The Logistics of Planetary Invasions: A Mechanistic Analysis

[Lord Solar]

In my experience the best ground armies I have created are units of Super-Heavy vehicles with triple CAP set to frontline attack with an officer that has 15%+ Ground Combat Offence bonus. These have so much more armor and HP that the infantry the NPR seem to use mostly is largely ineffective at killing them, while the CAP + attack + hit chance bonus + breakthroughs shred everything including the few vehicles the NPR uses. It's probably efficient to combine with some MAV, HCAP, or other anti-vehicle if you have it but mostly the CAP works just find. I've invaded a homeworld while being out-tonned several times over with this doctrine.
The only downside is you have to research super-heavy vehicles and super-heavy vehicle armor but that's pretty cheap and easily attainable by 50 years into a campaign.

[misanthropope]

meaning no unfriendliness, you didn't even _define_ the question "is ground combat broken?" let alone address it. 

remember the last time you were rewarded with an interesting tactical situation in ground combat?  no you don't, and neither does anybody else.  ground combat materially interferes with the space combat *and* empire management aspects of the game without providing an iota of gameplay.  it's not a game, it's a game simulator.  you spend copious time developing your pretty OOB and sigh "if only there were a game to play that used this". 

the conventional issue of "brokenness", referring to there being multiple valid playable strategies, doesnt pertain as so-called game balance has explicitly never been a design aim.  _do_ a substantial fraction of the units serve a valid function in the ground combat mechanism?  ha ha, no.

[Droll]

_do_ a substantial fraction of the units serve a valid function in the ground combat mechanism?  ha ha, no.

I think this point is best demonstrated by the fact that all effective armies will consist of infantry and anti-infantry tanks. With the occasional anti-tank unit thrown in for good measure. The only reason why you might have a tank with mixed weapons or assault infantry is RP.

In it's current state, ground combat is so abstracted that the only thing that really can be discussed is the maths, and not really any strategy beyond the most basic design advice. The abstraction helps aurora handle granular combat at grand scale but completely sacrifices any sort of decision making that the player may have have beyond "to invade, or not to invade?". Winning is literally just a matter of "send more".

Most if not all of the enjoyment I get out of the new ground combat system is really the RP aspect, it's why I model my armies down to the company level, despite the micromanagement headache that it causes.

[nuclearslurpee (OP)]

meaning no unfriendliness, you didn't even _define_ the question "is ground combat broken?" let alone address it. 

This is because I am not looking to answer that question, full stop. I am looking specifically at the question of planetary invasions, and common player complaints about them - they take too many troops, they take too long, they're too hard to supply, etc. In this respect I like to think I did at least a passable job of defining the question I wanted to investigate and then studying it reasonably well with some admittedly approximate analysis.

The much larger question about the goodness of ground combat mechanics as a whole is emphatically not something I intended, nor do I intend, to interrogate in this thread, and I would prefer that if folks want to continue the discussion in that direction they do it in a different thread which explicitly addresses those questions.

[Kristover]

I think you did a good job breaking down the mechanics and showing the math behind it all.  I also think the answer is 'ground combat broken' is kind of weird one because the level of abstraction which Steve chose to utilize for it doesn't result in a traditional gameplay loop in the same sense of space combat where there are many choices which can be made throughout the process.  For an RPer like me the ground system is completely suitable for what it is and there ARE choices to be made in force composition, timing of invasions, and strategic choices like 'should I bombard the hell out of this colony or should I try to capture it?'

Just a random musing, I think ground combat in an interstellar setting would be incredibly challenging to the point of absurdity. I'm former military and I know that it takes MILLIONS of tons of equipment, supplies, and fuel and the associated planes, trucks, and ships to keep even a modest 100K soldiers forward supplied and ready to fight at a distance on conventional Earth.  Now lets posit you're trying to invade an enemy homeworld of BILLIONS.  You would need millions of troops and an unfathomable amount of supplies, consumables, fuel, and parts constantly arriving.  When I play my campaigns, I'll invade and take outlying colonies and even important secondary worlds but generally I just bottle up home systems with orbital sensors, a large military base in system, and a credible ground force near by.  I MIGHT consider invading if it is a low militant/xenophobic race but if they're high militant/xenophobic, I'm not going anywhere near that meat grinder.

