Factorio

Factorio

Lihat Statistik:
Topik ini telah dikunci
Are solar panels worth building at all?
They produce so little power each and the price compared to even just a steam engine is way higher, so why use them? By the time I started running into problems with my starting steam set-up I would have had to produce hundreds of them to fix it or just work on nuclear power. So when am I meant to use solar panels? Are they ever worth the cost? New player here just trying to learn the game.
< >
Menampilkan 256-270 dari 278 komentar
Hedning 16 Apr 2023 @ 5:05pm 
Diposting pertama kali oleh Chindraba:
Diposting pertama kali oleh Hedning:
I can't find who suggested to not use landfill. Did anyone actually do that?
I'm sort of the responsible party there, even though I didn't say not to use landfill. What I said amounted to if you're going to use landfill in the calculation of cost for nuclear, use it in the calculation of cost for solar.
Indeed that is not saying not to use landfill, and a good point, though I would say use landfill but don't include it in any calculation for any of them since landfill is made so different from anything else it will be really misleading to include it. It's 20 ore on a 0.5 second recipe. Making one such ore equivalent to an ore used in a chain ending with a red circuit is beyond unfair. It's outright dishonest.

Diposting pertama kali oleh Chindraba:
Diposting pertama kali oleh Vyndicu:
The community uses Steam Storage Tanks with Nuclear Energy.
Of this I was unsure. Most of the folks I've dealt with didn't use tanks as storage. A few use a couple as a monitoring system to control the flow of fuel, turning off the reactors when not needed. I used the factorio.school website and checked out the first 2 pages of nuclear prints. 59% (70 of 118) had at least one tank, though about 20 of them had less than 20 so I'm not sure if they were part of a control system or a miniature reserve "in case".
Doesn't matter how popular something is, especially if it is likely based on a false belief. Anyone who never calculates it or runs nuclear for a lot of time, enough to realize the truth, is likely to believe the fuel is expensive. I did that. I was surprised when I learned the truth. I have seen lots of people be surprised. I have rarely seen anyone defend saving nuclear fuel after having been shown the calculations, Vyndicu being an obvious exception.
Hedning 16 Apr 2023 @ 5:06pm 
Diposting pertama kali oleh Khagan:
Every little helps :)
No, because starting a 10h timer an hour before I start my 1h 22 minute timer isn't going to do anything and it's not worth the effort. It's not a temporary solution, it is a solution that is obsoleted way before it can do anything of consequence.

Diposting pertama kali oleh Khagan:
Of course, when I start processing uranium I don't generally have beacons yet
This is why you don't try to rush it. Boilers are king early game and your 3000 (?) red circuits can be better used for other things than nuclear.
Diposting pertama kali oleh Vyndicu:
Diposting pertama kali oleh knighttemplar1960:
Like this:

https://steamcommunity.com/sharedfiles/filedetails/?id=2963361280

You can tell by the minimap that my personal roboport is no where near the construction zone. That's a spidertron building crew. They require no infrastructure or power from the power grid. They can be moved by remote control, the trunk holds more than a cargo wagon, and they can be set to request and autotrash things when they are in a logistic supply area just like you can.

Yeah, fair enough. Nothing wrong with a Spidertron construction crew. They still have to walk up to the water's edge and wait for it to be done.
Yeah but 10 of them with 3 personal roboports each makes short work of the project. Its like having a 15 cargo wagon construction train that doesn't require your presence or need rails.
Diposting pertama kali oleh Chindraba:

My nuclear plant (partially) and the dance it plays with solar power all day long.
https://steamcommunity.com/sharedfiles/filedetails/?id=2963445467
And, more stupid uranium than I know what to do with.
https://steamcommunity.com/sharedfiles/filedetails/?id=2963447571

I detect over building again :steamhappy::steamthumbsup:

I don't use steam storage tanks. They are redundant. There is no ebb and flow with nuclear like there is with solar. All you have to have is more capacity than you use and some way to handle power spikes if you have trimmed it too close.

I don't use steam storage tanks. Any time I need an extra 1.1 GW I just drop another nuclear plant. The solar panel and accumulators are only in the build to power the water pumps in case I can't build right on the edge of the water. The 4 pictured here are all I need to expand from first launch to mega base. I have enough material on hand to build 2 of these plants.

