HOI4 - Development Diary - October 12th 2016

  • 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.
Yes, they did. That is why when they announced Spearhead is gonna be paid feature general reaction was "WAT?!". Because what is Spearhead if not an improvement people are looking for? Taking that the possibility of pincer attack is presented on the official wiki I presume we can call it "normal".

I am looking forward to see what can be achieved in patch alone though.

The problem I have with this is that an armoured spearhead attack is almost as trademark as it gets when it comes to European WWII offensives.

How will this work if you don't have the DLC? Will the AI - Germany still be useless at punching through lines with tank divisions? Or will the AI be allowed to use it, just not the player? If that's the case, then how can they justify charging for a feature that they deem essential for the AI to work properly, but not essential for the players game experience?
 
The problem I have with this is that an armoured spearhead attack is almost as trademark as it gets when it comes to European WWII offensives.

How will this work if you don't have the DLC? Will the AI - Germany still be useless at punching through lines with tank divisions? Or will the AI be allowed to use it, just not the player? If that's the case, then how can they justify charging for a feature that they deem essential for the AI to work properly, but not essential for the players game experience?

When this function was announced devs weren't sure yet.

The intention is that it will be able to do this, though it is still a work in progress. While I would prefer the AI to play the game as much on the players terms as possible (and NOT use the blitz if the player is not capable of doing so), it may not be up to me, and it may turn out that is is a significant enough improvement to be desirable regardless of if the player can use it or not.
https://forum.paradoxplaza.com/foru...october-12th-2016.973797/page-2#post-21944955

Personally I would prefer AI to use it no matter what. Some extra difficulty is always nice.
 
How will this work if you don't have the DLC? Will the AI - Germany still be useless at punching through lines with tank divisions? Or will the AI be allowed to use it, just not the player? If that's the case, then how can they justify charging for a feature that they deem essential for the AI to work properly, but not essential for the players game experience?

Humans can already achieve it manually, so it's better for the AI to have it even if humans don't, since they can compensate. The current situation where the AI can't do it but humans can is part of the reason for vast imbalances / cakewalk AI.
 
  • 2
Reactions:
If you really thought that the Battleplanner would be able to micro your Panzers for you just as good as a player can, then I would say you have unrealistic expectations. Nothing Paradox wrote or said before launch would even "imply" this would be the case in the actual game...

Looking at how HoI3 AI control functioned, and adding in that the focus of how offensives are drawn in HoI4 is not the breakthrough arrow, but the end line you expect all your units to reach, it's not hard to figure out beforehand that the battle-planner would be a tool for broader assaults mainly, and one that will not handle tanks differently from infantry unless you give them different orders.

Having unrealistic expectations on a product =/= "being lied to".

Okay, at this point I'm pretty sure that you're either taking the piss, or that I wasn't clear enough, so I'm going to explain.
At no point did I say that I was expecting human-like behaviour. What I'm asking is that the AI executes plans correctly. For example, if I give an army only a frontline, or a fallback line, they should occupy provinces where their defensive values will be the highest possible, for example mountaineers in mountains, tanks in the plains. If I give them a frontline and an offensive line, their attack values should be optimised, for example mountaineers being placed to attack enemy mountains, panzers being placed to attack enemy plains, marines handling river crossings. Since the battle planner actually shows you what direction the offensive will take, attack values in those provinces should also be part of the equation when deciding which way to send units of certain types. I'm not asking perfection, I'm asking that the AI doesn't send all my panzers into a swamp, and my mountaineers into the Ukrainian plains. There's no human-like behaviour there, just number-crunching. My expectations are about as unreasonable as expecting the AI in an RTS to distinguish between air and land units.
HOI3 is a 7 year old game, that was vastly different from its predecessor, and was widely criticised. Is it also unreasonable to expect improvement after 7 years?
Again, let me quote the dev diary about battle plans.

Offensive lines and frontlines may also be overlapping. I prefer to assign a big front line of infantry to a frontline with a broad attack plan, and then manage my panzer divisions or other breakthrough units along smaller sections of that front and with a much narrower offensive line. In my barbarossa plan above Guderians panzer divisions are operating in this way together with von Rundstedt's infantry.

