Effective manpower cost (model calculation)

[YertyL]

So, motivated by the discussion of fusion reactors and the cost of manning buildings, I tried to do a little model calculation: If you treat manpower like any other power source, i.e. you want complete self-sustenance, how much do you need to invest for each free qualified worker?

Checking https://survivingmars.paradoxwikis.com/Colonists, colonists are children at 0-5 and seniors at 61+. I could not find any data on average age of death, so I am going to assume it's 80. So imagine we want a closed system of 80 colonists, and each sol, one dies and one is born. Assuming max shifts for all buildings.

• Children and seniors: 5 colonists will be children, 20 seniors. Effective workforce: 80-25=55
• Basic comfort: You need at least 66.4 average comfort for sustained births. 3 diners will fulfill social and meal for 3*10*3=90 colonists max, 3 grocers shopping and meal for 3*8*3=72, so I am going to take that as base comfort cost. 3 diners are 18 workers, 3 grocers 9. Assume you get some parks for relaxation and ignore other needs. Effective workforce: 55-27=28.
  
• Food and healthcare: A wheat farm with 6 botanists at 50 morale and 50% soil will feed 40 people, one at 100% soil 70, hydro and fungal are way worse. I am going to roughly estimate 1.33 farms needed, so 8 workers. You will also want at least 1 clinic with 2 shifts, so 4 more workers. Effective workforce: 28-(8+4)=16.
• Education: Of 55 workers, 27 can do unqualified jobs (service), so you will want about half your people to go to university. Basic time for uni education is 15 sols. Every sol, 1 person is born, and half of them go to university for 15 sols, so (rounding) 8 people will be at university at all time. Effective workforce: 16-8=8 qualified workers.

This seems pretty insane, and please correct me if I am wrong, but in this model calculation, if 80 people want to sustain themselves in the long run, only 8(!) of them will be qualified workers free to work somewhere else. This will go slightly up with better performance at the farms, but I did not calculate stuff like sanatorium, security, more costly needs, suboptimal food sources, waste, disasters etc. Now, how much do you need to invest in materials?

Space needed: again, rough calculation with basic buildings:

• 6 living complexes to house 6*14= 84 people for 6 large squares
  
• 3 grocers, 3 diners for ~2 large squares
  
• 1/3 university for 8 adults, 1/3 school for 5 children, and 1 clinic for ~ 1 large square
• 1.33 farms for 1.33 large squares.

So now we are at 10.33 large squares, but since we so far disregarded parks, which would optimally need 80 hexes a.k.a 8 large squares for relaxation, dome connections, and production buildings, I am just going to round up to 1 medium dome which offers 12 large squares and 3 single hexes. Water and oxygen are roughly 1 MOXIE and 4 vaporators for 1.33 wheat farms and a medium dome = ~ 4 water. You will want at least 1 oxygen and 2 water tanks as well.

Material cost: Again, roughly.

• Concrete: Medium dome 120, 6 living complexes 36, 3 grocers 15, 3 diners 18, 1/3 uni 10, 1/3 school 7, clinic 10, 1.33 farms ~11, gardens very roughly 10-20 ish, so ballpark 240 concrete. Upkeep roughly 6 + 3 +3 =12 concrete.
  
• Metal: dome 50, 3 diners 15, 1 MOXIE 4, 4 vaporators 8, tanks 3*3=9, so 86. Upkeep 2 + 8 + 3 = 13 metal
• Poly: dome 20, vaporators 20 = 40, 45 with playground. Upkeep 2
• Electronics: 5 for 1/3 uni, 2 for 1/3 school. Upkeep 1.6.
  
• power: Dome 30, diners 6, 1/3 uni 5, MOXIE 2, 4 vaporators 20, so 63 power.

So, again, this seems kind of insane to me, but this was for a qualified, free workforce of 8 workers.

So without techs and in the long run, the upkeep of each qualified worker in factories etc. will cost you about 1.5 concrete, 1.5 metal, 8 power, and ~ 0.2 poly and electro, with an initial investment of roughly 31 concrete, 11 metal, 5 poly and 1 electronics plus power plants per worker.

Other takeaway: 1 school 1 uni per ~240 stable population, obviously more if you want to grow. Martian education halves the time for uni education, which means 1 uni per 480 people, but more importantly, 4 more workers in your effective workforce, and a cut of effective worker upkeep by 33%. Farm automation will free 3 more workers, so those 2 techs almost halve your effective worker upkeep. They are very, very good.
This also means that initially, fusion reactors only pay for themselves in pure power upkeep (200 power for 24*8=192 power upkeep) while being a huge sink in ressource upkeep, of course, which is pretty hilarious to me. They are basically mid-game workshops.

