How to fix module balance

Place to discuss the game balance, recipes, health, enemies mining etc.
User avatar
thereaverofdarkness
Filter Inserter
Filter Inserter
Posts: 532
Joined: Wed Jun 01, 2016 5:07 am
Contact:

How to fix module balance

Post by thereaverofdarkness » Sat May 26, 2018 6:57 pm

The problems with current module balance include, but are not necessarily limited to, the following:

1.) Speed and productivity modules are useless without beacons.
2.) Speed modules are generally useless without being used in tandem with productivity modules.
3.) Efficiency 3 module is insufficient to be worth a module slot.
4.) Productivity as a whole is lackluster and has overwhelming drawbacks, leading to it being used only on the highest level intermediates.
5.) There is no justifiable reason, gameplay-wise, that productivity can't be used on non-intermediates.
6.) There is no reason other than top level products that productivity can't be used in beacons.


Here's the original module attributes:
Speed
Speed: +20% / +30% / +50%
Energy: +50% / +60% / +70%

Efficiency
Energy: -30% / -40% / -50%

Productivity
Productivity: +4% / +6% / +10%
Speed: -15% / -15% / -15%
Energy: +40% / +60% / +80%
Pollution: +5% / +7.5% / +10%
*can only be used for intermediates
*cannot be placed in beacons


Here's my suggested module attributes:
Speed
Speed: +25% / +37.5% / +50%
Energy: +25% / +50% / +75%

Efficiency
Energy: -30% / -45% / -75%

Productivity
Productivity: +6% / +8% / +10%
Speed: -10% / -12.5% / -15%
Energy: +10% / +15% / +20%
*can be used for all products
*can be placed in beacons - subject to dispute as explained below


You may notice I have suggested an enormous buff to productivity. Let me defend that. First, note that the maximum productivity bonus remains unchanged at +10%. This means that your ever important high-tech science, processing unit, and rocket part productivity is not increased, it only has its penalties decreased. As these final steps in production make up a tiny part of your factory anyway, the energy and pollution penalty on the setup is fairly unimportant overall. Those penalties were doing far more to hinder the already underpowered usage of productivity in lower production steps than they were to hurt your use of productivity at the top end of the chain. I removed the pollution penalty entirely because there's no gameplay reason to justify it, and it makes no sense that improved material efficiency would increase pollution. The pollution output is already increased by the increase in energy cost.

Should productivity modules go in beacons? I feel that it would cause no problems at all, except at the very highest levels of production, and even there it is disputable that it's really an issue. First, I want to tackle my point that they should be allowed on non-intermediate products. There isn't a single non-intermediate product which you produce in large amounts, with which you would gain a stronger benefit from using productivity than you would from simply putting it on rocket part production. It doesn't even come close. Sure, you could get a lot of resources by putting productivity on nuclear reactors or rocket silos, but at some point you'll have as many of those as you can use. You could put productivity on a Power Armor Mk II, but will you ever produce enough to get even one more armor suit? And would you want that? I don't see a need to deny productivity on non-intermediates. A bigger concern would be the high-end intermediates that really do cost a tremendous amount of resources. It starts with processing units, then moves on to high-tech science, and eventually you reach the most expensive intermediate: the rocket part. If you were going to make an argument that certain products shouldn't be allowed to have productivity modules, you should focus on these products.

And so I get to the point about beacons. There is a good case to be made in saying that putting productivity in beacons is bad because it allows productivity stacking on the highest levels of the production chain, potentially saving tremendous amounts of resources. There is also a good case to be made in saying that all other levels of production will get along just fine without productivity in beacons, as you primarily want to mix productivity with other module types anyway. But I think we should stop worrying and learn to love the productivity. When you look at the sheer amount of resource you can gain, it seems like a very large amount, and indeed you would be an idiot not to put productivity at the top of your chain. But if you look at it this way you are deceiving yourself. Resources in Factorio aren't ultimately limited. Using productivity modules doesn't cause an intrinsic gain in the player's net wealth; rather it is a way of improving time efficiency by enabling a mining outpost to be used for longer before another one must be located and set up. But furthermore let's look at how much you can really gain from putting productivity in beacons. You can squeeze probably a logistically viable maximum of 7 beacons around a mk 3 assembler, for a grand total of 18 modules. Using my suggested module attributes above, if you put in 12 productivity 3 modules, you get +120% productivity and -180% speed. With 6 speed 3 modules to counteract this, the final speed is +120% and the energy cost is +670%. Looks like a lot, right? Now if we had instead only used the 4 productivity modules we could fit in the assembler along with 10 speed modules and 4 efficiency modules, we would get +40% productivity, +440% speed, and +530% energy cost. That means the second setup runs 145% faster and costs 18% less energy, and all the first setup gains is about 57% more productivity than the second setup. I'm not saying using all those productivity modules isn't an optimal setup, just that it isn't as game-changing as you might think it is.

But feel free to disagree with me in the comments below.