Since fresh nuclear fuel in = spent nuclear fuel out I can use one train with the cargo wagons partitioned 50/50 to do both at the same time.

https://steamcommunity.com/sharedfiles/filedetails/?id=2963529525
https://steamcommunity.com/sharedfiles/filedetails/?id=2963529535
Terakhir diedit oleh knighttemplar1960; 16 Apr 2023 @ 10:10pm
Chindraba 17 Apr 2023 @ 1:59am 
Diposting pertama kali oleh knighttemplar1960:
I detect over building again :steamhappy::steamthumbsup:
Not really surprising, though. It's just in my blood.

Diposting pertama kali oleh knighttemplar1960:
https://steamcommunity.com/sharedfiles/filedetails/?id=2963529525
I do have a personal request, thought. Would you, please, turn off the grid for most of your screenshots. They tend to make it harder to enjoy the show. Sometimes, like for small prints where spacing it of interests, like turret range or wall spacing, it's helpful, but in the main it's actually harder to grasp the full thing.
And this thread has devolved into a debate on whether solar or nuclear power is technically cheaper if everything is already long since unlocked and as the base size approaches infinity.
Yes you can go the entire game without building solar panels (and even launch the rocket using coal power) but as with most of the different technologies in the game, each power source has their own intended use case as players progress through the game and different pros and cons allowing for different ways to play the game for those that feel so inclined.
Terakhir diedit oleh SolarCaveman; 17 Apr 2023 @ 5:13am
Diposting pertama kali oleh Chindraba:
Diposting pertama kali oleh knighttemplar1960:
I detect over building again :steamhappy::steamthumbsup:
Not really surprising, though. It's just in my blood.

Diposting pertama kali oleh knighttemplar1960:
https://steamcommunity.com/sharedfiles/filedetails/?id=2963529525
I do have a personal request, thought. Would you, please, turn off the grid for most of your screenshots. They tend to make it harder to enjoy the show. Sometimes, like for small prints where spacing it of interests, like turret range or wall spacing, it's helpful, but in the main it's actually harder to grasp the full thing.
A lot of my screen shots are all ready taken and since I'm visually impaired having the grid on helps with placement and building. I'll try to remember to turn it off for new screen shots going forward.
Chindraba 17 Apr 2023 @ 7:20am 
Diposting pertama kali oleh knighttemplar1960:
A lot of my screen shots are all ready taken and since I'm visually impaired having the grid on helps with placement and building. I'll try to remember to turn it off for new screen shots going forward.

I'm not impaired, and I use the grid often enough myself. I've even set it to be on in debug mode (if I have the options named right) so that F5 toggles is on/off for me. For some reason in screenshots it makes things worse, even though it makes things better in the live version.

having the grid on helps with placement
While it only makes a difference if avoiding off-by-one problems, I've found that I can make placement easier if I have one rail track somewhere in the print. Of course it only works for prints I make myself and connect with other prints also "tagged" that way. It's the same as using a grid snap of 2x2 but won't interfere with a relative snap when I want it for running tile placement. I just have to remember to keep rails out of the ♥♥♥ inventories unless I'm doing rails. That way the rail isn't built and I can delete the ghost after the print's been planted.

The other trick I use when making prints that go together, like pieces of a plant in smaller prints, is to keep one thing from the next-door print with some size to it. Roboports are great if they are close, or big poles and substations. Nothing with a recipe, though. I use that thing as a "key" to align the new print with the old print. If I wiggle the mouse while clicking then it says it can't place the print because a roboport, or whatever, is in the way. The big blue square even lets me place prints from zoomed-out map view.
Terakhir diedit oleh Chindraba; 17 Apr 2023 @ 7:20am
Vyndicu 17 Apr 2023 @ 9:41am 
Diposting pertama kali oleh Chindraba:
Diposting pertama kali oleh Vyndicu:
You are right. It is a non-trivial substantive space to turn over to solar energy. It won't be an issue since you always want to push outward to grow the factory which comes with significant expansion of the defense wall perimeter length and, consequently, more unused megabase's interior space.

As the one doing most of the intense math thus far, I'll issue the challenge directly. Simple enough to do. How many landfill tiles are currently under your entire collection of solar panels and what percentage of the space used by solar is that? Not sure? Easy solution:

For each solar array make a blueprint of it. In the first GUI that pops up after dragging the print maker tool over the field, select the "Tiles" box in the "Filters" section. If you need the size you can check the "Snap to grid" check box and read the Width and Height sizes in their own boxes. Clicking the normal "X" close button will avoid cluttering you library with temporary prints. If, as is highly likely, the count of landfill is reported in K, just add all the prints up by using the K for each. 64K landfill figured as 64000, even if the actual was 64850. Once all the sizes are added and the landfill are added, the difference of 1 or 2 K won't matter much in the percentages.