I would like eventually to see a bit more added to the ground combat side - I would like to see the A2A pods and the air to air aerospace game implemented and I think something like 'strategies' added where you can set something like 'elastic defense' or 'all-out attack' which would add modifiers and have counters.  I would also like to see some sort of concealment system which allows for increased special operations/guerilla play.  But all in all, I enjoy what we have.

[nuclearslurpee (OP)]

I think you did a good job breaking down the mechanics and showing the math behind it all.  I also think the answer is 'ground combat broken' is kind of weird one because the level of abstraction which Steve chose to utilize for it doesn't result in a traditional gameplay loop in the same sense of space combat where there are many choices which can be made throughout the process.  For an RPer like me the ground system is completely suitable for what it is and there ARE choices to be made in force composition, timing of invasions, and strategic choices like 'should I bombard the hell out of this colony or should I try to capture it?'

I personally agree. While there are some issues with the ground combat at the tactical level, which have been well-publicized elsewhere, I started this analysis intending to identify key problems and offer well-supported suggestions. Instead, even before I corrected some errors in my analysis I found myself overall very impressed with how well the strategic and operational balance works for homeworld invasions, IF the player adequately prepares for them with a long-term plan. Given how the ground combat is an all-new system in C#, continually being improved and with minimal playtesting since Steve is only one guy with limited time on his hands, it's really a good basis in the strategic sense and I'm sure Steve will continue to work out the tactical issues as time goes on.

[Agraelgrimm]

I think you did a good job breaking down the mechanics and showing the math behind it all.  I also think the answer is 'ground combat broken' is kind of weird one because the level of abstraction which Steve chose to utilize for it doesn't result in a traditional gameplay loop in the same sense of space combat where there are many choices which can be made throughout the process.  For an RPer like me the ground system is completely suitable for what it is and there ARE choices to be made in force composition, timing of invasions, and strategic choices like 'should I bombard the hell out of this colony or should I try to capture it?'

I personally agree. While there are some issues with the ground combat at the tactical level, which have been well-publicized elsewhere, I started this analysis intending to identify key problems and offer well-supported suggestions. Instead, even before I corrected some errors in my analysis I found myself overall very impressed with how well the strategic and operational balance works for homeworld invasions, IF the player adequately prepares for them with a long-term plan. Given how the ground combat is an all-new system in C#, continually being improved and with minimal playtesting since Steve is only one guy with limited time on his hands, it's really a good basis in the strategic sense and I'm sure Steve will continue to work out the tactical issues as time goes on.

What are you talking about limited testing? We are the testers lol.

[nuclearslurpee (OP)]

What are you talking about limited testing? We are the testers lol.

Before release, I mean. Although I'd say most of us are less "testers" and more "opinion havers". 

[Migi]

Your post is very intersting, looking at a planetary assault from a decade-long perspective is probably not what most Aurora players, even longtime players do.
I think it would be somewhat helpful if there was some sort of concept of planetary unit carrying capacity, to give the player a sense of scale (although beyond the scope of this thread).

[Veneke]

This is easily one of the best posts made on the mechanics behind aurora in quite some time.
 
One of the things that I think posts like this do is make it clearer what the player needs to do to accomplish a particular goal; and your analysis gives us a good ballpark range for what we need to do to ensure that we have an industrial base for the ground combat elements of a planetary invasion.
 
