• 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.

HOI4 Dev Diary - Chain of Command

Hi everyone and welcome back to regular dev diaries. This and upcoming diaries will be covering stuff happening in the 1.5 "Cornflakes" update as well as the unannounced expansion that will come out together with it. One of the main focuses of those can be summarized as "making players care more about armies, leaders and troops" (our DLCs tend to have 1-3 main focuses or missions). The first feature that touches on this, and the topic of today's dev diary is adding a military chain of command to the game.

After Hearts of Iron III, where something like organizing the soviet chain of command could take about an hour of the players time we decided that we wanted something that was a lot less effort to work with for HOI4. We basically settled on a flat level with field marshals with no restriction on commanded divisions, and generals with a limit on division count but with a different set of traits. Over time we felt that we lost a bit too much of the WW2 military flavor with this abstraction, so we started thinking about how to do it in a more interesting way.

Pasted image at 2017_09_13 02_48 PM.jpg

What we have done now for 1.5 is that field marshals are now leading an Army Group, which is a certain number of Armies (what we had before) led by Generals. There are then places in theaters as before. Theaters are like before just a geographical organizational tool for the player and don't have a commander or the like to keep them as flexible as possible. This means that we have a Theaters->Army Groups->Armies->Divisions structure now.
While the Generals still come with a soft cap for how many divisions they can efficiently command, the field marshals will now have a number of armies they can efficiently command.

I also want to make sure to point out that this is still very early on in development, so stuff is very likely to change, and some stuff aren't completely working as it should yet. So we are showing you this in progress rather than showing a completely finished feature, and as always any numbers you see are extremely subject to change ;) Also I very sneekily hid the topbar for now ;)
upload_2017-9-13_15-27-6.png


When it comes to controlling your troops the new system introduces some changes to the battle planner. You can either do a plan for each army in the army group, or have a central plan for the whole Army Group where each army has a part of the frontline assigned as its responsibility. You can also do a mix, in which case an Army will finish its plan and then fall back to executing the Army Group's plan. We are still iteration on this stuff though but I figured you all wanted to know how it would work in practice.

upload_2017-9-13_15-26-30.png


Something that does not really come across in the images is that we are working on ways to streamline the process for setting up fronts using the new army groups. This should make at least the basic cases feel smooth to set up, even with one more command level and more armies without a ton of extra clicking.

upload_2017-9-13_15-23-51.png


The sharp eyed reader will also notice that we have removed the skill level for generals. This is now replaced with separate skills of different kinds. Attack, Defense, Planning and Logistics. Attack and Defense do what you expect while Planning improves planning speed and Logistics lowers supply consumption. Field marshal stats apply together with army general stats at a reduced capacity, so you will always want to have a chain of command for best efficiency.

The chain of command feature is going to be part of the free update, although there is some cool DLC features that tie into it we will be revealing in later diaries. Also expect to read more details about the system itself like how things in combat are affected etc.

See you next week when we will be taking a look at national unity...
 
Question about how it is developing. The current system has a problem with divisions getting shuffled all up and down a front line. This change seems to allow for defining lanes, or clear AOs within a master plan. Do divisions stay in their lanes?

I might suggest not capping the number of divisions a general can control, but rather the width of an army group front. The eastern front could have a theater commander with 3 army groups that have limited width but can contain an unlimited number of divisions.

Also, think about having divisions attach directly to the FM, which would act as a reserve. Army groups needing support could pull divisions from the reserve, damaged units are rotated back to the reserve for recovery, and the AI could balance forces by pulling units from one AG back to the theatre reserve and then sending them where needed.

The holy grail would be unified front planning where I could assign a FM, draw the AGs with their desired division types and manuver, and then hit go. The FM then assesses and requests I attach the nessary forces to his reserve, and from there he assigns them to the AGs. If i lack it, then i try to build it, or I can flag a planned AG as Allied and then the call goes out for an expeditionary force. (Which could clean up that whole mess. The only EFs sent would be what was needed for a plan).

If things could get to that point, this game would be incredable. Plan driven AI! This the proposed update is a step in the right direction.

