Aurora 4x
C# Aurora => Development Discussions => Topic started by: Steve Walmsley on October 14, 2018, 07:32:54 AM
-
Next issue to resolve is collateral damage from ground combat. In VB6, it is based on readiness losses, but that doesn't even exist in C#.
My current line of thinking is for combat to generate a total damage value, then that is applied against a population as an energy weapon attack. This would be based on attacks, not hits. So if you use heavy bombardment weapons, it should have an impact on the population regardless of whether you hit any hostile units.
The collateral damage probably should not be linear with damage. So personal weapons fired 6 times should not be equal to a shot from a heavy anti-vehicle weapon or a heavy bombardment weapon. My current thinking is to ignore AP value and use the damage value cubed as the baseline, then divide by a large number, maybe a million.
Using that example and assuming a base ground combat tech of 10 (about TL4),
- An infantryman with personal weapons would generate 0.001 collateral damage per combat round (damage 10, shots 1)
- Light anti-vehicle (damage 20, shots 1) is 0.008
- Light bombardment (damage 20, shots 3) is 0.024
- Medium anti-vehicle (damage 40, shots 1) is 0.064
- Medium bombardment (damage 40, shots 3) is 0.192
- Heavy anti-vehicle (damage 60, shots 1) is 0.216
- Heavy bombardment (damage 60, shots 3) is 0.648
Putting that in terms of regiments, 1000 infantrymen would generate 1 collateral damage per round while 50 heavy tanks (about the same size but 2x cost) would generate 11.2 collateral damage, assuming HAV and HCAP. To put that in perspective, vs energy weapon fire a construction factory has 20 HP and a research facility has 400 HP.
Another consideration could be how densely the planet is populated. For planets close to capacity, ground combat should cause a lot of damage. While for frontier colonies, damage should be relatively minor. However, even in the latter case it is unlikely the population is evenly distributed so there should be some minimum modifier. If this was a factor, the base collateral damage should probably be higher or it would be almost irrelevant except for very large populations.
Comments welcome.
-
I think having it scale exponentially with damage works well.
Trying to guess at the absolute damage that should be inflicted is really hard without testing, though. So much depends on emergent gameplay; for example, it's unclear to me whether an average fight will be between 1000 infantry or 100,000, since it will depend on just how much production spent on ground units works out being cost effective.
Actually, it also occurs to me that some sort of diminishing returns (well, not returns in this case, but diminishing) is probably the way to go. A fight between 100,000 infantry should cause much more collateral damage than 1,000, but not 100 times more. Similarly, it should be very hard to complicate erase a planetary population by bombardment, since some will be hiding in bomb shelters/caves/etc.
-
I think having it scale exponentially with damage works well.
Trying to guess at the absolute damage that should be inflicted is really hard without testing, though. So much depends on emergent gameplay; for example, it's unclear to me whether an average fight will be between 1000 infantry or 100,000, since it will depend on just how much production spent on ground units works out being cost effective.
Actually, it also occurs to me that some sort of diminishing returns (well, not returns in this case, but diminishing) is probably the way to go. A fight between 100,000 infantry should cause much more collateral damage than 1,000, but not 100 times more. Similarly, it should be very hard to complicate erase a planetary population by bombardment, since some will be hiding in bomb shelters/caves/etc.
Yes, this will definitely need testing. I am running some experiments at the moment using formations I created for test purposes to get an idea of scale. However, as you say, it needs a full campaign to see how this will work in reality. I am getting fairly close now to running my first test campaign. Probably 5 player races plus precursors and I will start adding NPRs once everything seems to be running OK. I need to finish a few more areas, but I now have a growing to-do list for 'after campaign starts' :)
-
what I would keep in mind is not only how populated a body is but how many installations are at the planet/body too..
the chance to destroy the lone 10 factories on a earth like planet with 50 or 200million people should me very very low...
no idea how to code this / get it into a math formula -.-
...
but my problem with collateral damage at all is how realistic it is that fighting occurs manly in cities/industrial strongholds instead of open area/forest/mountains...
I like the idea of using the "pop/maxpop" factor, maybe there could also be a factor about something like "#workers" that are occupied by working in installations... so with a high employment rate the chance to hit an installation is high (as there are more of them relation to population) - with nearly no one working in an installation the chance is nearly Zero even if there are millions of people...
not sure if that is understandable.. sorry :-[
---
also, maybe an additional factor which represents the terrain of the planet/body would make sense... a body with deadly inviroment there would the fighting mostly in the populated regions.. in a desert world who knows... in mountains there would be more fighting in the mountains than in the cities etcpp...
