: **Thu Sep 08, 2022 5:58 am**

Nidan wrote: Thu Sep 08, 2022 12:56 am Let's try to move the discussion away from Kindergarden back to something fact based...

According to https://kirkmcdonald.github.io/calc.htm ... tASIURsgoA a single productivity module 3 requires 1200 coal + copper + iron, when using prod 3 along all processing steps (which, given everyone seems to be in agreement that miners should be the last thing to get modules, seems a sensible requirement).

To repay itself, a single prod 3 module thus needs to produce 1200 extra ore, which with 10% productivity, will happen after 1200 / 10% = 12000 mining cycles. (Or 12000 / 0,5/s = 6h 40min at speed 1.) (When using multiple prod modules, the modules will still repay themselves after 12000 mining cycles since cost increase and productivity increase will cancel out.)

It's even worse than that since the PM3 module are slowing down the mining drill by 45% so the 12000 cycle should be divided by less than by 0.5 second no ?

Edit : "at speed 1" nvm, i disagree on which speed is to be used to indicate but i agree with the reasonning

if you count 4000 ore for a module, it will take (4000x3)/0.0825 second to break even, or also 145K second, or 40 hours for 3 modules to pay for themselves.

If you count 1000 ore for a module it will take (1000x3)/0.0825 second; or 10 hours which is the same number as 6h40 considering the 45% speed reduction on the mining drill.

The 0.0825 is to me [0.5x0.55]x0.3 which is the base speed x0.55 to represent the actual speed of the mining drill when slowed by 3 PM3, multiplied by 0.3 is to only account for the extra 30% productivity generated by module. Or 0.0825 extra ore generated per second per mining drill equiped with 3 PM3.

It's worse than that if you already had mining productivity researched i feel, because them reducing the speed of the mining drill by 45% makes you lose on more of that "free ore" from the research per miner.

However i agree with your method of calculating things. There are some little difference on the parameter, but given that you explain the steps it is understandable why : you consider 6h40 with mining speed 1, i don't.

I would encompass the cost of the extra speed 3 module required to increase the speed of the mining drill back to 1 or > into what needs to be pay back for by the PM3 modules, so for a ratio of 1:1 PM3 speed 3, i would could 2400 ore instead of 1200 ore that needs to be repay.

Those little difference in perspective again, do no change that i agree with the way of doing the math contrary to what i read previously from other user

Posted: **Thu Sep 08, 2022 7:29 am**

mmmPI wrote: Thu Sep 08, 2022 5:58 am
Nidan wrote: Thu Sep 08, 2022 12:56 am Let's try to move the discussion away from Kindergarden back to something fact based...

According to https://kirkmcdonald.github.io/calc.htm ... tASIURsgoA a single productivity module 3 requires 1200 coal + copper + iron, when using prod 3 along all processing steps (which, given everyone seems to be in agreement that miners should be the last thing to get modules, seems a sensible requirement).

To repay itself, a single prod 3 module thus needs to produce 1200 extra ore, which with 10% productivity, will happen after 1200 / 10% = 12000 mining cycles. (Or 12000 / 0,5/s = 6h 40min at speed 1.) (When using multiple prod modules, the modules will still repay themselves after 12000 mining cycles since cost increase and productivity increase will cancel out.)
It's even worse than that since the PM3 module are slowing down the mining drill by 45% so the 12000 cycle should be divided by less than by 0.5 second no ?

Edit : "at speed 1" nvm, i disagree on which speed is to be used to indicate but i agree with the reasonning

Note that the paragraph you quoted only considers the cost of a single module in relation to the productivity bonus. You could ignore the conversion to time and instead keep it at "mining operations needed to repay itself"; wouldn't change anything for the comparison as the conversion to time cancels out anyways.
Accounting for multiple modules and the speed difference happens further down.

Posted: **Thu Sep 08, 2022 7:54 am**

Nidan wrote: Thu Sep 08, 2022 7:29 am
mmmPI wrote: Thu Sep 08, 2022 5:58 am
Nidan wrote: Thu Sep 08, 2022 12:56 am Let's try to move the discussion away from Kindergarden back to something fact based...

According to https://kirkmcdonald.github.io/calc.htm ... tASIURsgoA a single productivity module 3 requires 1200 coal + copper + iron, when using prod 3 along all processing steps (which, given everyone seems to be in agreement that miners should be the last thing to get modules, seems a sensible requirement).

To repay itself, a single prod 3 module thus needs to produce 1200 extra ore, which with 10% productivity, will happen after 1200 / 10% = 12000 mining cycles. (Or 12000 / 0,5/s = 6h 40min at speed 1.) (When using multiple prod modules, the modules will still repay themselves after 12000 mining cycles since cost increase and productivity increase will cancel out.)
It's even worse than that since the PM3 module are slowing down the mining drill by 45% so the 12000 cycle should be divided by less than by 0.5 second no ?

Edit : "at speed 1" nvm, i disagree on which speed is to be used to indicate but i agree with the reasonning
Note that the paragraph you quoted only considers the cost of a single module in relation to the productivity bonus. You could ignore the conversion to time and instead keep it at "mining operations needed to repay itself"; wouldn't change anything for the comparison as the conversion to time cancels out anyways.
Accounting for multiple modules and the speed difference happens further down.

I agree with the cost of a single module in relation to the productivity bonus, i also agree with one of your previous conclusion where if 3 PM3 takes 6-7 hours to repay for themselves, then 1 PM3 will also take 6-7 hour to repay for itself.

I agree with the accounting for multiple module and the speed difference happening further down.

I was unconfortable with expressing the time 1 PM3 needs to pay back for itself when you say "at speed 1", because that's not accounting for one of the negative aspect of the PM3 on mining drill, using 1 or 2 or 3 PM3 in a mining drill changes it speed and therefore it's understandable to find an equivelent "at speed 1" which in games needs to be obtained adding additionnal ressources, that you expressed later on, could be spent on different setups. It makes more sense now. The 6h40min is lower than what would occur in game if you were to place 3PM3 in a mining drill but it's meant as a mathematical intermediate value that is made more precise with another calculation adding the cost of ressources needed to reach that speed of 1 again or another higher speed depending on different example of setup. It's all good

The time you mention do not account for the additionnal energy consumption if one wanted to fully math out things

Which using PM3 would also require more ore to build the solar pannel and accumulator to keep the beacon working which is a hidden cost that lenghen the process of repayment even worse if you don't use solar and /or have many depleted miner as in real game condition a % of them would be at any given time unless perfect condition. It's understandable that this part cannot be precisely generalized and is not math out, but it could come with a disclaimer then :p

Factorio Forums

why prod modules are important

Re: why prod modules are important

Re: why prod modules are important

Re: why prod modules are important

Mining bonuses avoid having to produce module for every miner.

Or explain WHY only 10 000 beacons vs 30 000 PM3. like you only use 1 beacon per miner ?

Re: why prod modules are important

Re: why prod modules are important

Re: why prod modules are important

Mining bonuses avoid having to produce module for every miner.

Or explain WHY only 10 000 beacons vs 30 000 PM3. like you only use 1 beacon per miner ?

Re: why prod modules are important

Mining bonuses avoid having to produce module for every miner.

Mining productivity research bonus avoid having to produce module for every miner

Re: why prod modules are important

Re: why prod modules are important

Re: why prod modules are important

Re: why prod modules are important

Re: why prod modules are important

Re: why prod modules are important

Re: why prod modules are important

Re: why prod modules are important

Re: why prod modules are important

Re: why prod modules are important

Re: why prod modules are important

Re: why prod modules are important

Re: why prod modules are important