Actually, this can be solved by having HQ units, and having the subordinate unit be limited to a certain distance from its HQ, hence the shuffling, if any, will be limited to units within the HQ. And each HQ can have its own objective. If anything, this may actually simplify the AI.

Another thing I would like to have is to assign independent battalions to the HQ/Generals, so he can add those to battle without affecting the width (1 per 20 width?). Historically, there are lots of independent sub units, and currently a lot of the sub-units are underused due to game mechanics.
 
Actually, this can be solved by having HQ units, and having the subordinate unit be limited to a certain distance from its HQ, hence the shuffling, if any, will be limited to units within the HQ. And each HQ can have its own objective. If anything, this may actually simplify the AI.

Another thing I would like to have is to assign independent battalions to the HQ/Generals, so he can add those to battle without affecting the width (1 per 20 width?). Historically, there are lots of independent sub units, and currently a lot of the sub-units are underused due to game mechanics.

So... A return to the hoi3 system? I'll pass thanks.
 
Why? An army level objective would be equivalent of the current one. And you can now specifies army and even corp level objectives. In contrast, HOI3 only allow you to set divisional objectives.
On map hq and command range all helped to contribute to the grind of army management in hoi3. 4 may not be perfect but at least those devil's are gone.
 
No, they really are annoying.
 
A lot of us don't want it back -- Paradox included!
I know, Fads change over time. What one generation think is good another thinks is bad. But eternal truths, like modern Armies use OOBs, remain even if unpopular with this generation of HoI players.
 
I know, Fads change over time. What one generation think is good another thinks is bad. But eternal truths, like modern Armies use OOBs, remain even if unpopular with this generation of HoI players.
Your (flawed) argument is essentially that every WW2 strategy game should have the mechanics of a Gary Grigsby tite. Funnily enough, those games - as good as they are - do not appeal to casual gamers, a market Paradox have been trying - with great success - to crack for some years now.

If you want to use vaguely Orwellian terms like 'eternal truths' you may, but that does not change the fact that complex, confusing, unfun, non user-friendly systems like a full-scale OOB are divisive whilst adding little in the way of playability. I have played every HoI since II, and I don't miss this...

r8A9z1o.jpg


It is neither fun nor rewarding, and it doesn't make the game feel more 'cool' or 'interesting.'

Your view is a minority one, and if such a system was implemented the user numbers would absolutely plummet -- this may not matter to you, but I can assure you it does to Paradox.
 
Your (flawed) argument is essentially that every WW2 strategy game should have the mechanics of a Gary Grigsby tite. Funnily enough, those games - as good as they are - do not appeal to casual gamers, a market Paradox have been trying - with great success - to crack for some years now.

If you want to use vaguely Orwellian terms like 'eternal truths' you may, but that does not change the fact that complex, confusing, unfun, non user-friendly systems like a full-scale OOB are divisive whilst adding little in the way of playability. I have played every HoI since II, and I don't miss this...

r8A9z1o.jpg


It is neither fun nor rewarding, and it doesn't make the game feel more 'cool' or 'interesting.'

Your view is a minority one, and if such a system was implemented the user numbers would absolutely plummet -- this may not matter to you, but I can assure you it does to Paradox.

NOOO Make it STOP!!!!! My Eyes are bleeding!
 
complex, confusing, unfun, non user-friendly systems
That was the problem with it. It need to be clear, fun (immersive) & user-friendly. Just because it was badly implemented in HoI 3 does not mean the idea for it is bad.

And I fully agree I am in the minority.

Also HoI 3 TFH Sucks!!!!! I would not play it for an hour! I love HoI 3 Black ICE! The BICE team fixed the command range issue by making it much more reasonable.
 
Also HoI 3 TFH Sucks!!!!! I would not play it for an hour! I love HoI 3 Black ICE! The BICE team fixed the command range issue by making it much more reasonable.

Hey, I liked HoI3 TfH :p. I do think HoI4 is better (and I do like BICE, but sadly HoI3 BICE was a bit crashy for me) though :D. I'm not getting involved in any HQ argument malarkey, but if we start talking about flagships and tactical plotting I'll be there :p.
 
We think we may have found out why many of the crashes happen. Many text files need to be in ANSI, bu to add non standard letters they were in BOM format. Testing to see if this fixes things.