To build a planetary invasion force of 10m tonnes, with 2m tonnes of supplies, over the course of 20 years, takes 50 GFTF. 40 for the invasion force, and 10 for supply production. In terms of upgrades due to technology improvements over the course of 20 years, I'd probably just accept the lower tech troops or else extend the production time out and use all 50 GFTF to upgrade the worst of the obsolete elements. Honestly, 50 GFTF isn't unreasonable, but it certainly would never have occurred to me to build that many. I don't think I've had a campaign with more than 15, maybe 20. Which is a pretty good explanation as to why I always find it easier to nuke enemy planets from orbit!
 
The other extremely fascinating thing is that planetary invasions then resolve themselves so quickly. Your initial post which had ground combat last between a month and a year iirc was pretty reasonable. It seems shockingly fast that a planetary invasion would be over in a month or less. Appreciate that the trickiest of terrains could take 2 months, but that's evened out by others ending in days.

[nuclearslurpee (OP)]

The other extremely fascinating thing is that planetary invasions then resolve themselves so quickly. Your initial post which had ground combat last between a month and a year iirc was pretty reasonable. It seems shockingly fast that a planetary invasion would be over in a month or less. Appreciate that the trickiest of terrains could take 2 months, but that's evened out by others ending in days.

This depends quite closely on the relative size of forces involved - in the extreme case of invading a jungle mountains planet, the minimum force required is about 6,600,000 tons against 2m tons of defenders, but that can take you six months. Given that supplying an army for six months is a bit challenging, it can easily take longer as you start running out of supplies and transitioning to low-intensity warfare.

The time taken would also be sensitive to some of these parameters I've only estimated such as the shots-per-ton and tons killed per hit. Some empirical testing with different army compositions and detailed data collection would be really helpful to refine those parameters, but it may be the case that depending on army composition(s) an invasion could take more or less time as well.

[IanD]

Despair not!

After the Theory here is a practical example.

Brief history: Found two alien nprs within a month of each other about 20 years into game. Unfortunately I had put a sensor station in one of the home systems and they took exception to it. The other tried to claim jump a cost 0.23 colony with a ruined city. I took exception.
The first npr was only 3 jumps from Sol so they were dealt with first. This took some time (years), the 48 tank battalions of 75 heavy vehicles each evaporated in three weeks, lasting just long enough for me to get my first super heavy battalions in.
Fast forward 30 years. Still haven’t completely eliminated the first npr, but have finally found what I think is the home system of the claim jumpers. During this time a third npr had appeared, possibly activated by the claim jumpers. While I was preparing to invade the home world of the second npr, npr number three took exception to one of my surveillance stations and blew it away. Fortunately 17 29,000 ton cruisers have kept them contained for now.

The Invasion
I started the invasion of the npr number 2 homeworld. It had an initial visible ground force of 495,000 plus. However, in taking out the STOs by orbital bombardment the ground force dropped to 229,000 and continued to drop as I landed my forces.
 I made my initial assault with three army corps of seven brigades each consisting of four battalions of super heavy and ultra heavy vehicles, about 945,000 tons in all. Each battalion of approximately 9,800 tons had 10 logistics vehicles and each higher level had another 50.
All the combat battalions were set to front line attack. In the first 10x8 hour increments I lost 4 first generation super heavy tanks, one third generation ultra heavy tank, 2 construction vehicles and 30 logistics vehicles.

In the next 5 increments 9 available logistic battalions were dropped, and I lost 3 more first generation super heavy tanks, one static artillery piece and 8 logistics vehicles. I also got my first breakthrough. I then dropped 7 brigades of heavy infantry. This gave a total force of just over 1.2 million tons. The infantry were set to front line defence. Each brigade had four battalions each of 850 heavy infantry, 200 heavy machine guns plus command, logistics and construction elements.