 

[ChadPeterson]

While I agree with your assertion that manpower is expensive, it is not as bad as you lay out, even with no techs unlocked.

Primarily it is that your assumptions are oversupplying Basic Comfort services, which are not required to the extent that you lay out. Many savings can take place here which will make a large difference in the overall costs of manpower.

First, the seniors do not need to be provided with comfort services, as their job performance does not matter. They can be moved to another dome and eat unprepared meals just fine. This reduces your colonists that require comfort to only 60.

Second, the population only spends about 1/2 of each shift in a service building, so the service buildings can supply roughly twice the numbers you calculated. A diner therefore can supply 3*(10*2) = 60 people per day. And a grocer 3*(8*2) = 48 per day. This might not seem like enough, but recall that colonists can only eat once per day, so they will not visit both a diner and a grocer on the same day, so 1 diner and 1 grocer can theoretically handle 108 colonists, but 60 is more practical as there are some vacancies.

• Changing to 1 diner and 1 grocer reduces the Basic Comfort services manpower cost from your estimate of diner(18)+grocer(9)=27 to diner(6)+grocer(3)=9.

Third, it is not necessary to have the highest service level in these service buildings. They are "basic" comfort and can therefore run on 1 worker per shift and still serve the same number of colonists. This also applies to the clinic which only needs 1 worker on each of the three shifts.

• This reduces the manpower involved in basic services and health care from diner(18)+grocer(9)+clinic(4)=31 to diner(3)+grocer(3)+clinic(3)=9.

Moving from 31 colonists allocated to supplying basic services to the much leaner number of 9 frees up 22 colonists for more productive activities. This increases the colony from having 8 productive individuals to 30 of them, or an output multiplier of 3.75!

 

[Xeorm]

Don't think you'd need so many grocers and diners. Forgive me if I'm entirely wrong, but you should only need enough slots in total as each time they'll choose a single need to try and fulfill. So you'd need 3 buildings in total, which matches with how well my own colonies tend to do. Grocer, diner, and bar should take care of them, which means only 15 workers in service buildings. 12 more workers.

A farm doing potatoes+soybeans at 100% soil+overtime produces a bit over 80 people worth of food. We have the infirmary, might as well get some use out of it and overwork that farm. That'll save another 2 workers. Total 14.

For training you'd want to look at your replacement for specialists. Every year one colonist leaves the workforce. What is the chance this worker was a specialist and not one of the non-specialized? We have 55 total workers, of which 15 are non-specialists. 40/55 = 72%. They'll spend 15 of their 55 sols in training, or 27% of their time there. Therefore 19% of the workforce is training at any one time, or ~10. So you'll have 20 working productively, 15 in service, 6 in food, 4 as medics, 10 as students, and 25 unable to work.

That's not terrible I'd think. Add on additional techs like you mentioned and you'd have 8 more workers. Stem reconstruction would help too to increase the time people spend working instead of training.

But yea. You've got way too many people in service buildings.

 

[YertyL]

Don't think you'd need so many grocers and diners. Forgive me if I'm entirely wrong, but you should only need enough slots in total as each time they'll choose a single need to try and fulfill. So you'd need 3 buildings in total, which matches with how well my own colonies tend to do. Grocer, diner, and bar should take care of them, which means only 15 workers in service buildings. 12 more workers.

A farm doing potatoes+soybeans at 100% soil+overtime produces a bit over 80 people worth of food. We have the infirmary, might as well get some use out of it and overwork that farm. That'll save another 2 workers. Total 14.

For training you'd want to look at your replacement for specialists. Every year one colonist leaves the workforce. What is the chance this worker was a specialist and not one of the non-specialized? We have 55 total workers, of which 15 are non-specialists. 40/55 = 72%. They'll spend 15 of their 55 sols in training, or 27% of their time there. Therefore 19% of the workforce is training at any one time, or ~10. So you'll have 20 working productively, 15 in service, 6 in food, 4 as medics, 10 as students, and 25 unable to work.

That's not terrible I'd think. Add on additional techs like you mentioned and you'd have 8 more workers. Stem reconstruction would help too to increase the time people spend working instead of training.

But yea. You've got way too many people in service buildings.

I have to admit I still do not fully understand comfort mechanics, but if I supply a dome with 4 living complexes (56 people) with only 1 diner and 1 grocer, I immediately start getting "service building was full" messages in the double digits. If I supply with 2 grocers + 1 diner or 2 diners + 1 grocer, the messages are in the 5-10 range. On the other hand, even 2 diners 2 grocers are often close to, or at, max capacity. You can replace 1 diner with a gym, but that takes more space, and it made comfort still drop below 70 for me (just tried this).

Neatly seperating seniors , and undersupplying them with services is something I found pretty difficult in practice, since your senior dome cannot be fully seniors if you do not want them to die. Plus I calculated infirmary shifts very conservatively, so I don't think the numbers are way off in practice.

EDIT: The "minimal service workers" approach with 1 gym, 1 grocer and 2 small parks for 28 people produced an average comfort of 60 despite a residence comfort of 65. I think it is safe to say you need at least a diner in the long run if you do not have some other high comfort source.

 

[Xeorm]

I have to admit I still do not fully understand comfort mechanics, but if I supply a dome with 4 living complexes (56 people) with only 1 diner and 1 grocer, I immediately start getting "service building was full" messages in the double digits. If I supply with 2 grocers + 1 diner or 2 diners + 1 grocer, the messages are in the 5-10 range. On the other hand, even 2 diners 2 grocers are often close to, or at, max capacity. You can replace 1 diner with a gym, but that takes more space, and it made comfort still drop below 70 for me (just tried this).

Neatly seperating seniors , and undersupplying them with services is something I found pretty difficult in practice, since your senior dome cannot be fully seniors if you do not want them to die. Plus I calculated infirmary shifts very conservatively, so I don't think the numbers are way off in practice.

My understanding of the system is that it works something like: Everyone picks a need they want to satisfy. Then slot themselves. So which need they go for depends on a random roll for that day. This also implies that getting everyone perfect will leave a lot of extra slots. Which is also ok. They don't need to have max comfort either. You can still have very happy colonists even as they fail to find the right building. As long as on average they gain comfort they're ok.

Like I'm looking now at my terribly overpopulated colony and I have one medium dome with 133 people, mostly apartments, that had 37 full errors and 5 missing errors. Still has 70 average comfort. One diner and a grocery. That's way overcapacity and not what I'd do ideally but it does show the limits of what you can get.

They should be able to survive as long as they have some source of food. This doesn't need to be supplied in a diner or anything. Unprepared meals should work ok, as long as there's a stockpile nearby. They won't be happy, but that's alright.

 

[YertyL]

I just created an isolated basic dome with 2 living complexes, 1 grocer, 1 gym, 2 small parks, and exactly 28 people to test this. All 3 services are constantly close to, or at, max capacity despite not everyone having shopping/social/relax interest because of mixed specializations, and I get small numbers of "are full" complaints for all 3. Service buildings providing for (capacity*shifts) people just seems empirically true.
Average comfort also varies between 57 and 65.

 

[Xeorm]

I just created an isolated basic dome with 2 living complexes, 1 grocer, 1 gym, 2 small parks, and exactly 28 people to test this. All 3 services are constantly close to, or at, max capacity despite not everyone having shopping/social/relax interest because of mixed specializations, and I get small numbers of "are full" complaints for all 3. Service buildings providing for (capacity*3) people just seems empirically true.
Average comfort also varies between 57 and 65.

How long did it run? Did you try doing the same, but with more buildings? If the numbers were that small my overpopulated dome should have seen way more errors than it did.

 

[ChoGGi_2]

death age is (MinAge_Senior is 61):
self:SetBase("death_age", self.MinAge_Senior + 5 + self:Random(10) + self:Random(5) + self:Random(5))
66-86

 

[Xeorm]

death age is (MinAge_Senior is 61):
self:SetBase("death_age", self.MinAge_Senior + 5 + self:Random(10) + self:Random(5) + self:Random(5))
66-86

Average is still 80, so that's nice. Thanks.

 

[YertyL]

How long did it run? Did you try doing the same, but with more buildings? If the numbers were that small my overpopulated dome should have seen way more errors than it did.

A few sols. Specific capacity and number of "full" messages varies -- it does not go to 0 though, and is almost always highest for relaxation and lowest for social, which makes sense with the 6 vs 10 capacity.
I also tried it with barrel domes with 56 people and combinations of 1-2 grocers, 1-2 diners, and 0-1 gyms. I'm pretty sure the rule of thumb is right. I believe service buildings can serve 2* capacity in a shift, but colonists will also visit 2 service buildings (?)

If those self:Random distributions are uniform 0-(value), average is 76 though -- but that does not change exactly a lot in the approximation above. 5% lower overall need for food, space and services, but that's about it.

 

[Xeorm]

A few sols. Specific capacity and number of "full" messages varies -- it does not go to 0 though, and is almost always highest for relaxation and lowest for social, which makes sense with the 6 vs 10 capacity.
I also tried it with barrel domes with 56 people and combinations of 1-2 grocers, 1-2 diners, and 0-1 gyms. I'm pretty sure the rule of thumb is right. I believe service buildings can serve 2* capacity in a shift, but colonists will also visit 2 service buildings (?)

If those self:Random distributions are uniform 0-(value), average is 76 though -- but that does not change exactly a lot in the approximation above. 5% lower overall need for food, space and services, but that's about it.

Looking at comfort histories it looks like 2 visits each day. So yah. Probably 3*capacity seems valid for mostly full coverage. Though going by the numbers I see, -10 for a miss and +10-+20 for a hit implies being able to solidly increase the coverage. Should be able to manage higher capacity as quality improves.

 

[YertyL]

So I tested these two domes over the course of about 5 sols:


All non-specialized non-child colonists, isolated domes etc. The left dome has 28 people and 10,8 and 10 max visitors for social, shopping, relaxation. You get consistent small complaints about full shopping and occasional small complaints about full other services. Buildings are all always mostly full, and if you check "visitors today" at midnight, the max number of visitors in a day seems to be 2*shifts*(max inside), which is consistently almost achieved for all buildings. For some bizzare reason, the "visitors today" for the large park seem to cap at 4*(max inside) instead of 6, but otherwise it behaves like the gym, i.e. consistently just about satisfies 28 people ( 3* max visitors in general).
If you add total numbers of visitors today for all buildings in the small dome with 28 people, it's very close to 28*4, which tells me that every colonist actually visits 4 random services each day. Since each service can satisfy 2*max visitors each shift, we get the 3* max visitors for each interest rule of thumb at max shifts.
In the right dome, despite only 1 large park instead of 2 for ~60 people, complaints about missing shopping are actually higher than those about missing relaxation, though this might be because I trained "workaholic" in schools, which means - relaxation interest. There is some randomness here though, relaxation complaints still tend to be highest on average.
Average comfort for left dome sank to 66, for the right to 70. Residence comfort is 65.

EDIT: For a medium dome with 84 people, I can get just above 66 comfort with 2 diners and 2 grocer -- if I almost completely fill the relaxation need with parks (28 park hexes). Then the total space requirements with farms, clinic and education go higher than a medium dome though.

 

[Xeorm]

So I tested these two domes over the course of about 5 sols:
View attachment 490864

All non-specialized non-child colonists, isolated domes etc. The left dome has 28 people and 10,8 and 10 max visitors for social, shopping, relaxation. You get consistent small complaints about full shopping and occasional small complaints about full other services. Buildings are all always mostly full, and if you check "visitors today" at midnight, the max number of visitors in a day seems to be 2*shifts*(max inside), which is consistently almost achieved for all buildings. For some bizzare reason, the "visitors today" for the large park seem to cap at 4*(max inside) instead of 6, but otherwise it behaves like the gym, i.e. consistently just about satisfies 28 people ( 3* max visitors in general).
If you add total numbers of visitors today for all buildings in the small dome with 28 people, it's very close to 28*4, which tells me that every colonist actually visits 4 random services each day. Since each service can satisfy 2*max visitors each shift, we get the 3* max visitors for each interest rule of thumb at max shifts.
In the right dome, despite only 1 large park instead of 2 for ~60 people, complaints about missing shopping are actually higher than those about missing relaxation, though this might be because I trained "workaholic" in schools, which means - relaxation interest. There is some randomness here though, relaxation complaints still tend to be highest on average.
Average comfort for left dome sank to 66, for the right to 70. Residence comfort is 65.

EDIT: For a medium dome with 84 people, I can get just above 66 comfort with 2 diners and 2 grocer -- if I almost completely fill the relaxation need with parks (28 park hexes). Then the total space requirements with farms, clinic and education go higher than a medium dome though.

Hmm. Interesting. Since you like testing, do you know if they change much if you use non-food service buildings more? I did get used to people frequently having uneaten meals. That might account for some of the difference?

 

[SillySMS]

death age is (MinAge_Senior is 61):
self:SetBase("death_age", self.MinAge_Senior + 5 + self:Random(10) + self:Random(5) + self:Random(5))
66-86

Can you check how Stem Reconstruction affects that? Based on a scan of my oldest colonists, it seems to set the age at which colonists become seniors to 80; I've seen middle-aged colonists at 79 sols on Mars, and one truly ancient 100-sol-old senior (it's possible there are older seniors possible).

Average is still 80, so that's nice. Thanks.

If I'm understanding that formula correctly (66+1D10+2D5), the mean should actually be 77.5, with a bell curve stretching from 69 to 86.

If I understand things correctly this is the demographics of a steady-state colony before Stem Reconstruction:
Children (0-5): 7.7%
Youth (6-9): 5.2%
Adults (11-30): 25.8%
Middle Aged (31-60): 38.7%
Seniors (61-86): 22.6%

Overall, 69.7% of your population is of working age at steady state, and you need 7.7 children/100 colonists to maintain population.

After Stem Reconstruction, assuming it sets the minimum age of seniors to 80 sols and otherwise touches nothing:
Children (0-5): 6.2%
Youth (6-9): 4.2%
Adults (11-30): 20.7%
Middle Aged (31-79): 50.8%
Seniors (80-105): 18.1%

At this point, 75.7% of your population is of working age, and you need 6.2 children/100 colonists to maintain population.

Also, assuming the formula for birth/comfort on the wiki is accurate (which I do not take for granted), you can now get away with just 58 comfort for steady population, because your middle-aged colonists have that much longer to have children.

In any event, a few issues with your analysis:
1) Not that it matters much, but you should have about 55.7 working age individuals per 80 colonists.
2) I'm 99% sure that the default average time to graduate from a university is ~10 sols; I notice my first class of graduates tends to pop the first time the university requests maintenance. With Martianborn Ingenuity, that time goes down quite significantly for Martianborn colonists.
3) My typical per 100 is 2 diners, 2 grocers, 1 spacebar, 1-2 infirmaries (depending on whether I expect heavy sanity issues), and 6+ decoration tiles. I suspect your numbers on workers tied up in service buildings are pessimistic. That said, I frequently get "building full" errors, I just don't care, and I rarely see the fraction of children drop below 10%.

EDIT: Updated numbers on #3 to reflect reality instead of exhaustion-induced mistakes.

 

[ChoGGi_2]

Those self:Random(5) go from 0-5 not 1-5.

Edit:
StemReconstruction adds +20 to .MinAge_Senior and .death_age

 

[YertyL]

(...)

In any event, a few issues with your analysis:
1) Not that it matters much, but you should have about 55.7 working age individuals per 80 colonists.
2) I'm 99% sure that the default average time to graduate from a university is ~10 sols; I notice my first class of graduates tends to pop the first time the university requests maintenance. With Martianborn Ingenuity, that time goes down quite significantly for Martianborn colonists.
3) I have generally found that about 1 clinic, 1 diner, 1 grocer, 6 decoration tiles, and 0.5 spacebars covers 100 colonists reasonably well. I suspect your numbers on workers tied up in service buildings are pessimistic. That said, I frequently get "building full" errors, I just don't care, and I rarely see the fraction of children drop below 10%.

1) yeah, I rounded, this is obviously only a very rough estimate and there are huge variations depending on the specific dome layout, techs etc. pp.
2) My source was https://steamcommunity.com/app/464920/discussions/0/3211505894130913894/ which seemed solid. High performance will lower time taken, so that might be the difference in your practice?
3) I just tested this with 100 colonists and a full spacebar in creative mode. Average comfort hovers between 47 and 54, which is roughly what you would expect, since you cover about 1/4 of shopping needs (24/100) and about half of social (60/100) and relaxation (48/100), but with mostly high comfort buildings. Birth rate is still surprisingly high with ~15-30 couples above 55 comfort and an estimated 0.6-0.7 children per sol (7 in about 10 sols), though that is without any seniors being present yet, and already far below the ~1.25 children per sol needed for stability. I am pretty sure you would get pretty massive problems with births, earth sickness and renegades in the long run with so little comfort.

One other thing this did teach me though is that spacebars are pretty awesome, because they cover both social and relaxation simultaneously, making them effectively a 20 capacity high comfort building in most situations (i.e. when colonists want both social and relaxation).

Those self:Random(5) go from 0-5 not 1-5.

Edit:
StemReconstruction adds +20 to .MinAge_Senior and .death_age

So your percentage of seniors drops from ~20% (15/76) to ~15% (15/96). Good to know

 

[SillySMS]

-snip-

In my defense, I was pretty exhausted when I wrote the initial post... and when I wrote edits, and I'm still exhausted. I'll try to make both this post and my original consistent.

My typical per 100 is 2 diners, 2 grocers, 1 spacebar, 1-2 infirmaries (depending on whether I expect heavy sanity issues), and 6+ decoration tiles.