The keywords being "I prefer". Not I have to.


No I just felt like it was "implied" by the post about equally much as paradox "implied" that the Battleplanner would handle tanks by not saying it wouldn't ;)

Based on what exactly?


And they also implied that the normal battleplan offensives will be improved in the future and for patch 1.3 as well, for free. (

Quite sadly, I didn't buy the game on launch for it to function a year later.




This is why I don't use field marshals right now (unless I just don't give a damn about a theater). I acknowledge that the AI has the same "division shuffle" problem it has always had. I make front assignments shorter to forestall this issue.

I'm not talking about the whole "division shuffle" issue, I wasn't expecting them to get it to work perfectly. What I'm talking about is the AI specifically using the strategic redeployment command on frontline units in the release version, which instantly drops their ogranization to 0. This, combined with the AI reassigning units to random pockets, especially when micro-managing pincer attacks, meant that the whole frontline became extremely chaotic, with the battle planner being a more dangerous foe than any major power. This is a design flaw that should have been detected and fixed after some basic testing, and, in my opinion, proves that the game is unfinished and rushed.
 
  • 1
  • 1
Reactions:
For example, if I give an army only a frontline, or a fallback line, they should occupy provinces where their defensive values will be the highest possible, for example mountaineers in mountains, tanks in the plains.

Tanks will have higher defensive values in mountains then they will in plains (due to penalties to enemy)... So your plan for the best AI is to throw all units into mountains and leave plains undefended?

Because that is how an AI operates, it follows your exact commands... If you code an AI to defend where units are strongest the silly situation above would be the logical result. Coding basic things that you feel make sense and are super easy... is really hard to do for an AI.

I'm just trying to illustrate a case point with the above.

My expectations are about as unreasonable as expecting the AI in an RTS to distinguish between air and land units.

The Joke is on you. Actually most RTS AIs don't distinguish between air and land units at all for how they are handled. :) ( For example helicopters ). Pathfinding (basic gamerules you and AI follow) takes care of how they are allowed to move so the AI just tells what to attack and don't care if it's a tank or helicopter doing it.

The keywords being "I prefer". Not I have to.

Indeed. You don't have to use the battleplanner, and you clearly don't prefer it either, so why are you torturing yourself trying to get it to do things it wasn't intended to do? ( like handling encirclements and tanks to good efficiency ).

Quite sadly, I didn't buy the game on launch for it to function a year later.

Can you win with only the battleplanner currently?

Yes, easily - so clearly the game functions. You can play it and win with features as advertised.

Is there room for improvements? Ofcourse, in all areas, and AI specifically.
 
  • 3
  • 1
  • 1
Reactions:
Tanks will have higher defensive values in mountains then they will in plains (due to penalties to enemy)... So your plan for the best AI is to throw all units into mountains and leave plains undefended?

Because that is how an AI operates, it follows your exact commands... If you code an AI to defend where units are strongest the silly situation above would be the logical result. Coding basic things that you feel make sense and are super easy... is really hard to do for an AI.

I'm just trying to illustrate a case point with the above.

No, I'm saying it should optimise. Mountaineers and infantry are easier to break through on plains, so the units that are easier to be broken through should be placed in provinces that help them defend (mountains and river crossings), while tougher units that can hold their ground without bonuses, infantry with heavy tanks and armoured divisons, should be placed on plains.


The Joke is on you. Actually most RTS AIs don't distinguish between air and land units at all for how they are handled. :) ( For example helicopters ). Pathfinding (basic gamerules you and AI follow) takes care of how they are allowed to move so the AI just tells what to attack and don't care if it's a tank or helicopter doing it.

Yes, and the pathfinding doesn't run land units into uncrossable terrain, and air units don't avoid it.

Indeed. You don't have to use the battleplanner, and you clearly don't prefer it either, so why are you torturing yourself trying to get it to do things it wasn't intended to do? ( like handling encirclements and tanks to good efficiency ).