-
Another consideration could be how densely the planet is populated. For planets close to capacity, ground combat should cause a lot of damage. While for frontier colonies, damage should be relatively minor. However, even in the latter case it is unlikely the population is evenly distributed so there should be some minimum modifier. If this was a factor, the base collateral damage should probably be higher or it would be almost irrelevant except for very large populations.
Comments welcome.
Any thoughts about scaling the actual combat intensity based on size of planets as well?
For example at a certain point when fighting with massive amount of forces on tiny bodies adding more forces will achieve little except getting in each-others way and making sure each enemy artillery or bombardment weapons wipes out alot more of your own forces.
Or when having quite small forces fighting on a huge planet ( at least one with alot of population and installations ) leads to less contact between the forces and lower intensity even if both sides seek a fight.
-
*snip*
Comments welcome.
*snip*
The thing about combat is that it's very much constrained by logistics. What can I say, armies are ravenous when it comes to supplies, and you do need to be able to supply that. And that means you either fly it in, you rail it in, you sail it in or you drive it in. This inevitably means that no matter what, even fairly newly settled worlds are likely to see a lot of combat in and around the places they live. Even fully automated mining colonies are likely to see most combat happen around major mining complexes and any surface/orbit infrastructure. It's where the goods are, it's where the money and the people are, and it's where you can move supplies around the easiest.
That doesn't mean that there won't be combat outside those inhabited regions, many battles for sections of the planet will start with maneuvers to cut off or establish corridors of supply, and remote sections of roads, railways or bridges are really good choices for that because long supply lines are really hard to defend continuously. There's a reason WW1 and 2 saw extensive convoying, and it's this exact problem.
And the issue of logistical path constraints gets even worse the higher the colony cost, but interestingly enough there's reason here to have a cut off below which the risk collateral damage (well, to non-infrastructure installations) drops considerably. A heavily populated planet with a colony cost requirement is likely to have an expansive network of transit systems that move goods and people between major population areas, which may be covered, but below that population count there simply has neither been time nor the desire nor the economical development to fund such a convenience, and all travel will be in fully environment rated independent vehicles. Which means that every battle will effectively be an assault on a fortified city simply because of the way infrastructure constraints work in real life.
Sure, you are less likely to damage the facilities manned by the people in a sprawling environmental hazard protection complex (or a bunker in other words), but you are extremely likely to compromise their infrastructure and thus their lives, likely causing increased casualties through losses of infrastructure alone, while combat in the city itself will inevitably involve the civilian populace who have no option of escape.
It also makes sense in this case for defenders to prefer a defense further afield, because otherwise the risk to habitation zones becomes much greater.
-
So far we've been talking about damage to installations -- various types of factories and the like -- but Hazard brings up a good point. Should collateral damage apply to Infrastructure as well? If Col Cost 2.0 (or sometimes 3.0) worlds most frequently represent domed cities and pressurized bubbles, shouldn't combat damage those?
We'd need to be careful that Infrastructure doesn't become 'protective padding' on low Col Cost worlds (my homeworlds are always producing the stuff faster than it can be shipped) -- maybe use Agricultural worker percentage or Col Cost to determine losses?
Of the top of my head, maybe something like each point of Collateral Damage destroys (Colony Cost)% of Infrastructure? (Iterate each point, so 110 collateral damage leaves 33% Infrastructure, not zero.)
-
Habitation domes, clean air and water infrastructure, for Ocean worlds under water and over water cities. Infrastructure can mean a lot of things.
Keep in mind though that even a single 1000 soldier strong regiment can deliver 1 point of collateral damage per round (6 hours) at Techlevel 4, while heavier forces cause ever more damage per round for the same size. Heavy Bombardment is the worse in this regard at about 0.01 damage per size of the weapon. And as techlevels increase this only gets worse and worse, while orbital assaults are unlikely to be small affairs outside of minor settlements.
Being modeled as Energy Weapon attacks on facilities, collateral damage will also toss up dust into the atmosphere, so if a battle gets big enough and lasts long enough you are likely to see colony cost escalate. You may be able to build a nice buffer for some grace, but that will mean deliberately shipping infrastructure around so it requires either attention or wealth. And finally, unless you are dedicating construction factories to Infrastructure production you can't produce Infrastructure fast enough, even high trade good production planets can't keep up with that.