Quick points on the other elements of my module balance suggestions:
  • Speed module 1 gives no decrease in energy efficiency. It only speeds up the machine. Putting these in, you are paying a module slot and the cost of the module as an alternative to simply building more machines.
  • Efficiency module 3 now gives a suitable decrease to energy consumption, strong enough to counteract the hefty energy cost of using high level speed modules, or you can just put 1 of these in each machine and have module slot(s) left over for productivity modules.
  • Productivity 1 has a more substantial increase in productivity bonus as well as far weaker penalties, so you can safely flood your entire factory with these at all levels if you really want to capitalize on the resource efficiency.
Finally, a link to my mod in which these changes have been (mostly) implemented. https://mods.factorio.com/mod/Reavers_Module_Balance You can try the changes in-game and see how they play out for you.


edit: We've been discussing things in the thread and we seem to generally agree that it's not enough to buff everything else, productivity 3--and the way productivity and speed multiply together--needs to be nerfed. The simplest way to accomplish this is to make penalties apply separately from bonuses and multiply with each other. Here's how that would look:

4 productivity-3 modules
+40% productivity
-60% speed
throughput: -36%

4 productivity-3, 4 speed-3
+40% productivity
-60% speed
+200% speed
final speed: +20%
throughput: +68%

comparison with 8 speed-3
+200% speed
throughput: +200%

So this would mean that productivity modules only increase material efficiency, and always decrease rate of output.

If efficiency modules were done the same way, you would reduce your energy cost with the same number of efficiency modules, regardless of the number of other modules you have in your system. This means that the intrinsic value of an efficiency module grows with more beacons in use. Using this base idea, BlakeMW had a suggestion to decrease the actual energy savings from efficiency while adding a pollution reduction effect so that it costs more modules to eliminate energy cost, but you can get rid of pollution just as easily as before.

eff1: -20% energy usage, -5% pollution (total pollution w/ 1 module: -24%)
eff2: -30% energy usage, -15% pollution (total pollution w/ 1 module: -40.5%)
eff3: -40% energy usage, -30% pollution (total pollution w/ 1 module: -58%)

I think it's a great idea, but let us know what you think in the thread below.
Last edited by thereaverofdarkness on Mon Jun 04, 2018 7:56 pm, edited 3 times in total.

User avatar
bobingabout
Smart Inserter
Smart Inserter
Posts: 6860
Joined: Fri May 09, 2014 1:01 pm
Contact:

Re: How to fix module balance

Post by bobingabout » Tue May 29, 2018 8:21 am

the numbers aren't quite what I'd call right, however...

The -15% speed across the board on productivity modules is there for a reason. Balancing modules has come up MANY times before, and on one such occasion, I personally ran numbers on the productivity module, and the speed value being constant does make a lot of sense when you go through the maths. It gives it a feeling of the higher tiers being more useful and efficient, vs "it does the same, but a little bit stronger", an incentive to actually use a MK3 instead of just two MK1s. but Maths on the actual outcome of the machine showed some surprising results on why it should keep this constant that is hard to put into words off the top of my head.


Anyway, to look at the bigger picture of numbers, you change speed from 20, 30, 50 to 25, 40, 50... that gives you a lot of incentive to go for MK2, but very little to then increase it to MK3. I'd say instead go 20, 35, 50 to give the MK2 to MK3 boost more incentive, or even 25, 37.5, 50. And you do have the .5% in other places, why not here?

As a general rule, the improvement from a MK2 to MK3 should be AT LEAST the same as the improvement from a MK1 to MK2, so the original 20, 30, 50 has an improvement of 10, then an improvement of 20, so that fits. Your sugestion has an improvement of 15, then and improvement of 10, so that doesn't.
Creator of Bob's mods. Expanding your gameplay since version 0.9.8.
I also have a Patreon.

Frightning
Filter Inserter
Filter Inserter
Posts: 801
Joined: Fri Apr 29, 2016 5:27 pm
Contact:

Re: How to fix module balance

Post by Frightning » Tue May 29, 2018 10:49 am

The problems with modules cannot be fixed with just number tweaking. They are fundamentally a result of the way they work and are implemented in the game currently. Your proposed changes to prod modules will break the game (specifically allowing them in beacons=it's possible to make, for a 3x3 assembler, +160% productivity, with a speed penalty exceeding -100%...I'm not sure the game can even handle a speed penalty of -100%, let alone > that). Even just getting rid of intermediate product restriction will make prod+spd even more mandatory for large scale factories (this, because of the savings in low-level infrastructure and raw resource mining required to make a given high end product X at rate Y).