My one solar array is two sections, 392x490 and 196x882 with a total of 64K landfill. Total is 361424 tiles. 64K is a mere 17.7%. And, when building that I did try to maximize the dry land area. It's only making about a 44% of my total solar output but the rest is placed in pockets here and there and I'd never figure the usage without some extra tools. If I did, it'd only make the numbers worse as nearly half is going to be on water. (I tend to be neglectful of my landfill usage. I have about 15 wagon-loads waiting to use and the ability to produce 3.3K/min.)

If you are struggling to find unused solid ground, which is not a typical vanilla Factorio experience as I only have encountered heavy-water map very infrequently. Even then, I would have to either use the island pre-set or artificially crank up the number of water tiles (especially the coverage slider) at map generation to get anywhere near the experience you have had.

I have already explained this previously, and asking me to do the math or check out my megabase while ignoring unlimited solid land as part of the normal Factory Expansions is not helping you to make your case that I should consider landfill costs for solar energy.

As an aside, I don't have any savegame currently with a megabase power grid, and the last time I did one was several years ago. I believe it was at least several hundred chunks of solar energy, if not more. I do not know the megabase's total size. No substantive lakes or oceans exist in a way that they block my need for solar energy growth.


I just rerolled several maps at the unmodified Vanilla "DEFAULT," Almost all had approximately ~20% to ~35 % water tiles in the map preview area.

None of them would have required players to build solar energy over water tiles unless the player lowered the scale slider (distance between lakes) and bumped the coverage slider (size) of water on map generation.


Diposting pertama kali oleh Chindraba:
Diposting pertama kali oleh Vyndicu:
Those unnecessary bells and whistles are there to optimize the Nuclear Reactors. Technically, it is possible to build them without those bells and whistles, but then they won't run as well as they could (no-circuit-inserter putting in Nuclear Fuel cells at a different time and wasting neighbor bonus).

Those bells and whistles do not optimize the production of power, are completely non-essential. Building without them is not only "technically" possible, it is much easier to do and get it right. My 56-core reactor (28-core print tiled once) has run continually since built with never once a loss of 1kW from a missed neighbor bonus, or lack of steam tanks, or anything else. (Print available if you'd like to try it yourself.) It don't even use fancy inserters, yellow inserters all the way.

How many Reactors you have built and successfully run without bells and whistles don't matter. There always exists a better solution or blueprint to get extra energy.

What you call "fancy inserters" is anything but expensive and ensures that all reactors start simultaneously.

Red Signal on the power pole from green wires is 1 to indicate that a steam fluid storage out of view is full, and inserters should not swing. They will all swing at once as soon the signal changes to 0.

https://steamcommunity.com/sharedfiles/filedetails/?id=2963688014

It costs a few copper wires, an arithmetic combinator, and a storage tank. I have no idea what you meant by "fancy" inserters.

At higher Nuclear Energy levels, more storage tanks will work better since filling and emptying a single storage fluid is too short.


Diposting pertama kali oleh Chindraba:
Diposting pertama kali oleh Vyndicu:
The community uses Steam Storage Tanks with Nuclear Energy.

Of this I was unsure. Most of the folks I've dealt with didn't use tanks as storage. A few use a couple as a monitoring system to control the flow of fuel, turning off the reactors when not needed. I used the factorio.school website and checked out the first 2 pages of nuclear prints. 59% (70 of 118) had at least one tank, though about 20 of them had less than 20 so I'm not sure if they were part of a control system or a miniature reserve "in case".

I tested the blueprint with all the bells and whistles a few posts ago to determine the equalized ore cost.

That blueprint was not my own and was submitted relatively recently to a public Factorio blueprint repository website. Using someone else's blueprint was to remove any chance of my bias influencing calculating the equalized ore cost.


Here is one that explicitly mentions logic circuits to reduce the over-usage of Uranium Fuel Cell.

https://www.factorio.school/view/-NROpFqQ-21tVUfLcere

Another one.

https://www.factorio.school/view/-N7zPRrt-EHIdkSSJnm6

Another blueprint with dozens of arithmetic combinators included for checking fluid levels.

https://www.factorio.school/view/-MvgARHrqaUa58MBhvEt

Here is another one that doesn't explicitly say this, but it is likely using a logic-based inserter, as I did.

https://www.factorio.school/view/-N6-j5RQYB-0dJEdLf-V

Here is one blueprint the original poster fixed many years ago for releasing the Factorio 0.17 patch. The original blueprint had 29 likes in the second link below.

https://factorioprints.com/view/-LGzh1rL3S0dU-eJHfcr

https://www.factorio.school/view/-La0aDEtd6Y4XlrpCfgW


I would not be surprised if you missed those blueprints due to needing to know what to look for in the first place. The search for both Factorio Blueprint and Factorio School could be better, to be honest.