Cheers Gamer, great to hear :).
 
The thing that got me with the HoI3 OOB is that the AI just moved things around as a blob-o-divisions. It still essentially does that. I think it could be useful to have a sub-dividing logic (without really the need to create permanent corps. Imagine drawing a front line that automatically creates corps lanes as you draw it across. A corps would have a width of maximum 2 provinces, and the divisions would align themselves within that width with some units forward and a few 1 province back as a reserve. In the battle planner, an army covering a front of 8 provinces would create 4 corps lanes, with 4 arrows going to the objective line (rather than the single arrow as is now the case). The army would have divisions behind the silos as a reserve. If as it advances, the army frontage expands to 10 provinces, a 5 corps lane would be automatically created and populated with divisions from the army's reserve. If the front contracts, a lane would be eliminated, and its divisions sent back to the army level reserve.

I don't think there is a need to create corps HQ units, or corps commanders etc, but having a corps logic in how the AI arrays its forces could be very useful. Instead of a front AI that every day tries to manage all divisions all at once, break it down. Every day, from right to left corps lanes give orders to their 5-10 divisions to fight the battle in their lane according to their battle planner objective. Every 2 days, army logic fires to allocate resources to corp lanes (give a lane troops from the reserve, remove divisions to the reserve from a lane, or request forces from Army Group Reserve) and to adjust corp lane attack arrows if necessary. Every 5 days army group logic fires to allocate resources and adjust (if necessary) objective lines for the armies. Finally every 15 days, a strategic level check happens to assign general objectives to army groups, assign divisions to army groups, and adjust production and division recruitment strategy if necessary.

There wouldn't be a need for any interface changes really, people would still just make army groups and armies, but it could radically help AI performance in how it fights its fight by breaking things down into smaller pieces rather than trying to fight the whole front at once.
 
The thing that got me with the HoI3 OOB is that the AI just moved things around as a blob-o-divisions. It still essentially does that. I think it could be useful to have a sub-dividing logic (without really the need to create permanent corps. Imagine drawing a front line that automatically creates corps lanes as you draw it across. A corps would have a width of maximum 2 provinces, and the divisions would align themselves within that width with some units forward and a few 1 province back as a reserve. In the battle planner, an army covering a front of 8 provinces would create 4 corps lanes, with 4 arrows going to the objective line (rather than the single arrow as is now the case). The army would have divisions behind the silos as a reserve. If as it advances, the army frontage expands to 10 provinces, a 5 corps lane would be automatically created and populated with divisions from the army's reserve. If the front contracts, a lane would be eliminated, and its divisions sent back to the army level reserve.

I don't think there is a need to create corps HQ units, or corps commanders etc, but having a corps logic in how the AI arrays its forces could be very useful. Instead of a front AI that every day tries to manage all divisions all at once, break it down. Every day, from right to left corps lanes give orders to their 5-10 divisions to fight the battle in their lane according to their battle planner objective. Every 2 days, army logic fires to allocate resources to corp lanes (give a lane troops from the reserve, remove divisions to the reserve from a lane, or request forces from Army Group Reserve) and to adjust corp lane attack arrows if necessary. Every 5 days army group logic fires to allocate resources and adjust (if necessary) objective lines for the armies. Finally every 15 days, a strategic level check happens to assign general objectives to army groups, assign divisions to army groups, and adjust production and division recruitment strategy if necessary.

There wouldn't be a need for any interface changes really, people would still just make army groups and armies, but it could radically help AI performance in how it fights its fight by breaking things down into smaller pieces rather than trying to fight the whole front at once.

There is another reason that I think there should be headquarter units -- HQ units can serve as supply node and fix the problem with supply routes. Essentially, supply flows from high level HQ to low level HQ then to individual units. So instead of every units needs to trace route to capital, all it need to do is to trace route to HQ it is subordinate to. So, if you place your high level HQ to a large port, your supply routes through that port instead of god know where the AI trace it to. Much cleaner, much simpler, and much less computational intensive.

Automated reorganization of forces has it uses, but sometimes it screw things up too. You don't want your painstakingly organized mech/tank army being dispersed by an AI that does not know what it is doing, don't you? If you insist on automated for reoganization, then have an option to disable it for certain subunits, else don't bother.
 