Honestly, the most interesting idea to me is simply removing beacons altogether. As much as I don't like just removing things from the game, the meaningful choice between types of modules is destroyed precisely because of how prod+spd modules interact, and the fact that beacons allow you to effectively add module slots, for an energy price and subject to specific layout requirements. This means that there's really only one ideal setup, and that setup is prod+spd with alternating rows of beacons and assemblers. Without beacons, layouts are no longer constrained, the interaction between speed and prod is thus not grossly exploitable, and synergy between spd and eff (specifically 1xSpd3+3xEff3=most energy and pollution efficient setup) has a chance to be relevant.

bobucles
Smart Inserter
Smart Inserter
Posts: 1591
Joined: Wed Jun 10, 2015 10:37 pm
Contact:

Re: How to fix module balance

Post by bobucles » Tue May 29, 2018 1:18 pm

Speed penalty caps at -80%. That's pretty painfully slow and generally not a great idea to work with.

Productivity modules scale exponentially. The reason is because prod modules can stack down a production chain, giving a bonus to each individual step. Getting 40% more rocket per rocket is great but what really sells it is getting 40% more rocket parts, 40% more blue circuits, 40% more red circuits, 40% more green circuits, 40% more wire and 20% more iron place. Increasing productivity doesn't just give you more rocket. It gives more of every single step leading up to the rocket so instead of getting 1.4^4 resources (roughly 4x) resources you would get up to 2.2^4, totaling 20x yield or more.

That's the same reason why not every item gets productivity bonus. It cuts down the exponential by one step, which in this case brings a 5x multiplier down to 4x. Of course it would be nice if all the intermediate items got to have prod bonus. It's pretty messy that some intermediates inexplicably don't allow prod bonus when the game says it's supposed to allow it.

Productivity scales FAST. That's why beacons don't get to have it.

dragontamer5788
Fast Inserter
Fast Inserter
Posts: 153
Joined: Fri Jul 15, 2016 1:44 am
Contact:

Re: How to fix module balance

Post by dragontamer5788 » Tue May 29, 2018 5:49 pm

I was expecting this topic to be about the utter uselessness of Efficiency modules in general. Instead, this topic seems to be about buffing the most powerful module in the game.

Productivity is the best module, period. Bar none, there is absolutely no module that is as good as the Prod3 module. Prod3 modules inside your science lab will literally shrink the rest of your base by 20% (or in the inverse: allow your base to produce 20% more science). There is literally no other module in this game that makes the REST of your base better the more you use it. Even in science labs and rocket silos alone, Prod3 is completely and utterly bonkers in terms of the resource savings you get.

Speed modules are okay, but are especially good when paired with Productivity. This is more of a side-effect of how good productivity is.

Efficiency modules are never better than the equivalent number of solar panels, except for Eff1 modules maybe inside of Refineries (because refineries have such high power usage). Eff3 modules cost 1080 iron, so you can build roughly 27 solar panels for the cost of a single Eff3 module. The only time to build Eff3 modules is to build Power Armor 2, they're literally useless everywhere else.

Eff1 modules are cheaper than solar panels at least, but with exception of Refineries, there's no building in the game that actually has high-enough power usage to make Eff1 comparable to a solar panel. And btw, solar panels are the most expensive form of energy. Build nuclear instead or anything else really, and you'll do far better than Eff1 modules. Case in point: Eff1 in a furnace saves you 30% energy. An electric furnace is 180kW, so you only save 54kW per Eff1 module. Guess what? I'd rather have 2 x Productivity3 in my furnaces (effectively increasing the efficacy of my mining-drills by 20%) and build a 60kW solar panel at my solar farm.

-----------

If Eff3 modules are to be a serious option in this game (and they don't necessarily have to be), the proper balance would be to triple, or quadruple the Eff3 module's efficacy. 240% reduction in electricity costs per Eff3 would be going somewhere, but honestly, I might still choose Prod3 instead. Especially because every beacon uses +300kW anyway. Maybe if Eff3 reduced beacon energy AND building energy, AND was on the order of -240% reduction or so per module. Yes, -480% energy usage per Eff3 module. Its not really a lot. 27 solar panels x 42kW average power across the day is 1.13 MW.

So an Eff3 module in an Electric furnace at -480% energy usage is still saving you only 864kW. Even at 480% "savings" (so that electric furnaces generate energy instead of use energy), an Eff3 module makes less energy than its equivalent 27 x solar panels. That's how bad Eff3 is.

User avatar
thereaverofdarkness
Filter Inserter
Filter Inserter
Posts: 532
Joined: Wed Jun 01, 2016 5:07 am
Contact:

Re: How to fix module balance

Post by thereaverofdarkness » Wed May 30, 2018 5:59 am

bobingabout wrote:The -15% speed across the board on productivity modules is there for a reason. Balancing modules has come up MANY times before, and on one such occasion, I personally ran numbers on the productivity module, and the speed value being constant does make a lot of sense when you go through the maths. It gives it a feeling of the higher tiers being more useful and efficient, vs "it does the same, but a little bit stronger", an incentive to actually use a MK3 instead of just two MK1s. but Maths on the actual outcome of the machine showed some surprising results on why it should keep this constant that is hard to put into words off the top of my head.
The speed penalty balances for the productivity bonus. It ensures that the productivity module gives you increased resource efficiency at the cost of decreased output rate. The lower productivity of lower-tier productivity modules is why I gave them a smaller speed penalty. I don't see anyone flooding their factory with productivity 1, so it's pretty hard to dispute that buffing it isn't in order.