Diposting pertama kali oleh Chindraba:
Diposting pertama kali oleh Vyndicu:
My primary rationale is not to squander Nuclear Fuel Cells; others may share this or have a different rationale. I use circuit-enabled inserters and a Nuclear Steam Storage tank to check levels. Of course, it is also possible to only use Storage Tanks alone without any other benefit besides storing excess energy.

It is also possible, practical, simple, and common to use nuclear without the "benefit" of any storage tanks.

Consider this scenario and I build a Nuclear Power to produce an unspecific number of GW while only needing less. What happens to the unused excess power?

Or is it better to regulate/smartly my Nuclear power only to consume what is necessary to meet the power grid demands at that moment? A smarter Nuclear power at a miscellaneous cost of a few copper wires, a single arithmetic combinator, and extra storage tanks?


Diposting pertama kali oleh Chindraba:
*snipped*

None of that changes anything I didn't know before the Original poster created the thread.

I have used all three power generations on a singular power grid in the past as well.

I just arrived at a very different conclusion than others regarding solar energy, no more, no less.
Chindraba 17 Apr 2023 @ 12:06pm 
Diposting pertama kali oleh Vyndicu:
If you are struggling to find unused solid ground, which is not a typical vanilla Factorio experience as I only have encountered heavy-water map very infrequently. Even then, I would have to either use the island pre-set or artificially crank up the number of water tiles (especially the coverage slider) at map generation to get anywhere near the experience you have had.

Your experience differs from mine. Try out seed 3403635342 at the default, or any other settings you prefer. Blue spots large and in charge, as the saying goes.

Diposting pertama kali oleh Vyndicu:
None of them would have required players to build solar energy over water tiles unless the player lowered the scale slider (distance between lakes) and bumped the coverage slider (size) of water on map generation.

Or... unless they thought it was faster, less work and less hassle to use available land for intricate prints like factories, and use the water for prints where the "pattern" of the needed landfill is easy to figure; i.e. straight edges and square corners. Lay out some landfill for factory sections, trying to not waste landfill where not needed, then lay out a solid block for solar. I'm pretty confident that one is easier than the other, and I know which it is.

And, before you add in the "get more space" option, don't forget that more space comes at its own costs; ammo, time to do the clearing, time and expense of building the new wall, (including the recycled parts from the older, and obviously shorter wall) and the time to remove that old wall, since it's probably going to be exactly where the new solar field is going. Otherwise I've expanded in the wrong direction.

Diposting pertama kali oleh Vyndicu:
How many Reactors you have built and successfully run without bells and whistles don't matter. There always exists a better solution or blueprint to get extra energy.

The reverse is also true. How many you have built with all the trappings doesn't affect reality either. That nearly half the prints found didn't need what you consider essential suggests that the "better way" is the one which fulfills the needs of the player, not the needs of a cost comparison. In other, repeated, words;
There always exists a better solution or blueprint to get extra energy.

Which brings me, out of order, to
Diposting pertama kali oleh Vyndicu:
I would not be surprised if you missed those blueprints due to needing to know what to look for in the first place. The search for both Factorio Blueprint and Factorio School could be better, to be honest.

One: as specified, it was the School site only. Two: it could have been worse as well. Three: it was a pure, unfiltered, random selection. I went there to see if the "community" consensus matched the claim made. I choose the simple approach. Open factorio.school, type "nuclear" in the 'search titles' box, look at the requirements to build each of the prints on the first two pages. Storage tank listed, even 1, classified as group "with". Storage tank not listed, not even 1, classified as group "without". Likes, age, description, requirements, conditions to build, or reasons not to build were all ignored. Nothing involved required me to do anything other than click and count. I did, however, not count a couple prints as they weren't reactors. One was some form of "module" to attach to a reactor for control and the other was a plant for making fuel cells. I also did not count, or even look, how many used circuits, with or without tanks, to control the reactors. My search of blueprints could have indeed been better at proving my point, if I was making one with the prints, had I searched for uncontrolled or unbuffered versions. Random chance, however, was likely to answer my query better than trying to prove a point would.