Because if someone tells me that they do something a certain way because that's how they prefer it, then my assumption won't be that it's impossible to do it any other way. If, let's say, you knew nothing about a game like Call of Duty, and a developer told you "Yeah, it's a shooter, I prefer to play it in first person mode" what would you assume? Would you assume that the game
A: Has no third person mode.
B: Has a third person mode, but using it makes the game nigh on unplayable.
C: Has a third person mode, the dev just doesn't like it.
One of the reasons for the implementation of the battle-planner they gave was to decrease the amount of micromanagement needed on secondary fronts. If I have to manage every single panzer division manually, then that hasn't really been achieved, has it?

Can you win with only the battleplanner currently?

Yes, easily - so clearly the game functions. You can play it and win with features as advertised.

Yeah, and in Big Rigs - Over the Road Racing your AI opponent didn't move. Therefore you could win, and the game functioned perfectly.
 
  • 4
  • 1
Reactions:
No I just felt like it was "implied" by the post about equally much as paradox "implied" that the Battleplanner would handle tanks by not saying it wouldn't ;)

Well, if it was equally "implied" like you claim, it is good that you did not pay 40 Dollars to read his comment, isn't it? Because it would be horrible to pay for something that is "implied" and then not get it, right?
 
  • 1
Reactions:
No, I'm saying it should optimise. Mountaineers and infantry are easier to break through on plains, so the units that are easier to be broken through should be placed in provinces that help them defend (mountains and river crossings), while tougher units that can hold their ground without bonuses, infantry with heavy tanks and armoured divisons, should be placed on plains.

Exactly. So suddenly it's not so easy to design an AI anymore, is it? And it also needs to work fine in situation where there are only mountains and only plains, or anything in between...

Further you need to classify when something stops being infantry and starts being tanks or mountaineers. Is a unit with 1 tank + 24 infantry still a tank division for how the Battleplanner should handle it?

Yes, and the pathfinding doesn't run land units into uncrossable terrain, and air units don't avoid it.

Path-finding and AI are two different things. You don't say that your units are under AI control when they path around a cliffside in an RTS, do you?

then my assumption won't be that it's impossible to do it any other way.

It's not impossible to fight using the battle-planner either, since you can even use it for things it was not intended to be used for (as multiple people have shown multiple times). Since all your AI opponents use the battleplanner all their time it clearly can attack and conduct operations. If it was impossible to at all use the AI would not function at all and not move a single unit.

One of the reasons for the implementation of the battle-planner they gave was to decrease the amount of micromanagement needed on secondary fronts. If I have to manage every single panzer division manually, then that hasn't really been achieved, has it?

Only if you have an army made up 100% of tanks... If your army instead is 90% made up of infantry (like most peoples are) the Battleplanner can reduce how many units you need to manage by 90% if it takes care of them for you, right???

Yeah, and in Big Rigs - Over the Road Racing your AI opponent didn't move. Therefore you could win, and the game functioned perfectly.

Good point :) (although I did never say the battleplanner functioned perfectly, only that it's functional with room for improvement).
 
Last edited:
Because if someone tells me that they do something a certain way because that's how they prefer it, then my assumption won't be that it's impossible to do it any other way. If, let's say, you knew nothing about a game like Call of Duty, and a developer told you "Yeah, it's a shooter, I prefer to play it in first person mode" what would you assume? Would you assume that the game
A: Has no third person mode.
B: Has a third person mode, but using it makes the game nigh on unplayable.
C: Has a third person mode, the dev just doesn't like it.

C, out of the choices presented, is the only rational interpretation. It's also closest to what the devs have been saying about HoI4, since they do use the battle planner.

However, I think we need to add one.

D: Has a third person mode, which is better for some things but not for others, so the dev uses it for the things it does well and doesn't use it for the things it doesn't do well.

An RW example of this would be the Jedi Knight games. It's fairly typical to play those in FPS mode when shooting but switch to third-person when fighting with your lightsaber. Some people play Skyrim this way too, though I don't know if a consensus ever emerged about the best camera uses.
 