The real problem isn't how strong productivity is. It's about how strong the combination of productivity and speed is. If you only used productivity, you would always lose production rate to gain resource efficiency. It would always increase, not decrease, the size of your factory for the same rate of output. The problem is the way that speed so easily negates the penalty. It would work much better if the speed bonus and speed penalty were cumulative--so a -60% speed from productivity modules and a +100% speed from speed modules would result in a net -20% speed. That would probably fix the whole problem in one fell swoop. The same could be done for efficiency modules: make their energy reduction cumulative with the increase. This way, if you have +200% energy cost and you put in a single -75% module, it takes it down to -25%.
bobingabout wrote:Anyway, to look at the bigger picture of numbers, you change speed from 20, 30, 50 to 25, 40, 50... that gives you a lot of incentive to go for MK2, but very little to then increase it to MK3. I'd say instead go 20, 35, 50 to give the MK2 to MK3 boost more incentive, or even 25, 37.5, 50. And you do have the .5% in other places, why not here?
Reasonable. I was thinking about it myself. I think the 40% is a throwback to a side project I was doing in which the pollution increase had a positive benefit. I'll change it to 25/37.5/50.


dragontamer5788 wrote:I was expecting this topic to be about the utter uselessness of Efficiency modules in general. Instead, this topic seems to be about buffing the most powerful module in the game.
But I didn't buff it in the way you think I did. The productivity 3 module still gives the same productivity bonus and has the same speed penalty. The only thing buffed about it is the energy and pollution values which, where it was ever used anyway, are trivial. The main point was to boost the 1 and 2 to be usable on lower levels of your factory. I could even see fit to nerf the productivity 3 just because of how powerful it is at the top end of the chain. But the real problem with productivity is that it can be applied to extremely expensive things like high tech science or rocket parts. If it was relegated to low-end production such as plates, iron gear wheels, green circuits, etc. then you could gain a lot from them, but only by having a lot of them.

I'm not sure how to fix productivity, but my suggestion doesn't buff it in the way you say it does.
dragontamer5788 wrote:If Eff3 modules are to be a serious option in this game (and they don't necessarily have to be), the proper balance would be to triple, or quadruple the Eff3 module's efficacy.
Are you basing that on energy costs of vanilla speed/productivity setups? I drastically reduced the energy cost for productivity. When I ran the numbers after the fact, I found that efficiency was definitely viable. It could maybe use a small increase, but you must compare it with the energy cost numbers we're getting with the other module changes.
dragontamer5788 wrote:So an Eff3 module in an Electric furnace at -480% energy usage is still saving you only 864kW. Even at 480% "savings" (so that electric furnaces generate energy instead of use energy), an Eff3 module makes less energy than its equivalent 27 x solar panels. That's how bad Eff3 is.
That says a lot about how OP solar panels are, really. Part of the problem with modules is that energy production in Factorio is so easy that it's almost a non-factor, so energy savings has a low ceiling of value, and enormous energy cost increases aren't seen as much of a penalty. If energy production were actually scarce, efficiency would matter a lot more, and speed's energy cost increase would be a lot more of an issue. Once you start mixing efficiency modules into your setup, you reduce the total effective number of productivity modules you can make work.

bobucles
Smart Inserter
Smart Inserter
Posts: 1591
Joined: Wed Jun 10, 2015 10:37 pm
Contact:

Re: How to fix module balance

Post by bobucles » Wed May 30, 2018 12:14 pm

But prod1's aren't underpowered. They are in fact absurdly powerful. The only issue with them is players use blueprints and the game can't automate module swapping. So rather than play the module swapping game, it's less thinking to plop down a prefab prod3 blueprint instead of incrementally upgrading from 1 to 2 to 3.

Frightning
Filter Inserter
Filter Inserter
Posts: 801
Joined: Fri Apr 29, 2016 5:27 pm
Contact:

Re: How to fix module balance

Post by Frightning » Wed May 30, 2018 8:33 pm

bobucles wrote:But prod1's aren't underpowered. They are in fact absurdly powerful. The only issue with them is players use blueprints and the game can't automate module swapping. So rather than play the module swapping game, it's less thinking to plop down a prefab prod3 blueprint instead of incrementally upgrading from 1 to 2 to 3.
The main thing keeping people from abusing Prod1 as soon as their available is that they don't yet have Beacons for beaconized layouts, and by the time they do, they can also have at least tier 2 modules, if not tier 3 then or soon after. Without speed beacons, the speed penalty on prod modules is actually pretty relevant and makes one think twice about using them for the material savings (you need a lot more assembly machines which each cost a LOT more energy and produce tons more pollution to get the same throughput without beacons with speed modules helping out).

bobucles
Smart Inserter
Smart Inserter
Posts: 1591
Joined: Wed Jun 10, 2015 10:37 pm
Contact:

Re: How to fix module balance

Post by bobucles » Thu May 31, 2018 1:43 am

There's still plenty of time to enjoy the fruits of prod1 in the mid-late pre rocket game. Set up two of them in blue assemblers on key products like GC, potions and labs. It's a very easy bonus to take advantage of and the -30% speed penalty is still very manageable. On GC for example the 3:2 ratio turns into a very easy 1:1 setup for a tasty 8% GC bonus. It may not seem like much but prod1's are dirt cheap. It'll pay off super fast and it stacks into building higher tier modules anyway.