Side not on that, of the prints you so highly recommended, only one used landfill at all, and even that one used it, seemingly, for decoration - 396 landfill won't do much good otherwise. If the prints you like so well don't consider landfill important enough to include in the print, why would it be important enough for you to include in the cost calculations?

Diposting pertama kali oleh Vyndicu:
I tested the blueprint with all the bells and whistles a few posts ago to determine the equalized ore cost.

That blueprint was not my own and was submitted relatively recently to a public Factorio blueprint repository website. Using someone else's blueprint was to remove any chance of my bias influencing calculating the equalized ore cost.

Did that print include landfill?

Diposting pertama kali oleh Vyndicu:
It costs a few copper wires, an arithmetic combinator, and a storage tank. I have no idea what you meant by "fancy" inserters.

"Fancy inserters" are stack inserters. You're controlling it so that the motion of the inserters is regulated to "need" which is once every 200 seconds at most, by design. The speed and ability of the stack inserter is wasted resources when you could have used generic yellow inserters. A savings of 51 iron, 30 copper and 2 plastic, each (however that is equalized.)

Diposting pertama kali oleh Vyndicu:
Consider this scenario and I build a Nuclear Power to produce an unspecific number of GW while only needing less. What happens to the unused excess power?

Or is it better to regulate/smartly my Nuclear power only to consume what is necessary to meet the power grid demands at that moment? A smarter Nuclear power at a miscellaneous cost of a few copper wires, a single arithmetic combinator, and extra storage tanks?

Penny-wise and pound-foolish has a real application in the saving of costs.

Consider this scenario I need a loaf of bread. My regular, store-brand, loaf is $1.39. The name-brand loaves are on sale, 3 for the price of 2, and at only $2.19. I can save $2.19 with that deal. Well, not really. That "deal" is going to cost me at least $1.60, and probably $3.99 since the 1 loaf I need is all I need, and of the 3 I get 2 will mold before my household of 1 will finish the first one anyway.

With the scenario you've painted, the storage tanks are costing you an equalized cost of 6,570. How many hours of running the given reactor at say 50% capacity will it take you to save enough fuel cells to total an equalized cost of 6570? (Not counting the other costs which might have been missed, just the tanks themselves.) How many of those hours will the reactor, producing some unspecific number of GW in the base, be running before the power consumption is nearly 100% anyway? Short of massive overbuilding, which I am guilty of, a reactor built at one point probably won't be "sufficient" for very many hours and another core, or 12, is needed. First time builders might make major errors in selecting the initial GW output. Few experienced players will be off by much and the "savings" in fuel is likely to never reach the cost of creating that savings. Penny-wise and pound-foolish.

In consideration of what I've found while looking around, I'm going to deduct from the equalized costs the solar panels: 4,860, accumulators: 1,763, storage tanks: 6,570, and landfill: 76,300 as included by mistake. I'm keeping the logistic chests as a stand-in for a regular fast-transport belt system (mostly undergrounds which are expensive anyway), and not adjusting the cost of inserters because I don't know how you've "equalized" costs. New total is 230,036. In comparison to the similar output solar, as calculated by you at 4,234,376, solar costs 18.4x nuclear. And the original ball-park estimate originally given by someone (don't care who) of 20x seems to have reasonably close for a back of the napkin estimate.

Not that I expect you to accept these conclusions, just to place them into finalized form so others, who do accept my logic, will have the numbers readily available.
Hedning 17 Apr 2023 @ 12:21pm 
Diposting pertama kali oleh Chindraba:
If the prints you like so well don't consider landfill important enough to include in the print, why would it be important enough for you to include in the cost calculations?
lol, you're being too nice. We all know the answer is that it is 20 ore per tile, so it boosts the cost for nuclear in the comparison.

