• We have updated our Community Code of Conduct. Please read through the new rules for the forum that are an integral part of Paradox Interactive’s User Agreement.

Dev Diary #99 - Ground Combat & Army Rework

Hello everyone and welcome to another Stellaris dev diary. Today's dev diary is about some changes coming to ground combat and armies in the 2.0 'Cherryh' update. This will be the last dev diary before we take a break for the holidays, so there will be no diaries in the next week or the week after that. Stellaris dev diaries return on Thursday January 11th, 2018.

Defense Armies and Fortresses
Constructing Defense Armies have always been largely a meaningless exercise in Stellaris. While they are useful for reducing Unrest and occasionally might be able to beat off an unprepared attacker, the fact that a planet is capped on how many armies can be defending it while the attacker is *not* capped on how many armies are attacking, coupled with the general weakness of defense armies, means that defending a planet against a ground invasion is generally an exercise in futility and will at most delay an attacker by a few weeks. However, if we solved this by just making defense armies a lot stronger or capping the number of attacking units, the result would turn every invasion of a backwater colony into a big affair - something that is not particularly desirable when a war can involve several different actors with hundreds of planets between them.

For this reason, we have decided to rework Defense Armies into something that is actually useful, but requires a significant investment of resources to muster more than a token defense. Instead of being directly buildable by the empire, defense armies are created from certain buildings. The capital building will produce defense armies depending on its level, as will some other planetary uniques like Military Academy. If you want a planet to be well defended, however, you will need to construct Fortress building on its tiles. Fortresses require a pop to work them, do not produce any other resources than a small amount of Unity, but provide a significant amount of defense armies to protect the planet. Armies spawned by Fortresses are also impervious to orbital bombardment, and will not be able to be killed without first ruining the building itself. The armies generated by a building have their species and type set by the pop working it, so a Very Strong Battle Thrall will produce several powerful defense armies if placed on a Fortress, and special pops like Droids will produce their own variants like Robotic Defense Armies rather than the normal ones. Fortified worlds will also be able to be fit with an FTL inhibitor (the exact way they get them is not yet determined) that prevents enemy fleets from leaving the system unless the world is captured, which allows for the creation of Fortress Worlds to protect strategically important systems.
2017_12_21_3.png

(Building icon is a placeholder)

One more important change related to Defense Armies is a change to Unrest: Armies on planets no longer reduce Unrest directly. Instead, to handle a planet with high Unrest, you will need to construct Fortress-style buildings or take other measures (such as using Edicts) to reduce the planetary Unrest. This means you cannot simply capture a planet and then spam a dozen defense armies to immediately zero out the Unrest. As part of this, we will be balancing certain events and effect to ensure newly captured worlds do not instantly defect back to their former owner.

Finally, as part of all these changes Defense Armies have received a general buff and there are several new technologies that unlock additional tiers of forts and various improvements to Defense Armies' combat ability, meaning that they will grow stronger alongside the invention of new, more powerful assault armies.

Assault Army Management
A major aim of our changes to armies is to reduce the amount of unnecessary micromanagement of armies. For this reason, and to make Assault Armies' role more explicit, we have decided to change Assault Armies to always be based in space. Whenever not directly engaged in an invasion, Assault Armies will now always automatically embark onto their transports, ready to be used to invade another world. We also aim to fix the minor but immersion-breaking bug where transport fleets are giving endlessly increasing sequential names whenever they land and embark again.

