The Productivity Cap and why the FFF #375 recycler shouldn't be introduced

[whitecold]

I think a very nice option would be to limit the recycler to final products, which can't take productivity modules.
The reason would be much less the productivity cap, but opening up the space for alternate recipes as has been mentioned in a recent friday facts, for which the recycler is an issue as it is unclear what the result should be.

[Illiander42]

I think a very nice option would be to limit the recycler to final products, which can't take productivity modules.
The reason would be much less the productivity cap, but opening up the space for alternate recipes as has been mentioned in a recent friday facts, for which the recycler is an issue as it is unclear what the result should be.

That's actually a really neat solution that removes the need for a productivity cap (which is a good thing). And it's a simple rule: Can you put prod modules in it? Yes: No recycle; No: Can recycle.

It does reduce the options for where you recycle, but that also introduces balancing challenges for each quality.

Though we do have item voiding now...

[mrvn]

I think a very nice option would be to limit the recycler to final products, which can't take productivity modules.
The reason would be much less the productivity cap, but opening up the space for alternate recipes as has been mentioned in a recent friday facts, for which the recycler is an issue as it is unclear what the result should be.

That's actually a really neat solution that removes the need for a productivity cap (which is a good thing). And it's a simple rule: Can you put prod modules in it? Yes: No recycle; No: Can recycle.

It does reduce the options for where you recycle, but that also introduces balancing challenges for each quality.

Though we do have item voiding now...

Isn't a big use case of the recycler to get more advanced items by recyling all the low level results back into the factory? This wouldn't work if you can only recycle final products.

[Illiander42]

I think a very nice option would be to limit the recycler to final products, which can't take productivity modules.
The reason would be much less the productivity cap, but opening up the space for alternate recipes as has been mentioned in a recent friday facts, for which the recycler is an issue as it is unclear what the result should be.

That's actually a really neat solution that removes the need for a productivity cap (which is a good thing). And it's a simple rule: Can you put prod modules in it? Yes: No recycle; No: Can recycle.

It does reduce the options for where you recycle, but that also introduces balancing challenges for each quality.

Though we do have item voiding now...

Isn't a big use case of the recycler to get more advanced items by recyling all the low level results back into the factory? This wouldn't work if you can only recycle final products.

It makes the process more complicated, but doesn't change the basics. It does take away the "recycle your first step until it's legendary, then proceed as normal" option, which I feel is probably a good thing?

[SupplyDepoo]

I don't agree with the OP's rationale. The idea that you could cut an unlimited amount of gear wheels out of a single iron plate is preposterous. 300% productivity is already very generous.

[MeduSalem]

[...] and we would very quickly reach the incredibly low cap of 300% productivity, [...]

I laughed when I read that. Especially when I read the "incredible low". I can't tell whether you are serious or not. ^^

In the base game you will not be able to hit 300% even if you put legendary tier 3 productivity modules inside a machine.

Currently a tier 3 Productivity module only gives +10% per module. If legendary quality it will be +150% of 10%. 10% + 10%*150% = 25%. So a legendary module will be +25% productivity per module.

And there is no machine, not even the expansion (or at least none was shown yet) that will allow so many modules that you can go all the way up to 300% productivity.

Some of the new machines shown have 50% productivity by default. Some of them have 5 Slots. But still only results in 50%+5*25% = +175% productivity. ^^

At least it is not said that a legendary Machine will have higher default Productivity as well, just higher speed.

And you can't boost productivity using a beacon, so there are no other sources where it could come from.

Currently I see no way to actually reach the 300% cap in the base game or expansion. So I don't even know why they made that cap in the first place if you can't get there. Only gives the wrong impression that you would be able to do that but you can't.



And when it comes to mods, they can raise that 300% cap to whatever ridiculous amount they want so it is not of a concern for them.

[SupplyDepoo]

there are no other sources where it could come from.

FFF 376 # Productivity researches (infinite research)

[MeduSalem]

there are no other sources where it could come from.

FFF 376 # Productivity researches (infinite research)

Ah, thank you. I haven't had read that particular FF.