Over the next 17x8 hour increments I lost 294 heavy infantry, 387 heavy machine guns 27 logistics infantry, 5 heavy infantry battalion HQ units and 3 logistics vehicles. The infantry achieved their first breakthrough during this period. I managed my second supply drop, 42 Logistics battalions (each of 160 vehicles) and 36 infantry logistics battalions (each of 200 logistics infantry).
The next 40 turns saw a reduction in casualties. 213 heavy infantry, 165 heavy machine guns, 60 logistics infantry, 1 heavy infantry battalion HQ units and 18 logistics vehicles. This was effectively the end of serious resistance.

In increments 73 to 194 I lost 121 heavy infantry, 96 heavy machine guns and 15 logistics infantry. I managed a second logistics vehicle drop of 21 battalions on increment 118 and an identical third drop on increment 168. The heavy infantry were set to front line attack on increment 157. On increment 182 there were a very large number of breakthroughs which continued until the end.

On increment 195 the planet surrendered. It had taken 1,552 hours or just 64²/₃ days.
Overall the principal losses incurred per increment were as follows: I lost 0.53 super/ultra heavy vehicles and 2.5 logistics vehicles each increment during the first 120 hours.
After the infantry landed I lost 17.3 heavy infantry, 11.1 heavy MG and 1.6 logistics infantry per increment for the next 136 hours. Over the next 320 hours losses dropped to 6.2 heavy infantry, 4.4 heavy MG and 1.5 logistics infantry per increment.

Over the final 976 hours losses dropped again to 1 heavy infantry, 0.8 heavy MG and just 0.12 logistics infantry.

The invasion could be divided into four phases with losses dropping in every phase. The dropping of logistics supplies kept the pressure on, without a reduction in enemy casualties.
The big mistake I made was the infantry logistics battalions. I dropped 36 battalions, while the most I actually required was 7, none of which would have been totally consumed. Finally I only just had enough logistics vehicles dropping 93 battalions in all, that is 14,880 logistics vehicles fortunately they are relatively quick to produce. In addition ground unit production was an empire priority research target.

[Nori]

Your armor must have out teched their weapons because you sure suffered light super heavy losses. Sure seems like big vehicles are excellent for assaulting planets.

[nuclearslurpee (OP)]

Your armor must have out teched their weapons because you sure suffered light super heavy losses. Sure seems like big vehicles are excellent for assaulting planets.

Heavy armor is usually effective for planetary assaults as the NPR armies are usually PW/CAP-heavy and I'm not sure they build very many weapons heavier than MAV which itself is unable to penetrate heavy armor.

However compared to infantry formations it does take several times more BP to produce the same tonnage of heavy armored units, for example for the same BP cost as 100,000 tons of SHV (with SHV armor) you could build 900,000 tons of unarmored INF. The latter might actually be more optimal for winning quickly due to sheer volume of fire if you have the transport capacity (note that Lanchester's Law varies with the square of formation size, but only linearly with combat efficiency), however the infantry will definitely suffer higher losses so over a sustained campaign of assaulting several heavily defended worlds the costs in attrition could shift in favor of the SHV.

[IanD]

Your armor must have out teched their weapons because you sure suffered light super heavy losses. Sure seems like big vehicles are excellent for assaulting planets.

The best armour on the NPR ships I captured was laminate composite armour. I had only recently achieved biphase carbide armour so given the time ultra heavy vehicles take to produce even having maxed out ground force construction rate most of my forces would be compressed carbon armour and earlier types of armour, hence its principally the first generation units being knocked out. Thus I was not that far ahead of the NPR in armour tech and could not afford to leave any units on the sidelines regardless of their armour!

I have only 32 troop transports with a 50,000 ton capacity and 16 with a 55,000 ton capacity, so I prefer ultra heavy vehicles. Transport capacity is on the tight side. My new standard ultra heavy being armed with 1 x SHAV and 3 x HCAP. My formation size is limited to 10,000 or under. It is helpful that the NPRs don't appear to build more that medium vehicles, but they do build a shed load of them (and infantry).
I am getting tired of the asteroid and cometary forts the NPRs build, usually protecting nothing with up to 200,000 tons of ground forces plus several thousand tons of STOs visible.