I didn't look that closely at his calculations, but did the extra solar and steam storage for the reactor also increase landfill? Very clever adding some solar to the reactor so you can add 180 ore cost per panel in landfill cost.
Chindraba 17 Apr 2023 @ 12:46pm 
Diposting pertama kali oleh Hedning:
I didn't look that closely at his calculations, but did the extra solar and steam storage for the reactor also increase landfill? Very clever adding some solar to the reactor so you can add 180 ore cost per panel in landfill cost.
Removing the landfill, with or without removing the solar/tanks, would have corrected for that anyway. Dumping it all canceled nearly all fluff (close enough anyway) and put the original 20x guess might close to the mark. By ballpark wouldn't have been much better than the original, way out, 5-6x. Then again, costs, in ore, pollution or power, are not something I usually give much weight to anyway. My time and my goals are the costs I measure most. The time spend on a puzzle might "waste" hours of multi GW power, generate 3 or 4 tonne of pollution, squander 20% of an entire iron ore deposit, but I had a "good time" so it was a "good investment". My overbuilt map is nearly worthless, without a 85% revamp, on the road to meagbase, but it sure was fun.
Now I get to polish up the results, surprisingly disturbing, or is that disturbingly surprising, on the cyclical knock-on effects of pollution from using flamethrowers. (Way back in the history now.)
Diposting pertama kali oleh Chindraba:
Diposting pertama kali oleh knighttemplar1960:
A lot of my screen shots are all ready taken and since I'm visually impaired having the grid on helps with placement and building. I'll try to remember to turn it off for new screen shots going forward.

I'm not impaired, and I use the grid often enough myself. I've even set it to be on in debug mode (if I have the options named right) so that F5 toggles is on/off for me. For some reason in screenshots it makes things worse, even though it makes things better in the live version.

having the grid on helps with placement
While it only makes a difference if avoiding off-by-one problems, I've found that I can make placement easier if I have one rail track somewhere in the print. Of course it only works for prints I make myself and connect with other prints also "tagged" that way. It's the same as using a grid snap of 2x2 but won't interfere with a relative snap when I want it for running tile placement. I just have to remember to keep rails out of the ♥♥♥ inventories unless I'm doing rails. That way the rail isn't built and I can delete the ghost after the print's been planted.

The other trick I use when making prints that go together, like pieces of a plant in smaller prints, is to keep one thing from the next-door print with some size to it. Roboports are great if they are close, or big poles and substations. Nothing with a recipe, though. I use that thing as a "key" to align the new print with the old print. If I wiggle the mouse while clicking then it says it can't place the print because a roboport, or whatever, is in the way. The big blue square even lets me place prints from zoomed-out map view.
I use roboports and substations as my connection points too but I leave trees that I don't have to remove so when I drop a blueprint its a shift click so it has to be correctly placed.

I have a hard time using the minimap to do that so I have to zoom in and shuffle around a bit.
Chindraba 17 Apr 2023 @ 6:11pm 
Way back in history, about half-way into this thread, someone (they might recall who they are, but this isn't about them anyway) posted a list of pros and cons for a couple defense setups: 1) Local pumpjack feeding a row of flamethrowers and 2) solar power fed lasers. One of the flamer cons had me thinking.

  • Low Constant Pollution from producing fluid and firing flames, which can add a few biters to attack waves requiring more firing time to kill in a feedback loop in a megabase with a large-than-the-base-perimeter pollution cloud (Negative)

My thinking was, basically, "yup, that's gonna happen, but it can't be that bad." Well, I've a nasty habit of being wrong in this game when "it can't be that bad" is in my brain. Usually I end up finding out I don't "know the half of it," as a common dismissive phrase suggests. I decided to "know" the answer to "how bad can it really be."

Stop

This is going to be a really large wall of text.

  • If you're interested in how bad it can get, just skip to the bottom and find out, then wonder how, in the name of WUBE, I got that, come back to the top and read it
  • If you're curious as to what, on the face of Nauvis, I'm thinking and doing, read on, and wish you hadn't
  • If you value your sanity, move to the next post

Now, for the insanity to follow:

By that time I'd already done an experiment and learned that the pumpjack involved is of no concern, never generating enough pollution to diffuse (min of 15) and barely enough that its mark would sometimes stay on the map long enough for me to read the value. I also learned that the pollution effects were minimal, and the user making the list had seen and acknowledged those numbers. The testing was done in clean air, so diffusion didn't occur beyond one chunk. With a heavily polluted area, however, the additional pollution could contribute to further diffusion, and eventual absorption by a spawner, increasing the size of future attack forces. And as the item suggested the pollution from killing that bug would cause more bugs. I pictured a spiral of increase where I'm killing 200 at a time instead of 50. "Yup. Better check this out," says I to myself.

I set up some parameters for the small-scale model I needed. Remember, the question was "how bad can it be?" so some things are set to the worst option when there is a choice. For example, tiles absorb pollution based on what type of tile they are. Often, on my maps, those tiles are grass as I pick maps based on the "greenness" of the preview. For the model I used sand, the worst natural land tile there is for absorption of pollution. Likewise, trees are super handy with pollution, so I included none.