Last edited:
At the very least, could we take advantage of a reworking of the army commanders to refine the rosters? Setting Montgomery as a Field Marshal at the very start of the game is deeply a-historical, and the new system should allow them to still give him base skills that lead him toward that path. It might also be nice to limit the number of army commander available to some of the factions to reflect the early army command structures a bit better - maybe set up a system where you can promote from a pool of unavailable commanders at start when you need more actual army commanders to manage your forces.

Also, remove commanders when they're appointed to the high command?
 
Your (flawed) argument is essentially that every WW2 strategy game should have the mechanics of a Gary Grigsby tite. Funnily enough, those games - as good as they are - do not appeal to casual gamers, a market Paradox have been trying - with great success - to crack for some years now.

If you want to use vaguely Orwellian terms like 'eternal truths' you may, but that does not change the fact that complex, confusing, unfun, non user-friendly systems like a full-scale OOB are divisive whilst adding little in the way of playability. I have played every HoI since II, and I don't miss this...

r8A9z1o.jpg


It is neither fun nor rewarding, and it doesn't make the game feel more 'cool' or 'interesting.'

Your view is a minority one, and if such a system was implemented the user numbers would absolutely plummet -- this may not matter to you, but I can assure you it does to Paradox.

Really? The problem with HOI3 is that the AI does not use HQs properly, nor does it respects the command hierarchy -- hence you see the screw up.. If the AI actually respects the hierarchy, it would not have move a division too far away from its corp HQ, etc. And HOI3 the game does not use HQ other than for cosmetics -- the potential is wasted. For example, if HOI3 uses HQs as supply nodes (supply flows from high level HQ to low level HQ), than the supply route would e much cleaner and predictable, and much less computational intensive, as one only need to trace route to its HQ. And each of the HQs can have its own objectives -- lower level HQ objective is limited to within the zone of the high level HQ -- hence choices are much more limited and less computational intensive.

There is a reason why command hierarchy is invented in real life -- it reduces complexity.

Manually managed command hierarchy in HOI3 is a chore, but automated version that respect the hierarchy in HOI4 is much less so -- because you may not need to mess around with each individual divisions anymore. And hierarchy allow you to reduce the display to higher level objectives, so the intention of AI become much more clear.

I don't really see what your objection is. If you only set the topmost level object, then it will behave as the current implementation.
 
There is another reason that I think there should be headquarter units -- HQ units can serve as supply node and fix the problem with supply routes. Essentially, supply flows from high level HQ to low level HQ then to individual units. So instead of every units needs to trace route to capital, all it need to do is to trace route to HQ it is subordinate to. So, if you place your high level HQ to a large port, your supply routes through that port instead of god know where the AI trace it to. Much cleaner, much simpler, and much less computational intensive.

Automated reorganization of forces has it uses, but sometimes it screw things up too. You don't want your painstakingly organized mech/tank army being dispersed by an AI that does not know what it is doing, don't you? If you insist on automated for reoganization, then have an option to disable it for certain subunits, else don't bother.

I think that, in combination with one more level in the chain of command: Army Group, Army, Corps, (with possible 4th level represented by a Fieldmarshall in control of a theater?), the HQ idea with command distance and supply node would be very interesting and perhaps add another dimension to combat with more ability to cut off supply/disrupt organization of units.

Wald
 
At the very least, could we take advantage of a reworking of the army commanders to refine the rosters? Setting Montgomery as a Field Marshal at the very start of the game is deeply a-historical, and the new system should allow them to still give him base skills that lead him toward that path. It might also be nice to limit the number of army commander available to some of the factions to reflect the early army command structures a bit better - maybe set up a system where you can promote from a pool of unavailable commanders at start when you need more actual army commanders to manage your forces.

Also, remove commanders when they're appointed to the high command?

Leader need to have a political power attribute as well. Leaders with higher level attribute (plus political power) get automatically promoted when there is a empty higher level post opening up. If you want to dismiss or promote a leader, then you need to pay in political point to do so.

By the way, this may mean creation of high level commands may no longer be free, else it would have become a loophole for leader promotion.