It's very simple to not stack 4 prod1's to get -60% speed before speed beacons come into play.

User avatar
bobingabout
Smart Inserter
Smart Inserter
Posts: 6860
Joined: Fri May 09, 2014 1:01 pm
Contact:

Re: How to fix module balance

Post by bobingabout » Thu May 31, 2018 8:36 am

Like I said, I did Maths, and the conclusion at the end of Maths was that Productivity modules should keep the constant -15% speed. (The reason for the maths was to try and find what speed value would balance them)

And yes, I know, this is a contradiction from someone who has a mod that removes the speed penalty from their productivity modules.
Creator of Bob's mods. Expanding your gameplay since version 0.9.8.
I also have a Patreon.

dragontamer5788
Fast Inserter
Fast Inserter
Posts: 153
Joined: Fri Jul 15, 2016 1:44 am
Contact:

Re: How to fix module balance

Post by dragontamer5788 » Thu May 31, 2018 6:52 pm

thereaverofdarkness wrote:
dragontamer5788 wrote:So an Eff3 module in an Electric furnace at -480% energy usage is still saving you only 864kW. Even at 480% "savings" (so that electric furnaces generate energy instead of use energy), an Eff3 module makes less energy than its equivalent 27 x solar panels. That's how bad Eff3 is.
That says a lot about how OP solar panels are, really. Part of the problem with modules is that energy production in Factorio is so easy that it's almost a non-factor, so energy savings has a low ceiling of value, and enormous energy cost increases aren't seen as much of a penalty. If energy production were actually scarce, efficiency would matter a lot more, and speed's energy cost increase would be a lot more of an issue. Once you start mixing efficiency modules into your setup, you reduce the total effective number of productivity modules you can make work.
On the contrary, Solar Panels are the highest-cost, lowest-energy producer in entire the game.

A nuclear reactor costs 3500 iron (500 steel + 1000 iron), but provides 120MW. That's ~30kW per iron. Solar panels are 42kW for 40 iron. That's ~1kW per iron. I guess you need to have steam turbines and heat pipes too, but those only drop the overall efficiency of nuclear by 1/2 or so. Maybe ~15kW per iron once you factor in all of those steam turbines.

IE: Nuclear reactors (in 2x2 configuration) are somewhere on the order of 30x more cost-effective than solar. The fact that Eff. Modules can't even compete against the worst-cost effective form of energy in the game only demonstrates their utter uselessness.
bobucles wrote:There's still plenty of time to enjoy the fruits of prod1 in the mid-late pre rocket game. Set up two of them in blue assemblers on key products like GC, potions and labs. It's a very easy bonus to take advantage of and the -30% speed penalty is still very manageable. On GC for example the 3:2 ratio turns into a very easy 1:1 setup for a tasty 8% GC bonus. It may not seem like much but prod1's are dirt cheap. It'll pay off super fast and it stacks into building higher tier modules anyway.

It's very simple to not stack 4 prod1's to get -60% speed before speed beacons come into play.
Alternatively, just stick a speed module + 3x Prod 1 modules. Like, speed 1 + Prod1 is still a really good combo.

Zavian
Smart Inserter
Smart Inserter
Posts: 1460
Joined: Thu Mar 02, 2017 2:57 am
Contact:

Re: How to fix module balance

Post by Zavian » Thu May 31, 2018 6:56 pm

dragontamer5788 wrote:On the contrary, Solar Panels are the highest-cost, lowest-energy producer in entire the game.
In terms of ups cost, solar is the most efficient energy producer in the game. And sincwe megabases are often ups limited, that makes solar very attractive for megabases.

BlakeMW
Filter Inserter
Filter Inserter
Posts: 951
Joined: Thu Jan 21, 2016 9:29 am
Contact:

Re: How to fix module balance

Post by BlakeMW » Thu May 31, 2018 7:35 pm

Prod1 modules are SO GOOD. They aren't good in a flashy way, but they are good in a calculated ROI way. In many recipes, nearly everything above ore smelting really, building a setup of labs or assembler 2 30% larger and stuffing 2x prod1 modules in them gives a pretty fast ROI (often about 20 minutes) and the additional energy used is fairly insignificant relative to the "resources out of thin air" effect.