Then I stand corrected on the matter that one cannot hit the cap. ^^

But even so it is not like it affects all items. They wrote there that it will only benefit particular ones, and none of them you can do a research for currently; except for mining productivity.

So I would still argue it is not worth crying over something we don't have yet. If all those additional infinite researches would be available already and if the 300% cap nerfed people's existing bases to the ground as a result and they lost all the effort they put into it, then they would have a reason to cry. (even though mods could lift the limit anyway so nothing is lost).

Anyway looking at it from the "infinite" research perspective I would actually be glad that you can hit a cap through the research, because it gives incentive to actually use other modules instead of just slapping in PMs everywhere. But that is just my personal take on it, after being tired of seeing endless fields of SM3-beaconized PM3-filled assemblers. Rather replace it with endless fields of EM3-beacons & QM3-filled assemblers. ;D

But if you have to research it for each particular item, it will probably take a while to get all the researches done, even if you can only go up by 30 levels or something before you hit the cap. Only very few people will breeze through it with their 5k SPM bases and those are also the people who will likely use a mod to lift any restrictions soon after anyway. ^^

[Illiander42]

So I would still argue it is not worth crying over something we don't have yet.

The cap isn't in yet either.

We are explicitly talking about the expansion content.

[MeduSalem]

So I would still argue it is not worth crying over something we don't have yet.

The cap isn't in yet either.

You don't say. ^^

We are explicitly talking about the expansion content.

I was just as much talking about the expansion and I don't know what some people expect.
With the expansion in many places we get way, way more productivity than we are currently capped at. The multiplicative factors even of the stuff we will get will already be insane over the crafting cascade compared to now. People should be happy that the devs even considered to increase productivity from where it is currently capped at. But instead some people are already like "give them an inch, and they will take a mile".

Anyway even if the recycler would not become a thing (where we need the cap to prevent a positive-feedback loop from recycling)...

... it would still be ridiculous to have unlimited productivity in the vanilla-expansion. At least I would find it extremely questionable if you can stretch the input materials basically infinitely.
We can't do that in reality either. Even if you reduce or re-use all the waste/clippings, at some point you can't cut more cookies out of the amount of dough available. At least if you still want a cookie with some substance left and not filled with hot air.


So I can only repeat myself... leave the crazy stuff to mods. If you want 100 gear wheels out of 1 iron plate, then a mod can do that.

[Illiander42]

But instead some people are already like "give them an inch, and they will take a mile".

The main issue I personally have with the productivity cap is that there are infinite researches that stop having an effect after a certain level due to a hard-coded cap rather than a natural effect of other factors.

For instance, I don't have a problem that damage research stops being effective once you get it high enough that you 1-shot behemoth worms, because that's a natural cap from game mechanics.

Mining productivity has a natural cap where you can't charge enough robots to empty the box.

I'd be quite happy with the prod cap if the prod researches weren't infinite and stopped at +300%.

It's the combination of a hard cap and infinite researches that go over the cap that bothers me.

[MeduSalem]

The main issue I personally have with the productivity cap is that there are infinite researches that stop having an effect after a certain level due to a hard-coded cap rather than a natural effect of other factors.

For instance, I don't have a problem that damage research stops being effective once you get it high enough that you 1-shot behemoth worms, because that's a natural cap from game mechanics.

Mining productivity has a natural cap where you can't charge enough robots to empty the box.

I'd be quite happy with the prod cap if the prod researches weren't infinite and stopped at +300%.

It's the combination of a hard cap and infinite researches that go over the cap that bothers me.

Well granted, it is not really "infinite" then. And long-story-short... the terminology is just badly chosen.

But I think we are leaving the area for a meta discussion. Because eventually all of the infinite research is not worth it anymore.
Either you hit the point where it has no further effect, as with the mentioned damage one-shotting an enemy.
Or where it is very negligible. Because for every additional level the additional effect is basically less noticeable. Going from 10% to 20% may be still noticable but going from 1000% to 1010% basically won't matter because it is only a relative 1% increase.
Or the cost becoming prohibitively expensive (even if you could get more out of it).
And even if all of that was of no concern, eventually your computer will not be able to handle the growth needed to keep up with research.
So in practice none of them really are "infinite". It is all just a question of "how far before it is not worth it anymore".


For individual productivity research, in the example they showed it was +10% per level, so basically 30 levels are possible before it is not worth it anymore. ^^

I think that is fine with me because you will have to do the productivity research for multiple recipes separately and that takes... well... enough time already to the point I doubt most people will ever reach a completely "maxed out" state where they maxed absolutely every productivity research.
I mean, a couple people definitely will reach it, but those are the types that will not be held back and use a mod just to see how far they could get. But it is not very representative of the majority of the player base.


Anyway they could still re-balance the gains. Not like the values they have shown are necessarily the final ones. If they are cruel and re-balance it such that it just gives +2% productivity per level like Mining Productivity... Then you would have to do 150 levels respectively before hitting the cap (if you don't use PMs that is), at which point the costs may be prohibitively atrocious. But I doubt it is the kind of masochism most people want. At least I know I don't. ^^

[Illiander42]

So in practice none of them really are "infinite". It is all just a question of "how far before it is not worth it anymore".

But for everything else that's actually a question, with some tradeoffs and decisions to make.

ie. Gameplay.

With productivity researches the answer is just "level 30".

[FuryoftheStars]

So in practice none of them really are "infinite". It is all just a question of "how far before it is not worth it anymore".

But for everything else that's actually a question, with some tradeoffs and decisions to make.

ie. Gameplay.

With productivity researches the answer is just "level 30".

My guess for infinite is because they didn't want to have to add 30 individual techs for each of the products that will get the bonus?

[MeduSalem]

So in practice none of them really are "infinite". It is all just a question of "how far before it is not worth it anymore".

But for everything else that's actually a question, with some tradeoffs and decisions to make.

ie. Gameplay.

With productivity researches the answer is just "level 30".

With the other researches it is not a question which level it is either.

Like for all the damage increasing researches. You can calculate the level after which it would be pointless because you would already one-shot even a Behemoth. I think someone even did the math for how many levels for every damage type back in the day after the "infinite techs" came around, just can't remember it anymore.
Just that you can't practically reach it because of the exponential research cost to get there. So we are also stuck within a certain range of maximum research level. And wouldn't make much difference if you have a 1000SPM or 10000SPM base because none can finish the next level in decent time. And that is usually reached after like 10-15 levels already because then your computer would struggle for days & months to get to the next level.

Same for the speed upgrades. You can calculate the point at which research another level would be useless because it would go from A to B almost instantaneously (if there is not a max speed cap programmed into the game in the first place, don't know). Also can't research far because it has same exponential research cost. For worker speed it is the same as damage. Artillery speed scales even worse, so it is over after level 5-10 already.

Even the follower robot count has a max, because eventually you can't click fast enough to spam all the capsules; so the first destroyers would disintegrate before you can spawn the last one necessary to hit the follower cap. I bet that can also be calculated using an empirical value for how fast a human can click. I managed to get around 7 clicks/s. So within 120s (the time it takes for the destroyer to disappear) I can click 840 times (albeit I think either my finger or the mouse will break first). That results in 4200 active bots maximum for me. You get 10 per infinite. So I can do like 400 levels before it is worthless. That one might actually be the only one where you can realistically hit the usefulness cap because it only increases 100 packs per level. Just takes 150 hours with a 1000SPM base. ^^

Mining Productivity is a more complicated topic; there is a breakeven point, where all the materials going into the increasing cost of research eventually outweigh getting the next level of mining productivity and where it would just be cheaper to build more miners instead because it would be less wasteful. But I never did the math on that so I cannot say where that point is and if actually ever pays off to do mining productivity at all or if it is just for the prestige level and because it is the only thing that we actually can research in a decent time frame (next to follower robots).


So what makes it so much different for the upcoming productivity research. Only because it says right there "30 is the limit because more would break the balance of other game mechanics"?

