“When the foot seeks the place of the head, the sacred line is crossed. Know your place. Keep your place. Be a shoe.” – Mason
In Stellaris, there are multiple ways to solve various labor problems. For many civilizations, you need just enough food to support your biological POPs, while devoting the rest of your planetary tiles to more productive pursuits: energy, minerals, research, or special buildings. In some cases, you may also employ various kinds of thinking machines: robots, droids, and synthetics. Robots and droids have no happiness, but they have serious disadvantages when it comes to energy and research. Synthetics are much better than their weaker cousins, but they come with other associated disadvantages.
In many cases, what kind of POP you should use for what task is largely contextual. Between species traits, abundance or lack of certain resources (Am I swimming in credits? Lack food? Need more research?), the decision to build thinking machines is complicated and there are no easy answers. The right answer in 2230 might not even be close to the right answer 20 years later after the galaxy changes.
For non-slaving empires, happiness is an important factor in certain kinds of productivity. Non-slaving empires that conquer lots of unhappy POPs might find that building droids and robots meets their needs on some planets, since these thinking machines do not have happiness. With no slavery option, increasing productivity via oppression is not possible, so creating a deliberate underclass of mechanical servants makes sense in some contexts. There is also the issue of droids being useful for colonization, but I want to focus this discussion just on productivity; obviously, thinking machines have uses beyond production.
For slaving empires, the decision whether to use more slaves or thinking machines is much more important to answer. Slaving empires (whether they just enslave filthy xenos or are collectivists who enslave even their own kind) face a dilemma. On the one hand, slaves seem free; you grow them, and they work your tiles. Thinking machines cost minerals and energy, while also costing energy to maintain. Yet, thinking machines do not require the same kind of infrastructure as slavery, and thinking machines seem better at mining, one of the primary tasks of slaves. In the words of Snowpiercer, the sacred question for slaving empires is, “Who is the best shoe?” Biologicals or thinking machines?
The answer is not obvious at all, because it turns out both slavery and thinking machines have hidden costs that cloud the topic, as well as hidden benefits that only come into play if you pay really close attention to certain tooltips.
It is important to discuss the production impact of each kind of POP. For ease of discussion, for this first part, we will ignore species traits and focus only on droids. Synthetics and species traits will be discussed later. As robots are inferior in every way to droids, just assume for the purposes of this discussion that robots are droids that cost the same, but are less useful.
Droids get a +15% bonus to mineral production. They also get a -20% penalty to energy, while having a -40% penalty to research. They have no bonus or penalty to food.
Slaves get a +20% bonus to food and minerals. They get a -33% penalty to energy, and a whopping -75% penalty to research. Note that the research penalty applies to labs and any building with research output, including clone vats. Note also that this does not take into account slave processing facilities; as this building requires a tile, its use is situational.
For comparison: POPs with 0% happiness take a -50% penalty to mining and energy, and a -25% penalty to research. Interestingly, they get no penalty to food. (Bug? WAD?) Note that droids and slaves have no happiness impact on production (good or bad).
From this summary, it appears that slaves are better at mining and farming, while droids are better at energy and research. At this point, someone is going to point out that using either droids or slaves to harvest energy or research is stupid, since the penalties for both are substantial. But if you look at unhappy POPs, there is little reason to keep unhappy POPs free if you have the option of either enslaving them or purging them and replacing them with droids except for research. If you replace them with droids, you don’t need as much food, and if you enslave them, they make far more food.
This seems to weight the production of resources in favor of slaves, while research is best done by either unhappy, but free, POPs or droids. But there are costs associated with droids and slavery.
Slaves require food to grow. Food comes from either farms or orbital hydroponic farms. Farms and orbital modules require energy credits to function properly. A tier 1 farm costs 1 energy a month and generates 2 food (unless otherwise noted, all discussion of buildings assumes no tile resources in play). A tier 3 farm generates 4 food at the cost of 2 energy. A level 4 farm generates 5 food for 2.5 energy. With the exception of the special tier 5 farms (can only be built on a planet with empire capitals) and basic farms (crappy farms on new planets), all farms generate 2 food per energy credit. The orbital hydroponic farm generates 3 food for 1 energy. Planetary capitals generate 2, 3, 4 or 5 food for no energy cost.
Note that farms have a capital cost in terms of minerals, but since all buildings do, this is not something about which we should worry right now.