Combat Width, Retreating and Collateral Damage
Another change to ground combat is the introduction of new mechanics in the form of Combat Width. Combat Width is determined by the size of the planet, and decides how many armies can be taking and receiving damage at the same time: For example, if 20 assault armies invade a world held by 10 defense armies with a combat width of 10, all 10 defense armies will be immediately engaged in battle while only half the assault armies will be able to deal and receive damage, with additional assault armies joining the fray as the armies in front of them are destroyed. This means that it is no longer possible to take a well defended world without losses by simply throwing a hundred clone armies at it: If you wish to minimize losses (and thus War Exhaustion), you will need to invest in expensive, high-maintenance elite armies.
2017_12_21_1.png

(Interface not final)

We've also added the concept of Collateral Damage: As armies fight on the planet, civilians and civilian infrastructure is caught in the fighting. Each time an army deals damage in battle, it will inflict a random amount of Collateral Damage, which increases Planetary Damage similar to Orbital Bombardment (see below) and can lead to the death of Pops and the destruction of buildings and tiles. Some armies will deal more Collateral Damage than others: For example, Xenomorph armies are highly destructive and cost-efficient, but will wreak immense havoc on the planet, potentially leaving it in ruins in the process of capturing it for your empire.

While working on combat mechanics we also took the time to change the way Morale Damage works, making it something that is suffered by both sides (instead of just the loser) and making the effects of it more gradual, so that armies suffer a drop in combat efficiency once they are <50% morale, and then another, sharper drop when they are broken (0% morale). This should make certain armies, such as Psi Armies, highly effective against low-morale opponents like Slave Armies, but less effective against an unfeeling army of Droids. Finally, we've also tweaked the damage-dealing algorithm so that damage is less evenly spread among combatants, making it so that even an outnumbered force can destroy regiments and inflict war exhaustion on the enemy.
2017_12_21_2.png


Finally, we have made some changes to retreats. When an attacker retreats from a ground combat, there is now a significant chance that each retreating regiment is destroyed while attempting to return to space, making retreat a risky endeavour and eliminating the tactic of simply send in the same army again and again in wave attacks, instead making retreats something you do in order to preserve at least some of your army in a poorly chosen engagement.

Orbital Bombardment Changes
Finally, again in the interest of reducing the micromanagement needed during war, we've changed the way orbital bombardment works. Fortifications have been entirely cut from planets, so that there is no need to bombard lightly defended worlds before going in with the ground troops. Instead, we have added a requirement that planets cannot be invaded if there is a hostile Starbase in the system, so that transports cannot snipe worlds that are protected by defensive installations present in the same system. Orbital Bombardment, instead of being something you have to manage and wait for in every single planetary engagement, is now something you do to soften up a particularly well defended target, or simply to wreak havoc on the enemy's planet and drive up their War Exhaustion.

As a planet is bombarded, the fleet will deal Planetary Damage, ruining buildings and killing Pops. Bombarding fleets will also do damage to armies present on the planet (unless those armies are protected by a Fortress), and over a long enough time can decimate a defending force, though doing so will likely cause heavy damage to the planet and may delay the attacker long enough that the owner of the planet has time to build up their forces or inflict enough war exhaustion to force a peace. The rate at which the planet is damaged can also be slowed with the construction of buildings such as Planetary Defense Shield, further dragging out the process.

As part of these changes, we've consolidated the Bombardment Stances into the following:
  • Selective: Deals normal damage to armies/buildings and light damage to pops. Cannot kill the last 10 pops.
  • Indiscriminate: Deals heavy damage to armies, buildings and pops. Cannot kill the last 5 pops.
  • Armageddon: Deals massive damage to armies, buildings and pops. Can turn planets into depopulated Tomb Worlds with enough bombardment. Only available to certain empires such as Purifiers.

Attachments
Finally, on the topic of attachments, we have decided to cut them entirely from the game. We discussed a variety of ways to improve the way you assign them, but ultimately decided that we already have so many types of armies and not nearly enough combat mechanics to justify a significant investment of UI time that could go towards something like the Fleet Manager instead. The technologies that previously unlocked attachments will be changed to give other effects, such as direct buffs to certain army types.

That's all for today! As I said, we're now going on hiatus, so I'll see you again on January 11th with a dev diary about... well, that's a secret, actually. You'll just have to wait and see!
 