I've no issue with Infrastructure becoming a bit of padding. Get a battle going long enough and it's not going to help. Although civilian casualties would also be nice to see modeled, when it comes to ground combat.
-
I've coded collateral damage based on my original idea but in such a way that the formula for damage amount can be easily adjusted. At the moment, there is no capacity element to the formula.
Once the total damage to a population is calculated, it is allocated as a series of 2-point energy weapon attacks. This is because infrastructure has 2 hit points. A construction factory (20 HP) would have a 10% chance of being destroyed, etc.. I think it is realistic that a domed city or something similar is likely to take a lot of damage if heavy weapons are involved.
In addition to the installation damage, the collateral (energy) damage increases the dust level by 5% of the damage amount and inflict civilian casualties at the rate of 2,000 per point of damage.
-
Would it be possible to instead create a randomised list of all facilities on planet and assign damage from the top in descending order and according to HP every round of combat?
You'd refresh the list every construction pulse because you need to account for new construction anyway. I understand this can cause odd results like a 400 HP research facility eating all collateral damage during a construction pulse and just not breaking at lower tech levels, thereby not suffering any collateral facility damage during that construction pulse. On the other hand, the list refreshes every 5 days or so it's not likely to stay on the top of the list.
-
Would it be possible to instead create a randomised list of all facilities on planet and assign damage from the top in descending order and according to HP every round of combat?
You'd refresh the list every construction pulse because you need to account for new construction anyway. I understand this can cause odd results like a 400 HP research facility eating all collateral damage during a construction pulse and just not breaking at lower tech levels, thereby not suffering any collateral facility damage during that construction pulse. On the other hand, the list refreshes every 5 days or so it's not likely to stay on the top of the list.
Currently, damage is randomly allocated for every shot, based on the weight of installation size. So a research facility is 20x more likely to be hit than a construction factory (but also 20x less likely to be damaged by the hit).
-
Looks like a good start point. Now just really looking forwards to hearing about that first test campaign!
-
Would this make infrastructure a good way of "padding" a colony against collateral damage? I think it would be interesting to have an installation that protects against collateral damage by appearing like an installation much larger than it is. Something infrastructure sized that appears as research facility sized when it comes time to applying collateral damage, as a way to abstract fortifications or hardening of civilian infrastructure.
-
I've always thought that collateral damage was too high in VB6 Aurora and would happily see it generally become slightly less in C# Aurora. One round of ground combat on Earth (5-days) can easily produce a million dead civilians if the armies are large enough, even without using weapons of mass destruction. That is just crazy. No human conflict has ever produced civilian casualties at such rates. Nuking big cities is of course an entirely different matter. I dunno about the math, but any change that reduces collateral damage from what it used to be gets my vote.
Note that I'm not advocating getting rid of collateral damage, or reducing it to something meaningless.
Also, Infrastructure as padding against collateral damage sounds interesting.
-
Keep in mind that the Second World War, one of the most vigorous wars ever fought with one of the highest per day casualty rates, killed on average 10 000 soldiers per day and twice that many civilians. Certainly, a part of that was deliberate action to commit genocide, but another sizable chunk was battles being fought in and around urban areas as well as the results of deprivation of food and other resources necessary to maintain the war effort.
And while modern day weapons are considerably more accurate than WW2 weapons, they are in many cases also more powerful, and thus more likely to cause collateral damage.
-
I've always thought that collateral damage was too high in VB6 Aurora and would happily see it generally become slightly less in C# Aurora. One round of ground combat on Earth (5-days) can easily produce a million dead civilians if the armies are large enough, even without using weapons of mass destruction. That is just crazy. No human conflict has ever produced civilian casualties at such rates. Nuking big cities is of course an entirely different matter. I dunno about the math, but any change that reduces collateral damage from what it used to be gets my vote.
Note that I'm not advocating getting rid of collateral damage, or reducing it to something meaningless.
Also, Infrastructure as padding against collateral damage sounds interesting.
As the new rules currently stand, it would take 500 points of collateral damage to kill 1m civilians. That is one round of 500,000 infantry or 2250 heavy tanks (100 tons). While that is still higher than historical, it is lower than VB6 and populations in Aurora tend to be larger than those of historical combatants and therefore more likely to be densely populated. Also, for game play purposes the collateral damage losses need to be meaningful. BTW I'm not saying the current level is definitely right - that will depend on playtesting.
-
Collateral damage from ground combat can be meaningful even without heavy civilian casualties.
Civilians are useless without facilities after all, even if facilities are useless without civilians.
Because of this, it's completely alright if the threat of loss of workforce is much less direct and much more the result of the loss of habitable environment conditions. Populations tend to bounce back much faster in real life compared to the damage done to the economy and environment in wars after all.
-
As the new rules currently stand, it would take 500 points of collateral damage to kill 1m civilians. That is one round of 500,000 infantry or 2250 heavy tanks (100 tons). While that is still higher than historical, it is lower than VB6 and populations in Aurora tend to be larger than those of historical combatants and therefore more likely to be densely populated. Also, for game play purposes the collateral damage losses need to be meaningful. BTW I'm not saying the current level is definitely right - that will depend on playtesting.
I can imagine that Collateral damage is tricky to balance, because it needs to be balanced in both ways. Too much collateral damage and the entire point of offensive ground forces goes away, which is the ability to take over planets mostly intact instead of nuking from orbit.
I totally agree with your approach that a successful balance should promote and reward using specialist and highly trained + well led lighter units, supported by precision weapons ( which would minimize collateral damage ) over using brute force of the heaviest bombardment guns and mechs/tanks in the arsenal. Basically the German concept of Bewegungskrieg from WW2 using lighter mobile small tanks, fast moving formations and precision CAS to quickly breakthrough and go around most of the enemy forces so that they could force surrender without much fighting.
The heavy weapons need to have their own role still when it comes to breaking enemy fortress worlds, or escalating a drawn out war/stalemate where both sides are well dug in, but such heavy fighting should maybe lead to a 50% destroyed world on average IMO ( Still better than nuking, but giving you a bit of a feeling of was it worth it afterwards ).
I've suggested before using some sort of diminishing returns for planetary damage and population casualties, but that would require keeping track of what the maximum population / facilities was before combat started ( so that it becomes harder and harder to knock out the last bits of it ).
-
Bewegungskrieg is still practiced in modern warfare doctrines. Maneuver and denying the enemy the same are critical components of any conflict currently pursued, and is likely to remain so unless and until static defenses start being more potent again than fast moving vehicles with big guns/bombs.
-
I've suggested before using some sort of diminishing returns for planetary damage and population casualties, but that would require keeping track of what the maximum population / facilities was before combat started ( so that it becomes harder and harder to knock out the last bits of it ).
Maybe I add a new 'destroyed installation' type that can be the result of future collateral damage (making the rubble bounce). One for factory-size, one for research facility size, etc. So when an installation is destroyed, it is replaced by a 'destroyed' equivalent that can potentially absorb future collateral damage hits.
Or maybe just have a flag for destroyed, so you can would have separate lines for intact and destroyed construction factories, intact and destroyed research facilities, etc.
In the latter case, the destroyed installations could be salvaged for a portion of their materials.
-
I've always thought that collateral damage was too high in VB6 Aurora and would happily see it generally become slightly less in C# Aurora. One round of ground combat on Earth (5-days) can easily produce a million dead civilians if the armies are large enough, even without using weapons of mass destruction. That is just crazy. No human conflict has ever produced civilian casualties at such rates. Nuking big cities is of course an entirely different matter. I dunno about the math, but any change that reduces collateral damage from what it used to be gets my vote.
Note that I'm not advocating getting rid of collateral damage, or reducing it to something meaningless.
Also, Infrastructure as padding against collateral damage sounds interesting.
As the new rules currently stand, it would take 500 points of collateral damage to kill 1m civilians. That is one round of 500,000 infantry or 2250 heavy tanks (100 tons). While that is still higher than historical, it is lower than VB6 and populations in Aurora tend to be larger than those of historical combatants and therefore more likely to be densely populated. Also, for game play purposes the collateral damage losses need to be meaningful. BTW I'm not saying the current level is definitely right - that will depend on playtesting.
Maybe you could make the collateral-damage effects on population increase as the population of a planet increases? With modifiers for worlds with a colony-cost, where Infrastructure needs will concentrate populations.
So 500 collateral to a 0-cost colony with 1 million civilians might kill a hundred thousand, but the same 500 collateral to a 3-cost colony with 100m citizens kept alive by Infrastructure (A la a Warhammer Hive City) might kill 10 million or more!
-
Maybe I add a new 'destroyed installation' type that can be the result of future collateral damage (making the rubble bounce). One for factory-size, one for research facility size, etc. So when an installation is destroyed, it is replaced by a 'destroyed' equivalent that can potentially absorb future collateral damage hits.
Or maybe just have a flag for destroyed, so you can would have separate lines for intact and destroyed construction factories, intact and destroyed research facilities, etc.
In the latter case, the destroyed installations could be salvaged for a portion of their materials.
Yeah, that sounds like a brilliant way to do it.
It's also pretty realistic because I can totally imagine ground forces fighting among ruins and rubble where no further damage can be inflicted on installations and where all civilians have since long fled the field ( if hit-roll targets destroyed installations then don't apply any direct civilian casualties either ).
-
Maybe I add a new 'destroyed installation' type that can be the result of future collateral damage (making the rubble bounce). One for factory-size, one for research facility size, etc. So when an installation is destroyed, it is replaced by a 'destroyed' equivalent that can potentially absorb future collateral damage hits.
Or maybe just have a flag for destroyed, so you can would have separate lines for intact and destroyed construction factories, intact and destroyed research facilities, etc.
In the latter case, the destroyed installations could be salvaged for a portion of their materials.
Yeah, that sounds like a brilliant way to do it.
It's also pretty realistic because I can totally imagine ground forces fighting among ruins and rubble where no further damage can be inflicted on installations and where all civilians have since long fled the field ( if hit-roll targets destroyed installations then don't apply any direct civilian casualties either ).
Yes, the civilians aspect is a good idea.
-
Yes, the civilians aspect is a good idea.
If you want to make it really detailed you could even scale civilian casualties based on employment factor, or connect the civilian casualties to how much population is needed to run the corresponding installations.
( Hit an installation with alot of workers needed nearby, and they are likely in the field of fire too ).
-
Yes, the civilians aspect is a good idea.
With that in mind, are we going to see changes in what installations need civilian personnel?
I mean, it always weirds me out a bit when the Military Academy, Genetic Modification Facility, Sector Command, Space Port or Ground Force Training Facility take no personnel whatsoever to run. Deep Space Tracking Station? Sure, that's probably automated. Mass Drivers and Automated Mines? Well, obviously. But everything else would seem to me to be the sort of thing that requires large numbers of supporting personnel.
-
I guess you could also include a new installation similar to infrastructure being civilian shelters. They would have a capacity of civilians, would reduce civilian casualties and in turn greatly reduce the negative impacts on lost productivity, impacts of fallout and chance for civilians to surrender as a result of bombardment.
-
Yes, the civilians aspect is a good idea.
With that in mind, are we going to see changes in what installations need civilian personnel?
I mean, it always weirds me out a bit when the Military Academy, Genetic Modification Facility, Sector Command, Space Port or Ground Force Training Facility take no personnel whatsoever to run. Deep Space Tracking Station? Sure, that's probably automated. Mass Drivers and Automated Mines? Well, obviously. But everything else would seem to me to be the sort of thing that requires large numbers of supporting personnel.
Yes, those are good point. I will review that.
-
Another vote for more installations requiring workers! I'd go so far as to require DSTS require a crew of 500 or something but I can understand that being too fiddly and micro for most people.
-
That's pretty much how automines function currently. They're not 100% automated, but rather the crew is so small (compared to 50,000 people) it's not worth tracking.
-
I guess you could also include a new installation similar to infrastructure being civilian shelters. They would have a capacity of civilians, would reduce civilian casualties and in turn greatly reduce the negative impacts on lost productivity, impacts of fallout and chance for civilians to surrender as a result of bombardment.
Of course, if one of these WAS destroyed, it would lead to phenomenal casualties, even if its not the first target
-
That's pretty much how automines function currently. They're not 100% automated, but rather the crew is so small (compared to 50,000 people) it's not worth tracking.
Well yes, that's how I RP my auto-mines and listening stations - that there's a small crew maintaining and servicing them, but mechanically they are self-contained, automated packages that you just drop down wherever.
-
Keep in mind that the Second World War, one of the most vigorous wars ever fought with one of the highest per day casualty rates, killed on average 10 000 soldiers per day and twice that many civilians. Certainly, a part of that was deliberate action to commit genocide, but another sizable chunk was battles being fought in and around urban areas as well as the results of deprivation of food and other resources necessary to maintain the war effort.
Well, that is a bit misleading because such a vast amount of civilian casualties in WW2 came from the Holocaust and other similar war crimes. But if you look at actual urban battles, we have a great case with Arnhem, where the British paratroopers fought the Germans - so it was a surprise operation that did not give time for civilians to flee - and the town was subject of artillery and aerial bombing, as well as fierce street fighting over a period of 9 days. Dutch authorities recorded only 452 casualties, whereas combined British and German casualties were about 3.5 thousand dead. Arnhem had a population of about 90 thousand back in the WW2 period. So that's a very small percentage. In Stalingrad, less than two thousand civilians died, though numbers are really sketchy as Soviets drafted significant numbers into militia and factories, and many others had time to flee the city. In contrast, the 1939 siege of Warsaw caused 18 thousand civilian casualties in 20 days, most of whom perished due to aerial bombing, out of a total population of about 1.3 million. Still a fairly low percentage.
The overwhelming majority of civilian casualties came from the unrestricted strategic bombing campaigns, from the few extended sieges (like Leningrad), from war crimes and anti-partisan operations, and finally from disease & malnutrition & starvation. Now, these are certainly elements closely related to warfare, but they are not the outcomes of actual battles. They are, somewhat crudely, represented in Aurora by changes to the population growth via radiation and atmosphere changes; ie nuclear winter does not depopulate the Earth immediately, just slowly kills people off. As there is no current system that would model agriculture or maintenance of civilian populations, I don't mind that collateral damage to the civilian population being a bit too high, but the VB6 levels were far too high, in my opinion. Since Steve already confirmed that for C# the amounts of collateral damage will be reduced, I'm happy. I just wanted to clear up this very popular misconception that fighting itself, even in urban areas, produces large amounts of civilian casualties, which it does not.
-
Yeah, as far as I know civilians tend to be able to flee cities just fine before the battle gets there. Remaining alive after that fact tends to be a lot more challenging to them, but yeah.
-
I haven't read through this whole thread yet, but consider the potential for infrastructure to reduce ground combat or bombardment collateral damage.
You could 'shore up' a world thats being heavily fought over by removing unneeded installations wile simultaneously dumping on infrastructure, maybe only underground infrastructure? Consider it burying your buildings or something.
-
This could be done by simply giving underground infrastructure the highest amount of HTK for its size/cost. Say, double or triple that of normal infrastructure.
-
IIRC underground infrastructure no longer exists. It's now low-gravity infrastructure.
Also, right now all facilities have a chance to be destroyed equal to 2/size. Cost varies, with larger facilities costing more compared to their size.
-
And rubble is a thing that now exists, and it protects surviving installations from damage. So the more rubble, the less chances for stuff to get wrecked.
-
And therefore, yes, adding a whole pile of (extra) Infrastructure to a colony will reduce the chance of losing vital Installations. And since 25 Infrastructure takes up the same space/HTK as one Construction Factory or AutoMine -- but consumes less minerals -- it makes for better (or at least, cheaper) 'armour' for your industry.
-
So now it's not only an RP thing to do for humanitarian reasons, but a valid strategy, to pile X amount of infrastructure on a planet... just in case. Like the countries on Earth that still build bomb shelters.
-
And therefore, yes, adding a whole pile of (extra) Infrastructure to a colony will reduce the chance of losing vital Installations. And since 25 Infrastructure takes up the same space/HTK as one Construction Factory or AutoMine -- but consumes less minerals -- it makes for better (or at least, cheaper) 'armour' for your industry.
Since the chance for "rubble" to absorb hits is based on the ratio of destroyed installations to intact installations, excess infrastructure will actually protect you less than you'd expect.
Consider a Planet A with 50 factories vs Planet B with 50 factories and 500 infrastructure. After losing 25 size* worth of installations, Planet A has lost half its installations, and thus future collateral is reduced by 50%. After losing 25 size worth of installations, Planet B has lost 25% of its installations and thus future collateral is only reduced by 25%
Note that if you do the math the infrastructure will still reduce the number of factories you lose (though oddly enough it will actually increase the rate of population loss). It just wont reduce it by half like it might appear at first.
At first I was surprised and wondered if the infrastructure actually increasing population loss was a detrimental bug. However after thinking about it I don't think it's a big issue. It can also be thought of as an aspect of population density; on a planet with 5 million inhabitants and a bunch of factories, those people are probably mostly in urban centers and vulnerable to collateral damage, whereas if you have 5 million people and only a few mines they're probably mostly spread out across the planet.
*I think factories are 1 size and infrastructure is .1? But regardless of the actual numbers the underlying point still applies.
-
*I think factories are 1 size and infrastructure is .1? But regardless of the actual numbers the underlying point still applies.
It would be more accurate to say Infrastructure is size 2 and factories are size 50.