These costs mean that the first 2-5 POPs on a planet cost food, but that the food cost is irrelevant since it is covered by the capital. Additional POPs cost either 0.5 energy per POP (due to farm energy costs) or .3333 energy per POP (orbital farm).
Droids do not cost food. In fact, a planet with no biological POPs at all requires no farms and the food in the capital will be wasted. Droids do cost 1 energy per POP, in addition to their creation cost of 150 energy.
This makes it appear that slaves are far superior to droids for minerals and food. The energy cost of food is less than the energy cost of droids. But where it gets complicated is that, since droids require no food, a planet with pure droids can devote more tiles to resource production. This means more mines, and more mines stack with mineral processing facilities and the production targets edict to create more minerals. Take the same planet, and cover it with droids, and you might pay more energy, but you can hypothetically churn out more minerals while spending the exact same influence for the production targets edict.
So, now it seems like droids, when used on special droid planets, are superior at mining. But that’s not the whole story. Slaves need food to grow. Once you fill a planet, they don’t need food anymore. Starvation only hurts happiness; slave production is not affected by happiness. When a planet is full, you can do what I have done here:
The name of this planet is a dead giveaway as to just how horrible of a place it is. The only food comes from the capital which, thanks to adjacency bonuses and inherent mining, does not get replaced by a mine. This planet is also large enough to justify the use of a slave processing facility on a tile with no minerals to boost mineral output by 10% across the board. (It also boosts food, but I don’t care in this case.)
There are some other costs, though. With no food, if a POP is lost, it won’t grow back. You will have to resettle a POP from a breeder world at the cost of influence to replace them. This is not usually a problem, but there are some situations where instead of paying food, you end up paying influence to support your slaving habit.
Some players will immediately object that no one needs that much mineral production due to the need for energy credits, but that takes us to the next point: are slaves or droids better at energy production?
Droids seem better at energy by a small margin. Slaves take a bigger penalty with no buildings that can improve their energy production. But because droids cost energy to maintain, they more or less tax energy production at 1 energy per tile. If you are running tier 4 energy plants, that tax is 16.666% of your energy production from that building. That stacks with the 20% penalty droids have for a 36.66% penalty (again, assuming no resources on the tile), making them worse than slaves by a small margin. (This is a good time to point out that with lower tiers of energy plants, the tax in question is much harsher.) Note that both are terrible at energy, but when you are conquering enemies, you don’t always have the option of just converting every planet you conquer and enslave into mineral production. There will be times when either slaves or droids need to be employed in energy, and if a planet is already filled with POPs, enslaving them and keeping them in energy production is better than purging them and replacing them with droids. There are no planets in the galaxy that suck so bad that they cannot be used to generate energy credits to support the empire, even if the planet is resource poor. Blank tiles default to energy fairly well if we don’t need to worry about food.
For research, droids are marginally better even with their energy cost than slaves, but unlike energy, research is the kind of thing that you should only be doing on happy planets with the right tiles; that assumes you even do research on planets.
At the moment, slaves seem to enjoy a clear edge in comparison to droids. There does not seem to be a situation where slaves are not a better choice once a planet has fully grown. But there are disadvantages to slaves that go beyond energy and minerals. Malcontent slaves cost influence to suppress. You can easily be paying 1.5 or more influence a month in suppression once you have a high percentage of malcontent slaves; you will end up with plenty of malcontent slaves as you enslave those who do not share your ethos (unless you purge half the galaxy for some reason). This assumes you don’t spend influence to suppress emancipation movements and docile slaves; doing so adds even more influence costs to such a schema.
But what if you don’t suppress the malcontents? There is no obligation to do so; one can choose instead to save the influence cost. Doing so will require more… brutality. Malcontent slaves that are not suppressed will begin to succumb to the Interstellar Railroad. This will periodically delete slave POPs. On normal planets with proper food, this is more of an inconvenience than a real problem; on planets given the Glavnoye Koloniy treatment, it means paying a small sum of influence to move a new POP from a breeder world that has enough food to grow new POPs (or building a droid to replace it).
Beyond the Interstellar Railroad, there is the problem of a slave revolt. Revolting slave POPs, while not the equal of assault armies, are far more potent than the free garrison troops that spawn when a planet is attacked. On larger worlds, slave revolts will spawn enough armies to be a serious threat; thus, if you choose not to suppress factions, you will need to garrison worlds that contain significant numbers of slaves. Slave armies are themselves sufficient to this task; each slave army used will cost 0.20 energy a month. (Obviously, other armies are more or less expensive, and performance in ground combat varies widely.)
This means that if you choose not to pay the influence cost, you will have to pay the energy cost of garrisons. Even if you put 12 slave armies on a planet, that’s only going to cost 2.4 energy a month. This is still cheaper than droids, but it narrows the difference. It also means you have to take the trouble to garrison worlds.
Another problem with slaves compared to droids is the slave unrest event that lowers productivity. Suppressing docile slaves should stop this event, but that costs influence. If you choose not to pay the influence cost, you will face penalties to production on planets where it fires. This means that those who refuse to pay the influence cost will face uneven production as some planets face production issues. This might make budgeting slightly more difficult.
These hidden costs do not make droids necessarily a better option than slaves, but they complicate things. For societies that possess both slaves and droids, the long term choice may well depend on what resource you think is more precious. Those short on energy might find droids unacceptable; those who find influence to be more valuable might feel differently.
But what about species traits? Do these influence the choice between thinking machines and biologicals?
Biological slave POPs with productivity traits are even better than thinking machines at all tasks that relate to resource production. Research is different; because of how exclusive the natural engineers/sociologists/physicists traits are, it’s not possible to stack the traits in a way to overcome the 35% worth of research penalties from being slaves (Intelligent grants +10% across the board, and the other traits each only grant +15%).
The Weak trait reduces mineral output, but only by 5%. I’m not sure it’s enough to make a difference in the overall analysis. Slaves and droids still have the same costs in general.
This leaves us with one other important consideration. Do synthetics change anything?
What makes synthetics different is that they can be enslaved. They have their inherent bonuses, but they also can benefit from the slavery bonuses.
This means that, under some circumstances, synthetics are actually better than biological slaves. But some species are better at mining minerals than even synthetics. Industrious adds +15% to mineral production, while very strong adds +10%. While this is a lot of genetic effort to get bonuses to just mineral production, it does mean that slave species with the right combination of traits even beat out the best mechanical beings have to offer.
So, have I missed anything in my analysis?
In Stellaris, there are multiple ways to solve various labor problems. For many civilizations, you need just enough food to support your biological POPs, while devoting the rest of your planetary tiles to more productive pursuits: energy, minerals, research, or special buildings. In some cases, you may also employ various kinds of thinking machines: robots, droids, and synthetics. Robots and droids have no happiness, but they have serious disadvantages when it comes to energy and research. Synthetics are much better than their weaker cousins, but they come with other associated disadvantages.
In many cases, what kind of POP you should use for what task is largely contextual. Between species traits, abundance or lack of certain resources (Am I swimming in credits? Lack food? Need more research?), the decision to build thinking machines is complicated and there are no easy answers. The right answer in 2230 might not even be close to the right answer 20 years later after the galaxy changes.
For non-slaving empires, happiness is an important factor in certain kinds of productivity. Non-slaving empires that conquer lots of unhappy POPs might find that building droids and robots meets their needs on some planets, since these thinking machines do not have happiness. With no slavery option, increasing productivity via oppression is not possible, so creating a deliberate underclass of mechanical servants makes sense in some contexts. There is also the issue of droids being useful for colonization, but I want to focus this discussion just on productivity; obviously, thinking machines have uses beyond production.
For slaving empires, the decision whether to use more slaves or thinking machines is much more important to answer. Slaving empires (whether they just enslave filthy xenos or are collectivists who enslave even their own kind) face a dilemma. On the one hand, slaves seem free; you grow them, and they work your tiles. Thinking machines cost minerals and energy, while also costing energy to maintain. Yet, thinking machines do not require the same kind of infrastructure as slavery, and thinking machines seem better at mining, one of the primary tasks of slaves. In the words of Snowpiercer, the sacred question for slaving empires is, “Who is the best shoe?” Biologicals or thinking machines?
The answer is not obvious at all, because it turns out both slavery and thinking machines have hidden costs that cloud the topic, as well as hidden benefits that only come into play if you pay really close attention to certain tooltips.
It is important to discuss the production impact of each kind of POP. For ease of discussion, for this first part, we will ignore species traits and focus only on droids. Synthetics and species traits will be discussed later. As robots are inferior in every way to droids, just assume for the purposes of this discussion that robots are droids that cost the same, but are less useful.
Droids get a +15% bonus to mineral production. They also get a -20% penalty to energy, while having a -40% penalty to research. They have no bonus or penalty to food.
Slaves get a +20% bonus to food and minerals. They get a -33% penalty to energy, and a whopping -75% penalty to research. Note that the research penalty applies to labs and any building with research output, including clone vats. Note also that this does not take into account slave processing facilities; as this building requires a tile, its use is situational.
For comparison: POPs with 0% happiness take a -50% penalty to mining and energy, and a -25% penalty to research. Interestingly, they get no penalty to food. (Bug? WAD?) Note that droids and slaves have no happiness impact on production (good or bad).
From this summary, it appears that slaves are better at mining and farming, while droids are better at energy and research. At this point, someone is going to point out that using either droids or slaves to harvest energy or research is stupid, since the penalties for both are substantial. But if you look at unhappy POPs, there is little reason to keep unhappy POPs free if you have the option of either enslaving them or purging them and replacing them with droids except for research. If you replace them with droids, you don’t need as much food, and if you enslave them, they make far more food.
This seems to weight the production of resources in favor of slaves, while research is best done by either unhappy, but free, POPs or droids. But there are costs associated with droids and slavery.
Slaves require food to grow. Food comes from either farms or orbital hydroponic farms. Farms and orbital modules require energy credits to function properly. A tier 1 farm costs 1 energy a month and generates 2 food (unless otherwise noted, all discussion of buildings assumes no tile resources in play). A tier 3 farm generates 4 food at the cost of 2 energy. A level 4 farm generates 5 food for 2.5 energy. With the exception of the special tier 5 farms (can only be built on a planet with empire capitals) and basic farms (crappy farms on new planets), all farms generate 2 food per energy credit. The orbital hydroponic farm generates 3 food for 1 energy. Planetary capitals generate 2, 3, 4 or 5 food for no energy cost.
Note that farms have a capital cost in terms of minerals, but since all buildings do, this is not something about which we should worry right now.
These costs mean that the first 2-5 POPs on a planet cost food, but that the food cost is irrelevant since it is covered by the capital. Additional POPs cost either 0.5 energy per POP (due to farm energy costs) or .3333 energy per POP (orbital farm).
Droids do not cost food. In fact, a planet with no biological POPs at all requires no farms and the food in the capital will be wasted. Droids do cost 1 energy per POP, in addition to their creation cost of 150 energy.
This makes it appear that slaves are far superior to droids for minerals and food. The energy cost of food is less than the energy cost of droids. But where it gets complicated is that, since droids require no food, a planet with pure droids can devote more tiles to resource production. This means more mines, and more mines stack with mineral processing facilities and the production targets edict to create more minerals. Take the same planet, and cover it with droids, and you might pay more energy, but you can hypothetically churn out more minerals while spending the exact same influence for the production targets edict.
So, now it seems like droids, when used on special droid planets, are superior at mining. But that’s not the whole story. Slaves need food to grow. Once you fill a planet, they don’t need food anymore. Starvation only hurts happiness; slave production is not affected by happiness. When a planet is full, you can do what I have done here:
The name of this planet is a dead giveaway as to just how horrible of a place it is. The only food comes from the capital which, thanks to adjacency bonuses and inherent mining, does not get replaced by a mine. This planet is also large enough to justify the use of a slave processing facility on a tile with no minerals to boost mineral output by 10% across the board. (It also boosts food, but I don’t care in this case.)
There are some other costs, though. With no food, if a POP is lost, it won’t grow back. You will have to resettle a POP from a breeder world at the cost of influence to replace them. This is not usually a problem, but there are some situations where instead of paying food, you end up paying influence to support your slaving habit.
Some players will immediately object that no one needs that much mineral production due to the need for energy credits, but that takes us to the next point: are slaves or droids better at energy production?
Droids seem better at energy by a small margin. Slaves take a bigger penalty with no buildings that can improve their energy production. But because droids cost energy to maintain, they more or less tax energy production at 1 energy per tile. If you are running tier 4 energy plants, that tax is 16.666% of your energy production from that building. That stacks with the 20% penalty droids have for a 36.66% penalty (again, assuming no resources on the tile), making them worse than slaves by a small margin. (This is a good time to point out that with lower tiers of energy plants, the tax in question is much harsher.) Note that both are terrible at energy, but when you are conquering enemies, you don’t always have the option of just converting every planet you conquer and enslave into mineral production. There will be times when either slaves or droids need to be employed in energy, and if a planet is already filled with POPs, enslaving them and keeping them in energy production is better than purging them and replacing them with droids. There are no planets in the galaxy that suck so bad that they cannot be used to generate energy credits to support the empire, even if the planet is resource poor. Blank tiles default to energy fairly well if we don’t need to worry about food.
For research, droids are marginally better even with their energy cost than slaves, but unlike energy, research is the kind of thing that you should only be doing on happy planets with the right tiles; that assumes you even do research on planets.
At the moment, slaves seem to enjoy a clear edge in comparison to droids. There does not seem to be a situation where slaves are not a better choice once a planet has fully grown. But there are disadvantages to slaves that go beyond energy and minerals. Malcontent slaves cost influence to suppress. You can easily be paying 1.5 or more influence a month in suppression once you have a high percentage of malcontent slaves; you will end up with plenty of malcontent slaves as you enslave those who do not share your ethos (unless you purge half the galaxy for some reason). This assumes you don’t spend influence to suppress emancipation movements and docile slaves; doing so adds even more influence costs to such a schema.
But what if you don’t suppress the malcontents? There is no obligation to do so; one can choose instead to save the influence cost. Doing so will require more… brutality. Malcontent slaves that are not suppressed will begin to succumb to the Interstellar Railroad. This will periodically delete slave POPs. On normal planets with proper food, this is more of an inconvenience than a real problem; on planets given the Glavnoye Koloniy treatment, it means paying a small sum of influence to move a new POP from a breeder world that has enough food to grow new POPs (or building a droid to replace it).
Beyond the Interstellar Railroad, there is the problem of a slave revolt. Revolting slave POPs, while not the equal of assault armies, are far more potent than the free garrison troops that spawn when a planet is attacked. On larger worlds, slave revolts will spawn enough armies to be a serious threat; thus, if you choose not to suppress factions, you will need to garrison worlds that contain significant numbers of slaves. Slave armies are themselves sufficient to this task; each slave army used will cost 0.20 energy a month. (Obviously, other armies are more or less expensive, and performance in ground combat varies widely.)
This means that if you choose not to pay the influence cost, you will have to pay the energy cost of garrisons. Even if you put 12 slave armies on a planet, that’s only going to cost 2.4 energy a month. This is still cheaper than droids, but it narrows the difference. It also means you have to take the trouble to garrison worlds.
Another problem with slaves compared to droids is the slave unrest event that lowers productivity. Suppressing docile slaves should stop this event, but that costs influence. If you choose not to pay the influence cost, you will face penalties to production on planets where it fires. This means that those who refuse to pay the influence cost will face uneven production as some planets face production issues. This might make budgeting slightly more difficult.
These hidden costs do not make droids necessarily a better option than slaves, but they complicate things. For societies that possess both slaves and droids, the long term choice may well depend on what resource you think is more precious. Those short on energy might find droids unacceptable; those who find influence to be more valuable might feel differently.
But what about species traits? Do these influence the choice between thinking machines and biologicals?
Biological slave POPs with productivity traits are even better than thinking machines at all tasks that relate to resource production. Research is different; because of how exclusive the natural engineers/sociologists/physicists traits are, it’s not possible to stack the traits in a way to overcome the 35% worth of research penalties from being slaves (Intelligent grants +10% across the board, and the other traits each only grant +15%).
The Weak trait reduces mineral output, but only by 5%. I’m not sure it’s enough to make a difference in the overall analysis. Slaves and droids still have the same costs in general.
This leaves us with one other important consideration. Do synthetics change anything?
What makes synthetics different is that they can be enslaved. They have their inherent bonuses, but they also can benefit from the slavery bonuses.
This means that, under some circumstances, synthetics are actually better than biological slaves. But some species are better at mining minerals than even synthetics. Industrious adds +15% to mineral production, while very strong adds +10%. While this is a lot of genetic effort to get bonuses to just mineral production, it does mean that slave species with the right combination of traits even beat out the best mechanical beings have to offer.
So, have I missed anything in my analysis?
- 2
- 1