Last edited:
  • 1
Reactions:
@Wiz

A consideration: if we're going to be able to create full fortress worlds, any chance for us to go all in on a choke point defense system with a new mid-game ascension perk (say...3rd on) that would let us upgrade a planet with X number of fortress and perhaps Y number of shield generators to have the full "Shielded" modifier that xenophobe FE's sometimes have?
 
Some people want to play optimally and don't want playing optimally to be a chore.
Attachments have never been "optimal". Adding Hunter-Killer Drones to an Assault army provides a 15% increase to damage, at the cost of 83% of simply recruiting a second Assault Army for a 100% bonus. Even with Gene Warriors the disparity shrinks only to 16.6% (assuming equalization of EC to Minerals), which is still more than the 15% buff, and that is before we consider the added hitpoints. Even if we add the impact of maintenance, an army that will pump all its resources into quantity rather than quality will always win. Wiz referred to this himself when he mentioned the impact of cheap clone troops in this very thread, even if he meant his in a more general sense.

Let's take a look at the stats as per the wiki:
  • Clone Army: 60 minerals -- average 281.5 DPS, 250 Health
  • Gene Warriors w/ Hunter-Killer Drones: 300 minerals, 50 energy -- average 517.5 DPS, 400 Health
For the price of a single unit of Gene Warriors with HK Attachment you get 5 Clone Armies with a total of 1407.5 DPS and 1250 Health. In addition you save 250 Energy. The Clones' maintenance is somewhat higher (1.65 vs 1.00), but (a) this can be compensated by drawing from the EC you saved at recruitment, and (b) you don't even need 5 Clone Armies to match 1 Gene Warrior unit; even 2 could do the job, for a maintenance of only 0.66 Energy per month.

In short, if you've been using Attachments all this time to play "optimally", you've been gimping yourself. They're a trap option, and although it's true that this might warrant criticism, a player would only have to compare the numbers to notice this.

You'd have a point once we get to 2.0, as then the addition of Combat Width would finally make individual army quality (and thus attachments) more important, but as it seems that you've already unintentionally worked with a non-optimal setup so far, I don't see how it should get worse if attachments were retained and you'd simply ignore them. And just to point it out again: the AI does not use attachments either, and they are easy to mod out of the game altogether if you really wanted to go that far. There is no need to take them away for everyone else.

Alas, they will quite likely be impossible to mod back in, as I imagine PDX will remove the necessary strings from the hardcoded vanilla files, preventing mods that seek to reintroduce this feature from doing so.

If they're effectively useless and have no relevancy to the game, they shouldn't be in the game.
By the same logic we should not have a feature to rename ships or characters. I disagree with the notion that it is wrong to have a feature that adds flavor only (though it would have changed with 2.0 anyways, see above).

But they do still exist, as flavour text and select stat buffs, just "baked in" ones instead of fiddly ones you need to waste time manually enabling.
The attachment system enabled you to customize various combinations of army, species and addon depending on the theme you were going for. Now the appropriate technologies will become active automatically for all your armies, with no respect to how much that might actually fit to your empire's headcanon.

It's bad enough that our Slave Armies all wear Power Armor -- this is how it works right now, because the equipment in question is handled as a technology rather than an attachment already. It's the perfect example of how it will apply to all attachments with Cherryh, and how awkward the end result may appear if you value flavor.

If you simply don't care about whether your Commissars will be added only to Slave Armies, only to auxiliaries recruited from that particularly warlike native species you just conquered, or all your armies because you're so uber-authoritarian ... then let's just agree to disagree. It's quite possible that I am in the minority with my intense focus on the game's storytelling potential, but I valued this optional level of fine control as a mechanical representation of my empires' themes, and I remain unconvinced that it absolutely had to be removed just because a portion of the playerbase resented performing 30(!) mouse-clicks per stack of 10 armies instead of just ignoring them like the AI did.