And if they made the progression so horrendous expensive that you can't reach level 30, wouldn't that make it just like the others above? ^^

Even if the cap weren't there, it would likely also suffer the same as Mining Prod. eventually, where the increased research cost overtakes the diminishing return of a few % more productivity. I wouldn't even wonder if that is reached before level 30. Because there you only get relatively seen 3% more productivity AND only for that particular item, but the next level will likely have a relatively seen higher cost increase than 3% on ALL ingredients and not just the one that you made more productive. ^^
At that point it would also only be increasing productivity for the sake of being able to swap out productivity modules to replace them with quality ones or whatever. But not because you would actually be more productive & save on items because of it. ^^

My guess for infinite is because they didn't want to have to add 30 individual techs for each of the products that will get the bonus?

Definitely, because it would be annoying otherwise.



I think they just should rename these techs to "Repeatable Technologies".

Repeatable doesn't necessarily imply that you can do them forever and that there may be a limited amount of times you can repeat them, whether because of practicality reasons or because there is a coded cap. ^^

[Illiander42]

Mining prod is either never worth it or always worth it, depending on your metric.

---

Notice how all your examples are exactly how I described them? The limit is an effective limit based on other gameplay mechanics, not a hard "it doesn't work past here"?

---

For damage upgrades, you have a decision of which breakpoint you consider it not worth it anymore.

1-shot behemoth biters with green ammo from a turret? 1-shot behemoth worms with yellow ammo from an SMG? Those give you very different answers.

For lazers, do you want turrets to 1-shot behemoth biters, or do you want distractor bots to 1-shot behemoth worms?

For mining prod, are you going for filling half a blue belt with speed mods in the miner? A full blue belt with efficiency mods? 1 ore per tick? (Actual max mining speed due to engine limits) Again, these give you very different answers.

For robot speed, the question is "How far do you want it to go in one tick?"

---

You see how each of these makes you answer a question before it tells you what the number is? You have to define what maxxing it out looks like for yourself, and the answer will change from game to game.

That last bit's important, btw. Because it makes the question interesting.

Item prod research always has the same answer.

[FuryoftheStars]

You see how each of these makes you answer a question before it tells you what the number is? You have to define what maxxing it out looks like for yourself, and the answer will change from game to game.

That last bit's important, btw. Because it makes the question interesting.

Item prod research always has the same answer.

Not that I disagree with your stance, but in fairness, the same can be said about this research, too. Are you going for maxing out productivity with the largest machines full of prod modules? Or are you going for maxing out productivity without the use of any modules?

The answer to that, imho, is going to be the same as the rest of the questions you posed: as far as you can conceivably go.

------------------------------------------------------

Same for the speed upgrades. You can calculate the point at which research another level would be useless because it would go from A to B almost instantaneously (if there is not a max speed cap programmed into the game in the first place, don't know).

The Creative mod has "super" bots that can move from the roboport to the destination in 1 tick once in a state where they're ready to move (well, within a range where you could conceivably monitor and measure it, at least). So I'd say no, there is probably no max speed cap in the game beyond the data type the worker bot speed is stored at. (And actually, after typing this, I decided to look in the API and files. The property speed (which is tiles per tick) is a double (2^53), so that's conceivably the true max speed. There is a max_speed property, but from what I can tell, it's not used on vanilla (flying) robot entities.)

My guess for infinite is because they didn't want to have to add 30 individual techs for each of the products that will get the bonus?

Definitely, because it would be annoying otherwise.



I think they just should rename these techs to "Repeatable Technologies".

Repeatable doesn't necessarily imply that you can do them forever and that there may be a limited amount of times you can repeat them, whether because of practicality reasons or because there is a coded cap. ^^

That may not be a bad idea to rename it.


Edit: I happened to decide to look at the technology prototype to see if there was a level defining property on them, and there is: max_level
The description says it's only loaded if the normal and expensive properties are not defined, but from what I can see in the vanilla (pre-2.0, obviously) files, this property is only over used as max_level = "infinite". It's very possible that what we've seen in the FFF is still under development and that come closer to release they may update the productivity researches to be a defined number (like 30 for those that are +10% per level). Or, they could leave them as infinite for mods to have an easier time in adjusting the productivity cap. *shrug*

[MeduSalem]

Not that I disagree with your stance, but in fairness, the same can be said about this research, too. Are you going for maxing out productivity with the largest machines full of prod modules? Or are you going for maxing out productivity without the use of any modules?