This effect isn't that prominent in normal play, though speedrunners have embraced it now, especially in labs, since the ROI is easily quick enough to shave some time off a speedrun, but it's really prominent in Marathon games, where it becomes totally a non-brainer to shove 2xprod1 in everything because the cost of the prodified setups is utterly trivial to what the research costs, and +8% over many tiers really adds up, potentially at least +30% production out of thin air (okay to be fair, out of a small additional amount of fuel burned in boilers, but the coal mined to run prodified setups is much less than the ore you don't have to mine, it's only in recipes as bad as smelting in electric furnace where it becomes close to burn 1 coal to get 1 free plate, many recipes have a 1:6 ratio or higher).

In terms of ROI time, 2xprod1 is actually pretty similar to a beaconized Speed3+Prod3 setup, the 2x prod1 setup only gives +8% stuff out of thin air and it runs at -30% speed, wheras the Beaconized setup gives +40% and runs at +340% speed, getting massively more free stuff at a much faster rate almost exactly balances out the much higher cost of the beaconized setup. Pretty good balance actually because ROI is an important factor, if something takes hours to get a ROI it's not very good, or if it takes 5 minutes it might be too good especially if low effort. Recognized tier3 setups can have a ROI of about 30 minutes.

Anyway. Setups I use are:

2 or 3x eff1: In many machines it's cheaper than producing more solar/accu and it reduces pollution. Often not as good as prod1 but worth putting in machines where a prod1 is impossible or unsuitable.
2x Prod1. Lots and lots.
3x Prod1 or 2x Prod1 + 1x Speed 1 in oilrefining, the 3x1 prod1 for recipes like sulfuric acid which tend to be mostly idling anyway.
3x prod1 + 1x speed1 in Assembler 3, good pre-beacons. Occasionally I'll use 4x prod1, the speed gets very low but feels worthwhile for things like high tech science pack.
4x prod3 + speed3 beacons

I don't find a lot of use for the tier2 modules. Obviously I'll use them in a transitional role on the way to prod3 + speed3 beacons but there never really seems to be a phase of the game which is dominated by the use of tier2 modules, I transition from tier1 to tier3 pretty quickly.

My problems:

Eff2 and Eff3 are effing useless. The ROI is ridiculously terrible and in particular is ridiculously terrible compared with just producing more clean electricity from solar/accu. Maybe it's partly because eff1 modules are so good, though even if eff1 modules were bad, eff2 and eff3 would still be just as bad.

Tier2 modules don't really need to exist at all. I don't have a huge problem with them, but they don't really seem to add much to the game.

I also wouldn't mind if effects are changed from additive to multiplicative when it comes to the sum of positives and negatives, so if you have -60% speed and +100% speed you'd end up with 80% speed rather than 140% speed. I feel this would improve the balance overall, beaconized setups would still be excellent but wouldn't be quite as absurdly powerful. It could also make eff modules potentially useful, -50% energy usage on the total energy consumption of prod and speed juiced assemblers might be worth having.

User avatar
thereaverofdarkness
Filter Inserter
Filter Inserter
Posts: 532
Joined: Wed Jun 01, 2016 5:07 am
Contact:

Re: How to fix module balance

Post by thereaverofdarkness » Thu May 31, 2018 8:23 pm

bobingabout wrote:Like I said, I did Maths, and the conclusion at the end of Maths was that Productivity modules should keep the constant -15% speed.
I did the maths, too. I'm fairly certain you made a mistake. Not necessarily an error in calculation, but rather an error in interpretation. Also, keep in mind your original maths were based on the assumption that the material savings is a direct earning, which I am trying to debunk/dispel. If you re-calculate for actual throughput of materials, ignoring resource efficiency, you'll get a different result. The result suggests that productivity modules at any level are extremely effective when used on the few most expensive products iff they are used with speed beacons. The productivity 1 module does not justify use on lower intermediates in large numbers of modules, whether you put them in pairs in tier 2 assemblers or try to justify the cost of tier 3 assemblers to cram in a speed module with them, because they don't work without beacons and they aren't strong enough at that level to justify beacons. Some people might feel the tier 1 productivity is useful as a material savings on a small number of projects when considering the material cost of the module, but this tiny advantage won't be drastically increased by fixing the module to the point it can be used in the rest of the factory. What we should be focusing on is rate of throughput, and the answer is that productivity modules should always reduce it. Only speed modules should increase it.

Let me re-iterate this extremely important point: The energy cost and pollution bonus of the productivity module does not impact its usefulness where it is already used, but does impact it strongly where it isn't used. Thus, removing or drastically reducing those penalties only helps to increase the use of the module at lower levels, without increasing its power where it is already most powerful.

And furthermore, if energy scarcity was at all substantial, the energy cost would actually matter and we could balance it accordingly. We can't balance energy cost penalties with Factorio's current easy power system.


dragontamer5788 wrote:On the contrary, Solar Panels are the highest-cost, lowest-energy producer in entire the game.

A nuclear reactor costs 3500 iron (500 steel + 1000 iron), but provides 120MW. That's ~30kW per iron. Solar panels are 42kW for 40 iron. That's ~1kW per iron. I guess you need to have steam turbines and heat pipes too, but those only drop the overall efficiency of nuclear by 1/2 or so. Maybe ~15kW per iron once you factor in all of those steam turbines.

IE: Nuclear reactors (in 2x2 configuration) are somewhere on the order of 30x more cost-effective than solar.
That's ignoring all of the heat exchangers, steam turbines, heat pipes, and centrifuges that go into running a reactor. In the end, it'll cost around twice that in raw materials just to place it down. But once again you're still ignoring the cost of running it, the uranium ore that goes into it and the scarcity involved. I would agree that uranium ore is entirely too abundant (it's literally more common than iron!!), but still you can't quote figures if you ignore all that.

Solar panels just keep going. They're reliable and require almost no logistics (just a few electric poles), a non-varying ratio of accumulators to run through the night (not very many, with default ~2 minute nights), and zero upkeep. They are placed once and done. I've played with solar panels reduced to 18% of their effective power output (Realistic Power mod by Kenira) and I use solar panels all the time. They take up a lot of space but they give me reliable and steady power. I use some mods that make other power generation harder as well, but other sources of power still have a far greater power density relative to the amount of infrastructure required to generate them. But when you account for resource depletion, pollution generated, fuel logistics, and other issues, it's easy to see solar panels as a good option to turn to. A lot of it has to do with available space, but that's all due to the fact that my game has a pretty balanced set of power generation options, making everything good in its place, and when I play on rail worlds, extra space is what I have. And when I am keeping my pollution low, time is also on my side. So it just goes with the way I play. Point is, the 60kW solar panels are entirely too powerful.

In fact I'm going to dig a bit deeper on this one. During development of the game, they made solar panels too powerful. But rather than nerf panels, they gave us modules to soak up all that excess power we didn't know what to do with. But then players quickly figured out how to abuse speed and productivity with beacons, and quickly started drawing enormous amounts of power, and complaining that they were getting tired of placing solar panel blueprints with bots when the real problem was with power penalties and the efficiency module. So Wube dev team-rather than fix the problem-gives us nuclear power for even more power generation. And here we are way, way down the rabbit hole with this mess. I'm trying to explain how to fix it, but first I need you all to understand where it went wrong way back near the beginning, and how piling on more band-aids will never fix the problem. Someone else said that simply changing the numbers won't fix the modules. He was partially right. The real truth is that the problem can't be fixed without digging all the way to the bottom, dredging up the heart of the problem, fixing it at the source, and then re-building it up again from there. And the longer we avoid this uneasy truth, the harder it will be.

vedrit
Filter Inserter
Filter Inserter
Posts: 276
Joined: Sun Aug 03, 2014 2:25 am
Contact:

Re: How to fix module balance

Post by vedrit » Thu May 31, 2018 8:44 pm

thereaverofdarkness wrote:In fact I'm going to dig a bit deeper on this one. During development of the game, they made solar panels too powerful. But rather than nerf panels, they gave us modules to soak up all that excess power we didn't know what to do with. But then players quickly figured out how to abuse speed and productivity with beacons, and quickly started drawing enormous amounts of power, and complaining that they were getting tired of placing solar panel blueprints with bots when the real problem was with power penalties and the efficiency module. So Wube dev team-rather than fix the problem-gives us nuclear power for even more power generation. And here we are way, way down the rabbit hole with this mess. I'm trying to explain how to fix it, but first I need you all to understand where it went wrong way back near the beginning, and how piling on more band-aids will never fix the problem. Someone else said that simply changing the numbers won't fix the modules. He was partially right. The real truth is that the problem can't be fixed without digging all the way to the bottom, dredging up the heart of the problem, fixing it at the source, and then re-building it up again from there. And the longer we avoid this uneasy truth, the harder it will be.
You may well be right, that the problem was introduced way back when and addressed in all the wrong ways. That brings up an important question: How do you fix it without upsetting your entire playerbase because of a change to core items? It's Belts vs Bots all over again.

User avatar
thereaverofdarkness
Filter Inserter
Filter Inserter
Posts: 532
Joined: Wed Jun 01, 2016 5:07 am
Contact:

Re: How to fix module balance

Post by thereaverofdarkness » Thu May 31, 2018 9:00 pm

vedrit wrote:That brings up an important question: How do you fix it without upsetting your entire playerbase because of a change to core items? It's Belts vs Bots all over again.
Personally (and I'm sure I'm wrong) I feel like the answer is to do the right thing and accept that you'll upset them. I really do think it'll be good for them to learn to put up with letting go of a bad thing they didn't know they don't want.

BlakeMW wrote:Pretty good balance actually because ROI is an important factor, if something takes hours to get a ROI it's not very good, or if it takes 5 minutes it might be too good especially if low effort. Recognized tier3 setups can have a ROI of about 30 minutes.
That's an excellent post and you make an excellent point. I would like to point out, however, that the ideal return on investment time can vary greatly depending on how a person plays. Also, your math is too clean to compare to the real world very well. I don't know what it's like for other players, but it can take me 5+ hours easily to get a ROI for my nuclear reactor--because that's how long it takes to set it up and get it running. That has a lot to do with my difficulty mods and the way I play slow, and how much experience I have with nuclear power (not much). So gameplay is a lot dirtier and seat-of-the-pants in style. One thing I notice when I play Factorio is that I don't play in terms of costs and benefits. I play in terms of I place down items and automate it and solve a logistics puzzle, and I get rewarded with free stuff forever. The free stuff forever keeps piling up until the chest fills up, and is always there when I want more. Until it isn't. Cue next logistic puzzle, or blueprints. Point being, to make a good cost analysis, you have to consider the way the player does or doesn't perform their own cost analysis during play, and account for that.

My numbers were made on the fly, they are highly analog in development style and I balanced them based on a simple input-output guess-and-check style, using my own gameplay as the model with which to test the numbers. My method can account for the organic errors your math will miss, but also has other flaws such as how my playstyle differs from others (drastically, I'm positive). I'm ranting, but the thing I want you to take away from this is an understanding of some of those factors you might have missed before.

bobucles
Smart Inserter
Smart Inserter
Posts: 1591
Joined: Wed Jun 10, 2015 10:37 pm
Contact:

Re: How to fix module balance

Post by bobucles » Fri Jun 01, 2018 3:31 am

That brings up an important question: How do you fix it without upsetting your entire playerbase because of a change to core items? It's Belts vs Bots all over again.
It's beta. Tweaking balance is the entire point of beta.

That being said I haven't found a module idea that is much better than the current system.

Zavian
Smart Inserter
Smart Inserter
Posts: 1460
Joined: Thu Mar 02, 2017 2:57 am
Contact:

Re: How to fix module balance

Post by Zavian » Fri Jun 01, 2018 4:08 am

I think that limiting beacons to use diminishing returns in some way, so that assemblers can only benefit from 4-6 speed beacons would help break the 4x4 pattern. It would also make more room for belts in heavily beaconed setups.

eg 1 beacon (2 speed modules) = 100% boost
2 beacons (4 speed modules) = 100% + 50% = 150% boost
3 beacons (6 speed modules) = 150% + 33% = 183% boost
4 beacons (8 speed modules) = 183% + 25% = 208% boost
5 beacons (10 speed modules) = 208% + 20% = 228% boost
6 beacons (12 speed modules) = 228% + 17% = 245% boost

Hopefully that is enough of a nerf that for cost efficient layouts, most players are likely to switch to 4-6 beacons per assembler. Unfortunately if ups is the sole consideration, then 8 beacons per assemblers would probably still be the most ups efficient layout.

dragontamer5788
Fast Inserter
Fast Inserter
Posts: 153
Joined: Fri Jul 15, 2016 1:44 am
Contact:

Re: How to fix module balance

Post by dragontamer5788 » Fri Jun 01, 2018 4:51 am

bobucles wrote:
That brings up an important question: How do you fix it without upsetting your entire playerbase because of a change to core items? It's Belts vs Bots all over again.
It's beta. Tweaking balance is the entire point of beta.

That being said I haven't found a module idea that is much better than the current system.
Make Eff2 and Eff3 modules generate electricity by bringing the energy costs below 0%. Oh, and make them like -200% or -300% power usage. (ie: 2x Eff3 would bring an electric furnace from 180kW power usage to -540kW power usage, aka +540kW for the rest of the base to use). At least then they'd have a use, and maybe that use is comparable to Speed3+Prod3 all over everywhere else.

I personally think that Speed3 + Prod3 is kinda fun to play with. It takes an absurd amount of engineering to mass produce modules, and the payoff is definitely worth it.

Hedning1390
Fast Inserter
Fast Inserter
Posts: 226
Joined: Fri Jan 05, 2018 8:47 pm
Contact:

Re: How to fix module balance

Post by Hedning1390 » Sat Jun 02, 2018 10:24 am

I think allowing prod modules on more products doesn't make much sense.
-Things like kovarex is balanced around getting only 1 extra. The 40 required for a cycle would turn from a negative (takes long time to get going) to a positive (high base number to multiply).
-On things like barrels you can get items by putting them in a loop which I disagree with.
-On structures such as inserters it really doesn't matter so why change it?

I very much disagree with buffing the lower tiers. They already massively cheaper, to a point where if you are only going for one rocket you will only get to build 4 tier 3 modules, the ones going in the rocket silo. Why reduce the sense of progression and push t3 even further away from where most people are?

Productivity in beacons is interesting. It creates a strategic choice. Right now you just use as much productivity as you can in everything. If you could have productivity in beacons you may still want some speed, so it's something to figure out, and may be different from item to item.

Efficiency needs more of a buff. I think eff3 should have something like 5000% to make sure a single module, even in a beacon, will put the machine on it's lowest consumption level.

Post Reply

Return to “Balancing”

Who is online

Users browsing this forum: No registered users