So a scratch pad and I started to do some math. (Actually, I did the math and the scratch pad just kept track of what I'd done.) I happen to like math. I also happen to not like boring repetition, even in math. It rapidly got that way. Next answer; spreadsheets. That worked, sort of. Enough that I realized there was more I needed to include in my model.

One was how the pollution is handled when a bug is added to the attack force. I had the initial steps down: spawner absorbs pollution at a defined rate (20 + 0.01 * chuck's pollution) and once it has absorbed its threshold amount it sends one of its bugs to join the attack, costing some determined, and defined amount. Say the threshold is 240 pollution (3 x Big Biter at 70% evolution) and the bug sent is Medium Biter (20 pollution to attack), what happens to the remaining 220 pollution the nest absorbed? Two options: 1) it's tossed out, or 2) it's kept in the bank as a balance. Option 1 really means that sending the medium biter cost all 240 even though the wiki says 20. Option 2 means bugs will be sent a lot sooner, and faster. Well, 'how bad?,' option 2 wins. Another one was the seemingly unrestrained ability of spawners to add bugs to the attack force. If a chunk has enough pollution to get things going, and with the bank from above, a spawner could, in theory, send a bug every second. (Well, all the math is 64-ticks, but that's close enough for "thinking" about it.) I wanted "bad" but not apocalyptic! Suffice it to say there was a bit more research needed. And, the spreadsheets were not up to the task. I'm naturally a programmer, and a systems-type person, so the next level was a simple (though final version not quite so simple) Perl script to handle it all.

Here, in final form, are the parameters I applied
  • One-chunk wide column.
  • 7 operational chunks: flamer back-area, flamer line, battle zone, dead zone, 3x spawner zone
  • Terrain of chunks all sand, the worst natural "ground" rate
  • Chunk absorption rate normalized to 64 tick intervals, matching spawner check cycle (0.006335/sec)
  • First chunk past flamethrower range completely clear of spawners and worms. (Medium worm range is 30)
  • Second chunk past flamethrower range 3 spawners making 2 medium spitter and 1 big biter
  • Third chunk past flamethrower range 2 spawners making 1 big spitter and 1 medium biter
  • Fourth chunk past flamethrower range 1 spawner making 1 big biter
  • Any new chunks (diffusion can create chunks) absorption rate = 0.01024 after creation (out-of-map)
  • Use the same perimeter design as my first test (for consistency)
  • Outer edge of flamethrower range at, or just inside, chunk boundary
  • Outer wall inside flamethrower range
  • Evolution > 70% (Enable Big spitters, disable small biters and small/medium spitters)
  • Threshold of spawner absorption based on 70% evo
  • Expansion ignored, no effect on the calculations (settle range is min 3 chunks distant)
  • Sideways dissipation presumed to be equal coming in as going out
  • Chunk before the operational ones, inside the perimeter, is the pollution source - set at a constant value - and diffuses into the operational area under normal rules.
  • Each cycle will have the pollution spread, then the spawners absorbing what they can, then the ground gets its share.
  • All bugs 'sent' in the minute cycle are instantly available to the attack party, no attack delay
  • The attack force, if it exists, will attack every minute, plus 48 ticks (57 pollution cycles)
  • Pollution created per attacker killed: .06 (6 avg. pollution from killing avg. of 100 attackers)
  • Pollution from laser kills is zero, duh. (That didn't take me much math.)
  • Each spawner can "own" up to 7 bugs, as observed in a 99+% evo map
  • Each spawner will "spawn" a bug every 5 seconds. [Base on observations for a 99+% map]

Some clarifications, the attacks will happen, if there are any bugs in the army, every minute. Normal is random 1-10 minutes, but I figure a large base is likely to have one, somewhere, nearly every minute, and for worst-case, that's all going to be "here". The initial pollution cost per kill is based on the pollution seen in my earlier test. The "production" pollution behind the flamers is a constant source (never looses pollution from diffusion). The spawners freely, w/o regard to pollution, spawn bugs up to their limit. As per the wiki. Also, as the wiki states it happens faster as evo goes up, I decided to base it on 99% evo observations. "Worst it can get."

I ran the script, and the numbers looked too safe. I modified the script to allow me to vary the initial pollution values for the operational chunks, to modify the existing pollution in the production area, and to alter the cost/kill of the flamers. I added the ability, somewhat, to detect and report if and when the attack force reached some "stable" level. Increased the 'time' from 100 "minutes" to 236 cycles of attacks, equating to just a tiny bit over 4 hours of game time. Created a table to compare the results easily.