It's a bit sad that it would easily be possible to give both sides of this argument what they want, but apparently there is no room for compromise here.
 
So, I like most of these changes. I think they'll actually perform the way you're describing, and in a positive fashion...unlike certain other changes being made in this update...

I'm pretty indifferent to the removal of Attachments. I actually did use them sometimes, but they were tedious. There really wasn't much wrong with them besides being a bit underpowered...and being super tedious to assign one by one.

My real question is...what about Generals? They're currently pretty useless as well, and that's unfortunate considering how important Admirals are. Though, Admirals may become less important now that there are no doomstack options. Admirals were great because you could put one in charge of ALL of your ships. Whereas Generals only benefitted a small number of armies. Unless you plan on upping the leader caps or something, I don't expect either type will see much use, considering how important governors and scientists are.

Edit: And seriously...why does ANYONE try to play "optimally" in such a suboptimal game? (not that optimal games exist at all, but still). Just do the things that make sense and are interesting in non-moronic ways.
 
Liking the info from this Dev Diary but one thing I want to know.... which I'm sure has been said several times in the last 16 pages. That's a lot of pages... I'm not reading them. Can we PLEASE get the ability to rename military units!?
 
once the system is yours, the starbase is yours and they have to take that down also.

I guess... but presumably, if you've taken the system, the starbase is in shambles.

And this new system sounds like it might really allow for an empire that specializes in ground warfare. If they can't defend the worlds they take, and given that they'll already presumably be at a disadvantage in space, those empires would have a hard time going on the offensive.
 
Last edited:
@Wiz Considering the idea that you'll have to have multiple fleets and high probability that any invading force will consist of a few frontline fleets and a few "bombing" fleets, I would like to suggest forsaking "transport" vessel class all-together and instead make "drop capsule" weapon type for normal ships. It might be a nice thing to also add weapons for actual planetary bombing, removing or significantly crippling fleets bombardment capability without those specialized weapons.

Rationale: from what I've got it's pointless to attack a planet unless all of it's fortresses are destroyed. So basically you'll send a fleet to hover over the planet and will be forced to check once in a while if they've got fortresses or not. Considering that it will take a while and that fleet will be nailed to the planed to the time being - that will hardly be one of your main force fleets. As we already have an algorithm that can tell if your land victory is probable or not for "land armies" order, why not merge the fleet and the transports so you can issue the "invade" order and just wait until it's done and get a nice notification once the invasion is complete. It would obviously require some way to setup an invasion casualty stance, similar to bombardment stance - how much of remaining resistance is considered "ok" to start the land phase.

As a bonus it will also make army fleets less fragile and possibly actually armed and will result in less stupid AI situations when one 500 points station can basically lock a 1k+ points transport fleet (unarmed otherwise) for a few hours of real time. This way it will be a fight and usual fight resolution rules will apply and AI will either win or retreat.

That might be a lot of work on adding new components/visuals though, so maybe consider that as an idea for future work :)

Part of the value of transports, I think, is it keeps you from traveling in an "invasion bubble." With the FTL change, you'll need to keep a secure line of travel back to your home territory that the transports can pass through. Otherwise, if you don't bother to hold what you've taken, the enemy can just take those systems back and cut you off from the troops you'll need to take planets. (And since those planets can pin your fleet down to one system, that will be a big deal.)

Although I could totally see a future special forces feature in a DLC/update that would do something like this. Small troops that can sabotage buildings or degrade enemy morale, and which are dropped from warships. That would be great. (I've mentioned John Scalzi's Old Man's War before... imho his ghost troopers would be the perfect fit for something like this.)
 
Last edited:
So, I like most of these changes. I think they'll actually perform the way you're describing, and in a positive fashion...unlike certain other changes being made in this update...

I'm pretty indifferent to the removal of Attachments. I actually did use them sometimes, but they were tedious. There really wasn't much wrong with them besides being a bit underpowered...and being super tedious to assign one by one.

My real question is...what about Generals? They're currently pretty useless as well, and that's unfortunate considering how important Admirals are. Though, Admirals may become less important now that there are no doomstack options. Admirals were great because you could put one in charge of ALL of your ships. Whereas Generals only benefitted a small number of armies. Unless you plan on upping the leader caps or something, I don't expect either type will see much use, considering how important governors and scientists are.

Edit: And seriously...why does ANYONE try to play "optimally" in such a suboptimal game? (not that optimal games exist at all, but still). Just do the things that make sense and are interesting in non-moronic ways.

I agree. And this worries me about what they're going to do about Generals as I feel Generals are just as important as Governors, Scientists, and Admirals are. I don't think they should be gotten rid of, but somehow made more valuable just like how they made Admirals more important.

I feel that attachments can be quite fun and engaging (and yes a bit tedious at times) but overall they added a sense of wonder and imagination to the game.

I understand that the folks at Paradox had a tough time coming to the decision to remove attachments and it probably wasn't an easy one as well. I don't think they'll remove the ability to mod attachments back into the game (@Wiz clarification on that please?) as they're probably people like me who liked playing with attachments. So thank you Paradox for all the hard work you did working on attachments and I do hope that in the future you'll figure out a way to bring them back.

It does intrigue me to see what they do with the tech that gave us attachments but overall I'm happy with the new changes (except for the removal of attachments).


Attachments have never been "optimal". Adding Hunter-Killer Drones to an Assault army provides a 15% increase to damage, at the cost of 83% of simply recruiting a second Assault Army for a 100% bonus. Even with Gene Warriors the disparity shrinks only to 16.6% (assuming equalization of EC to Minerals), which is still more than the 15% buff, and that is before we consider the added hitpoints. Even if we add the impact of maintenance, an army that will pump all its resources into quantity rather than quality will always win. Wiz referred to this himself when he mentioned the impact of cheap clone troops in this very thread, even if he meant his in a more general sense.

Let's take a look at the stats as per the wiki:
  • Clone Army: 60 minerals -- average 281.5 DPS, 250 Health
  • Gene Warriors w/ Hunter-Killer Drones: 300 minerals, 50 energy -- average 517.5 DPS, 400 Health
For the price of a single unit of Gene Warriors with HK Attachment you get 5 Clone Armies with a total of 1407.5 DPS and 1250 Health. In addition you save 250 Energy. The Clones' maintenance is somewhat higher (1.65 vs 1.00), but (a) this can be compensated by drawing from the EC you saved at recruitment, and (b) you don't even need 5 Clone Armies to match 1 Gene Warrior unit; even 2 could do the job, for a maintenance of only 0.66 Energy per month.

In short, if you've been using Attachments all this time to play "optimally", you've been gimping yourself. They're a trap option, and although it's true that this might warrant criticism, a player would only have to compare the numbers to notice this.

You'd have a point once we get to 2.0, as then the addition of Combat Width would finally make individual army quality (and thus attachments) more important, but as it seems that you've already unintentionally worked with a non-optimal setup so far, I don't see how it should get worse if attachments were retained and you'd simply ignore them. And just to point it out again: the AI does not use attachments either, and they are easy to mod out of the game altogether if you really wanted to go that far. There is no need to take them away for everyone else.

Alas, they will quite likely be impossible to mod back in, as I imagine PDX will remove the necessary strings from the hardcoded vanilla files, preventing mods that seek to reintroduce this feature from doing so.

By the same logic we should not have a feature to rename ships or characters. I disagree with the notion that it is wrong to have a feature that adds flavor only (though it would have changed with 2.0 anyways, see above).

The attachment system enabled you to customize various combinations of army, species and addon depending on the theme you were going for. Now the appropriate technologies will become active automatically for all your armies, with no respect to how much that might actually fit to your empire's headcanon.

It's bad enough that our Slave Armies all wear Power Armor -- this is how it works right now, because the equipment in question is handled as a technology rather than an attachment already. It's the perfect example of how it will apply to all attachments with Cherryh, and how awkward the end result may appear if you value flavor.

If you simply don't care about whether your Commissars will be added only to Slave Armies, only to auxiliaries recruited from that particularly warlike native species you just conquered, or all your armies because you're so uber-authoritarian ... then let's just agree to disagree. It's quite possible that I am in the minority with my intense focus on the game's storytelling potential, but I valued this optional level of fine control as a mechanical representation of my empires' themes, and I remain unconvinced that it absolutely had to be removed just because a portion of the playerbase resented performing 30(!) mouse-clicks per stack of 10 armies instead of just ignoring them like the AI did.

It's a bit sad that it would easily be possible to give both sides of this argument what they want, but apparently there is no room for compromise here.

I accidentally hit the wrong button. I was actually agreeing with you as I too am in the same minority for valuing story telling purposes in terms of the attachments and how they fit my empires' themes. I don't think that attachments should had been removed entirely and that some sort of compromise could have been reached.
 
Last edited:
so, assault armies are always spacebased and defence armies automated/building base
so?
there will be no way to bring in additional off world garrisons or etc when rebellion is likely?
 
My real question is...what about Generals? They're currently pretty useless as well, and that's unfortunate considering how important Admirals are. Though, Admirals may become less important now that there are no doomstack options. Admirals were great because you could put one in charge of ALL of your ships. Whereas Generals only benefitted a small number of armies. Unless you plan on upping the leader caps or something, I don't expect either type will see much use, considering how important governors and scientists are.

I think with these changes generals will be more useful, mostly because of combat width. With generals, your armies are stronger so you can minimize losses with them. Before Cherryh people dont use generals, because its more efficient to build more armies than assign generals (mostly), but after Cherryh as you cant throw 1000 armies at once into an enemy planet and come out without losses, you will want to save those Gene Warriors with the extra strength a general gives.

For defending generals... well you mostly need obly one, and drop him wherever you need him (you can spot a transport fleet coming to your planet, and reassign a leader somewhere else takes only 10 days).

And I dont know about the leader cap, unless you had like 5+ sectors and/or a lot of science ships, leader cap were never an issue. I was fine without any bonus leader cap civic/tech since 1 governor is enough for all core systems. After cherryh I will think about picking that aristocratic elite civic.
 
Last edited:
Will the habitability of a world impact the effectiveness of the preferred climate of combat troops? Something like combat effectiveness being reduced by how inhospitable a world is to whatever species an army is composed of, so invading a tomb world for example would be more difficult for soldiers adapted to any of the normally habitable climate types. If such a mechanic was introduced then robotic/machine armies would have an advantage over organic troops by being able to ignore world habitability. Techs improving the base habitability of planets could also be tweaked to render the penalty to be negligible or totally absent for late-game armies.
 
Some of these changes such as combat width I'm fine about, others I'm less keen on. I used assault armies to garrison all the time. Attachments I'm torn on. Never used them but I feel that instead of removing them they should have been reworked and given a ui makeover.

As with so much else in 2.0 its a case of wait and see I guess.
 
Last edited:
This means that it is no longer possible to take a well defended world without losses by simply throwing a hundred clone armies at it: If you wish to minimize losses (and thus War Exhaustion), you will need to invest in expensive, high-maintenance elite armies.

Hoping for good balance. A militarist empire shouldnt get much War Exhaustion for throwing in clone or slave armies. Even mass armies should be a tactic for some empires.
 
Last edited:
This means that it is no longer possible to take a well defended world without losses by simply throwing a hundred clone armies at it: If you wish to minimize losses (and thus War Exhaustion), you will need to invest in expensive, high-maintenance elite armies.

I guess I can still spam my clone armies and let some of them die. That's fine, I didn't expect them to survive.