Obvious Answer

On a serious note I think we need more hands-on details. More spreadsheet theory-crafting, or more empirical testing. ^^

The answer to that, imho, is going to be the same as the rest of the questions you posed: as far as you can conceivably go.

Anyway, I agree, it will be as far as one is dedicated to go. ^^

But that said, at least I will be looking more deeply into the matter with excel spreadsheet madness when the time comes. Because not every of the productivity researches will be equally worth it. Some will be more so than others due to their relative importance in all the recipes they are involved in. And it will change at various points where doing one type of research will be more worth it over the other due to increasing costs. Definitely something that will be worth my time spending in sheets to find the best path.

And then there is also the matter of quality. Obviously one wants to get that going asap, as not to end up with too much stuff one has to recycle later. So a large portion that previously went into doing Prod. modules will now rather go towards quality modules instead. Will see how that plays out.

The Creative mod has "super" bots that can move from the roboport to the destination in 1 tick once in a state where they're ready to move (well, within a range where you could conceivably monitor and measure it, at least). So I'd say no, there is probably no max speed cap in the game beyond the data type the worker bot speed is stored at. (And actually, after typing this, I decided to look in the API and files. The property speed (which is tiles per tick) is a double (2^53), so that's conceivably the true max speed. There is a max_speed property, but from what I can tell, it's not used on vanilla (flying) robot entities.)

2^53? Holy damn. Well... If Einstein knew about this I am sure he would want to have a word with them. xD

Edit: I happened to decide to look at the technology prototype to see if there was a level defining property on them, and there is: max_level
The description says it's only loaded if the normal and expensive properties are not defined, but from what I can see in the vanilla (pre-2.0, obviously) files, this property is only over used as max_level = "infinite". It's very possible that what we've seen in the FFF is still under development and that come closer to release they may update the productivity researches to be a defined number (like 30 for those that are +10% per level). Or, they could leave them as infinite for mods to have an easier time in adjusting the productivity cap. *shrug*

The base game could go with that 30 cap, so that players don't even waste time attempting to go beyond only to find out it was for nothing.
But I guess it will be more important for modders in the end. They would have to override that limit if they put one there.

[FuryoftheStars]

But that said, at least I will be looking more deeply into the matter with excel spreadsheet madness when the time comes. Because not every of the productivity researches will be equally worth it. Some will be more so than others due to their relative importance in all the recipes they are involved in. And it will change at various points where doing one type of research will be more worth it over the other due to increasing costs. Definitely something that will be worth my time spending in sheets to find the best path.

And then there is also the matter of quality. Obviously one wants to get that going asap, as not to end up with too much stuff one has to recycle later. So a large portion that previously went into doing Prod. modules will now rather go towards quality modules instead. Will see how that plays out.

I'm sure all of that would affect things for different players in different ways. But ultimately, I feel that more so affects the order in which one does them more so than how far on each one.

[MeduSalem]

[...]

You see how each of these makes you answer a question before it tells you what the number is? You have to define what maxxing it out looks like for yourself, and the answer will change from game to game.

That last bit's important, btw. Because it makes the question interesting.

Yea, I know the questions are "there", but they are just for having theoretical discussions on the forum-meta-game. These questions have no practical applications in the game. xD

Because... the truth is that the answers to these questions don't really change from game to game. One will run into the same prohibitive research cost wall every time after which you just let it be. (unless you mod the game and change certain aspects of it)

For playing the game I have become more pragmatic because of it; I played too much & did enough theory crafting. I have spreadsheets for the time lengths for various SPM amounts and I just have my fixed cut-off points because I know I cannot realistically get past a certain research wall without wasting an enormous amount of time I am not willing to waste anymore. ^^

I already accepted that one cannot win against the exponential development of research costs; and they basically made asking these questions above "meh". For me in practice the "infinite" research simply is not "infinite" because of the time factor involved. So they act just like hard caps in practice.