I DO NOT like or completely trust these results on a gut level. They seem counter to everything nearly everyone here has accepted about the effect flamers would have on the attacks. Most people, on the side of flamers anyway, seemed to accept that there'd be some, not much, and mostly insignificant, impact from using flamers. I've done the best I could to collect facts and understand the mechanics of the game as they apply here. I even discussed it in some of the Discords for Factorio. In short, like it or not, I actually think the results are valid, and prior thinking, for me and others, was based on a lack of information.

Finally, trusting my programming skill or not, the results make me sure I've missed something, somewhere. If you know Perl, know Git, and know something about the internals of Factorio which apply here, please do check out the repo and find the problems with it. It's 'dirty' code, as it was supposed to be a quick one-and-done thing. Of note, the number of columns coded in are designed to fit the table in the code box here. Using it yourself adjust as you wish to fit your screen.

https://gitlab.com/Chindraba/fact_pol_sim

Results

  • Column heading: Pollution is the amount in the 'production' chunk which diffuses into the area at the normal 2% rate
  • Column heading: Pol. Base is how much pollution each of the chunks starts with, there's an extra .00010 in there to guarantee that the first diffusion process will diffuse
  • Row label: Laser 0, 0 pollution per 100 kill from lasers
  • Row label: Flamer 6, 6 pollution per 100 kill from flamethrowers (my experimental results)
  • Row label: Flamer 12, 12 pollution per 100 kill from flamethrowers (possible)
  • Row label: Flamer 24, 24 pollution per 100 kill from flamethrowers (stupidly high)
  • Row label: Flamer 50, 50 pollution per 100 kill from flamethrowers (insanely high)
  • Row label: Flamer 100, 100 pollution per 100 kill from flamethrowers (just because)
  • Table values are the number of bugs killed after 236 cycles, 4 hours of game time
  • Row label: Size, the attack size which seems to be the stable value for the above row, xx+ for when it bounces between two numbers and xx is the higher one, -xx when no such number found, and xx is the last size
******************************************************************************** Pollution: 100: 500: 500: 500: 1000: 1000: 5000: 5000: 10000: 5000: 100000 Pol. Base: 20: 20: 100: 500: 20: 500: 20: 500: 20: 500: 500 -------------------------------------------------------------------------------- Laser 0: 2079: 4383: 4403: 4458: 8051: 8145: 9784: 9832: 9821: 9832: 9897 Size: 7: 19+: 19+: 19: 35: 35: 42: 42: 42: 42: 42 Flamer 6: 2089: 4386: 4407: 4461: 8058: 8152: 9784: 9832: 9822: 9832: 9897 Size: 7: 19+: 19+: 18+: 35+: 35+: 42: 42: 42: 42: 42 Flamer 12: 2101: 4389: 4410: 4464: 8064: 8158: 9784: 9832: 9822: 9832: 9897 Size: 7: 18+: 19+: 19: 35: 35+: 42: 42: 42: 42: 42 Flamer 24: 2123: 4396: 4416: 4471: 8076: 8172: 9784: 9833: 9822: 9833: 9897 Size: 7: 19: 19: 19+: 35+: 35: 42: 42: 42: 42: 42 Flamer 50: 2168: 4410: 4430: 4485: 8106: 8199: 9785: 9833: 9822: 9833: 9897 Size: 7: 19+: 19+: 18+: 35: 35+: 42: 42: 42: 42: 42 Flamer 100: 2268: 4437: 4459: 4512: 8159: 8254: 9785: 9833: 9822: 9833: 9897 Size: 7: 19: 19+: 18+: 35: 35+: 42: 42: 42: 42: 42 ================================================================================
Terakhir diedit oleh Chindraba; 18 Apr 2023 @ 6:25am
Khagan 17 Apr 2023 @ 6:35pm 
Diposting pertama kali oleh Chindraba:
I DO NOT like or completely trust these results on a gut level. They seem counter to everything nearly everyone here has accepted about the effect flamers would have on the attacks. Most people, on the side of flamers anyway, seemed to accept that there'd be some, not much, and mostly insignificant, impact from using flamers.

Isn't that exactly what your results show? When the base pollution rate is low, the flamer pollution causes a small but measurable increase in the number of bugs killed compared to laser defence. When the base pollution rate is high, the flamer contribution is completely negligible. Unless I'm misreading your table.
< >
Menampilkan 256-270 dari 278 komentar
Per halaman: 1530 50

Tanggal Diposting: 2 Apr 2023 @ 1:01am
Postingan: 278