I'm curious because I don't really understand what people mean when they write this. I can get why people might not like the way the battle planner works, but how exactly is it broken?
I think because of shuffling? Is there anything else?
 
I'm curious because I don't really understand what people mean when they write this. I can get why people might not like the way the battle planner works, but how exactly is it broken?
I think because of shuffling? Is there anything else?

The frontlines getting longer and longer the more your units advance. Also leading to useless redeployment of your units.
 
The frontlines getting longer and longer the more your units advance. Also leading to useless redeployment of your units.
How horrible! You mean they don't abandon half the front?
 
  • 3
  • 1
Reactions:
I'm pretty sure that is as designed.

How horrible! You mean they don't abandon half the front?

I'm guessing the issue @Lifthrasil is referring to is when you have multiple adjacent frontlines, and they overlap or do other funky things. For example, if I have a front with three 24-division armies with equal frontage each over 30 provinces, say, it's not unheard of two end up with two of those armies to take up all of the frontage bar one province, with the third army's front being squashed down to one province.

This, to me, is the lesser of two evils, as I find if those three armies fronts start to overlap such that all three fronts cover the same 30 provinces after a little while, is more frustrating, as the troops are a mixed-up mess between the armies.

Both are solvable with lots of repetitive battlefront micro, so it's not as if they're not manageable, but it looks like the Spearhead/Blitz feature will make those kind of situations far more enjoyable, and I can't imagine given the general anti-micro (UI lack of notifications leading to map micro aside) approach to design that the battlefront micro if taking that approach to the game is the devs goal.

Or I could be wrong - apologies if I am Lifthrasil.
 
  • 1
Reactions:
I'm guessing the issue @Lifthrasil is referring to is when you have multiple adjacent frontlines, and they overlap or do other funky things. For example, if I have a front with three 24-division armies with equal frontage each over 30 provinces, say, it's not unheard of two end up with two of those armies to take up all of the frontage bar one province, with the third army's front being squashed down to one province.

This, to me, is the lesser of two evils, as I find if those three armies fronts start to overlap such that all three fronts cover the same 30 provinces after a little while, is more frustrating, as the troops are a mixed-up mess between the armies.

Both are solvable with lots of repetitive battlefront micro, so it's not as if they're not manageable, but it looks like the Spearhead/Blitz feature will make those kind of situations far more enjoyable, and I can't imagine given the general anti-micro (UI lack of notifications leading to map micro aside) approach to design that the battlefront micro if taking that approach to the game is the devs goal.

Or I could be wrong - apologies if I am Lifthrasil.
The way fronts work doesn't take into account multiple armies on the same front. It treats it like its the only army ok the front
 
The way fronts work doesn't take into account multiple armies on the same front. It treats it like its the only army ok the front

I'm not criticising, just asking, but are you sure this is the case? From what I've seen, it seems to take some account of multiple fronts (for example, it will often not (but not always) overlap multiple fronts (as long as there's no initial overlap, and no nearby coastlines), but rather just squish some up and spread others out, which would be a highly improbably outcome if no account of multiple armies were taken).
 
How horrible! You mean they don't abandon half the front?

Actually they do. These troop redeployments are beyonf catastrophic. Or do you actually believe that any field marshall or general would redeploy division from the eastern front to the western front in the middle of a push? Constant thinning of parts of the front line where the opposition has troops to spearhead through your front line and so on.

Anyways, people who are happy with this can keep paying. I wont.
 
Actually they do. These troop redeployments are beyonf catastrophic. Or do you actually believe that any field marshall or general would redeploy division from the eastern front to the western front in the middle of a push? Constant thinning of parts of the front line where the opposition has troops to spearhead through your front line and so on.

Anyways, people who are happy with this can keep paying. I wont.
They don't abandon it. There's an empty spot. The AI finds the closest(I'm assuming) unit that isn't alone. It moves it. Of course currently that part isn't working but if you read the diaries you would know they are fixing that. And the original post I was quoting before that was them saying that they had to pay to fix it
 
  • 1
Reactions: