Air Changes | Developer Corner

[pnt]

Well, I have never seen the HOI4 source code, so I can't say how things actually are calculated.

My understanding is that there is ana initial determination of eligible planes that can participate, which depends on mission efficiency, weather, time of day - and in the case of fighters, also detection. If the net number of participants on one side is at least a factor 2 higher than the opponents, then my understanding is that additional bonuses are applied. But then there are limiting caps, and stats of participating aircraft are folded in, so knowing some of the variables does not fully reveal the underlying formula.

However, details aside, the current algorithm favors larger formations (well, at least up to a point), whereas in reality large formations tended to hit diminishing returns. Of course, having many fighters fighting each other in an air zone does not mean that they are all in the same spot). But bombers were more survivable in large "herds" (just like prey animals), and were typically attacked in sequence where escorts were peeled off, and then formations were broken up so that finally smaller groups of bombers could be taken down separately. Having superior numbers was still helpful, but there are many ways air combat could be made much more tactical and more streamlined at the same time. And i looks like Paradox is heading in that direction.

 

[HES]

Thanks for your reply PNT

I found it difficult to fully explain the nature of the problem here and probably muddied it by referring to the Battle of Britain when I really had the Battle for France in mind.

In 1940 before the fall of France, the British split the RAF between France and the UK, keeping some fighters home to protect against air attack. In the game, I find the AI immediately sends tactical bombers against the UK as soon as war commences and in this scenario, most raids are around 400-450 bombers hitting a single region.

So my defence is to have at least 400 fighters ready to defend any targeted region and even at 400:450 it is enough to discourage the AI and call off the raids on that region. The AI then merely moves its bombers to another region without fighters, causing me to relocate the fighters from the first region to the newly contested region, until the AI realises it is contested again and withdraws moving to another region to do it all over again.

The 400 odd fighters left for home defence are sufficient to discourage the AI and get them to go off and find other targets after a month or so of probing the air defences, losses aren't a problem not least as I deliberately build high quality fighters and they are aided by AA and radar. The key is to disrupt the bombers rather than shoot them all down.

The trouble is that to cover all 4 air zones in the UK in a 'set and forget' mode to see these tactical bomber raids off, (really important if France is fighting for its life) I need 400 fighters in each region, meaning a total of 1600. This is excessive, unrealistic and not actually necessary, if the fighters can protect all of the region in the UK within their range. Placing the fighters in Yorkshire allows the range to be covered without having to spec into that in the fighter design.

Italy, France and Germany all have the same problem when they are under attack and not being able to just allocate sufficient fighters to manage home air defence, where all regions are within range of the airbase, is quite frustrating. Being able to set some fighters to home defence would then allow you to use the surplus fighters to engage in the actual land battles, very important for France in particular.

 

[pnt]

Yes, at the moment it is an unavoidable game of whack-a-mole, where you have to swap the mission areas to match the AIs. As I mentioned above, I am hoping that the path Paradox has embarked upon can in the longer term resolve this at a more fundamental level. However, in the short term, my experience is that a mixed approach is usually helpful.

If you play Germany, max out radars in Schleswig-Holstein and Brittany (locations that also help you with naval combat) to improve detection, and get the centralized control air force spirit as soon as possible. This gives you an immediate, flat 10% bonus to mission efficiency when switching mission areas. Then, I found that having 100-200 fighters permanently assigned to at-risk areas in combination with "fire brigades" that keep switching between provinces worked best in 1.11.11, since you start out with a sufficiently high number of fighters to discourage the AI, while at the same time leaving no empty targets to swap to. In addition, it can be worth building some AA in high-infra provinces. AA is currently a bit OP compared with fighters - but is not a long-term investment in the sense that once you win you can't move it to a more suitable location. But this is just my experience. Others may have better short ideas!

 

[DukofDeth]

Thanks for your reply PNT

I found it difficult to fully explain the nature of the problem here and probably muddied it by referring to the Battle of Britain when I really had the Battle for France in mind.

In 1940 before the fall of France, the British split the RAF between France and the UK, keeping some fighters home to protect against air attack. In the game, I find the AI immediately sends tactical bombers against the UK as soon as war commences and in this scenario, most raids are around 400-450 bombers hitting a single region.

So my defence is to have at least 400 fighters ready to defend any targeted region and even at 400:450 it is enough to discourage the AI and call off the raids on that region. The AI then merely moves its bombers to another region without fighters, causing me to relocate the fighters from the first region to the newly contested region, until the AI realises it is contested again and withdraws moving to another region to do it all over again.

The 400 odd fighters left for home defence are sufficient to discourage the AI and get them to go off and find other targets after a month or so of probing the air defences, losses aren't a problem not least as I deliberately build high quality fighters and they are aided by AA and radar. The key is to disrupt the bombers rather than shoot them all down.

The trouble is that to cover all 4 air zones in the UK in a 'set and forget' mode to see these tactical bomber raids off, (really important if France is fighting for its life) I need 400 fighters in each region, meaning a total of 1600. This is excessive, unrealistic and not actually necessary, if the fighters can protect all of the region in the UK within their range. Placing the fighters in Yorkshire allows the range to be covered without having to spec into that in the fighter design.

Italy, France and Germany all have the same problem when they are under attack and not being able to just allocate sufficient fighters to manage home air defence, where all regions are within range of the airbase, is quite frustrating. Being able to set some fighters to home defence would then allow you to use the surplus fighters to engage in the actual land battles, very important for France in particular.

You only need a bit over 300 - there's a bit of code in the defines that means that if there are more than 300 aircraft in the zone on air defense it will move on. Those 300+ could the the crappiest ones you have; the AI just notes the numbers and moves on. With 310 aircraft per zone [set them to high priority so accidents won't run them down too much] you should be fine. I've never seen the German AI reach out to the Irish zone even after France has fallen.

 

[pnt]

That sounds a little odd. I have seen the AI send aircraft into air zones where there are more than 1000 fighters. Perhaps there is also some other condition for the >300? Like a certain air mission?

 

[GrandVezir]

If all zones in range have 300+ defending FTR/HFTR, the AI has nowhere to divert to and chooses from among the defended zones as it would if they were all undefended.

 

[pnt]

I set up a little test using a recent game. If you look in the attached file, you can see that Germany has 2200 fighters in Ukraine, 1000 in Belarus, and none in the Baltics. And the Soviet counter is 1300 planes in Ukraine. So there has to be a little more to it.

 

[DukofDeth]

That sounds a little odd. I have seen the AI send aircraft into air zones where there are more than 1000 fighters. Perhaps there is also some other condition for the >300? Like a certain air mission?

Sorry - I forgot to be more specific. If there are more than 300 fighters on air defense [intercept or air superiority] then the AI will stop strat bombing that zone. I semi-confirmed this myself just last night. As the Lithuanian-Polish Commonwealth/Kingdom the Soviets were hammering the Baltics because all the fighters assigned [evenly] to my four armies spread along the border all decided to fly in Eastern Poland. So as soon as I could get 300+ into the air in the Baltics the bombing switched to Western Poland till I could get 300+ there. At one point they were still relentlessly bombing me and I had 300 in two groups of 150 so I shoved the size up to 160 each and as soon as the groups began to fill out the bombers stopped.

There was plenty of activity in various zones later all in support of their attacks but no more strat bombing. You can see for yourself.
I do have some mods but just graphical ones - makes it easier to see what is going on.

I was following BitterSteel's guide for "This is Lit" and one of his jokes in the title is that Lithuania is overpowered. Strikes true since the mighty Soviet army is really shying away from engaging me. Aside from one position its just level two and three fors and the army on border is all 9/2 infantry/artillery with some support. My shore guards are just 9/0 infantry - I sort of spammed them out following the first naval invasion.

 

[DukofDeth]

@YaBoy_Bobby I was wondering is - as a part of the air changes - you might add in some of the more popular QoL improvements from the mod community. The whole "Colored Air Force" spectrum would be nice and even the "Colored Navy" icons would also be great. With a simple on-off toggle to accommodate the color blind [and some of the really masochistic purists] that would really improve the presentation of information. With the different aircraft/ship types all color coded and even the different colors to mark airbases that are in use by the player the AI and ones that are empty - it all goes a long way to helping the player follow along.

 

[LordInsanity]

With these QoL changes on the horizon, can the list of air equipment types also be reworked to make it less of a pain to navigate? Usually by lategame it's loaded full of obsolete air equipments made earlier in the game and from capitulated countries.

 

[DukofDeth]

With these QoL changes on the horizon, can the list of air equipment types also be reworked to make it less of a pain to navigate? Usually by lategame it's loaded full of obsolete air equipments made earlier in the game and from capitulated countries.

They are allowing sorting by type which will make it a lot easier. I doubt we will be able to have obsolete gear vanish from the presentation though that would be really nice.

 

[Mister Analyst]

@YaBoy_Bobby

Regarding this:

The final change I want to discuss today is an addition to the combat system. We are adding a new mechanic for intercepting planes in regions enroute to their target. In most cases, engaging planes in their target region will still be most effective. But in others, such as when range is a factor, the region being crossed has a bunch of engagement and spotting bonuses, and in a few other cases leaning into combat in an intermediary region can be a good idea.

Do you have any thoughts on increasing or reworking attrition/accidents to bomber wings for each air region crossed while flying to a target and returning home?

 

[pnt]

Sorry - I forgot to be more specific. If there are more than 300 fighters on air defense [intercept or air superiority] then the AI will stop strat bombing that zone. I semi-confirmed this myself just last night. As the Lithuanian-Polish Commonwealth/Kingdom the Soviets were hammering the Baltics because all the fighters assigned [evenly] to my four armies spread along the border all decided to fly in Eastern Poland. So as soon as I could get 300+ into the air in the Baltics the bombing switched to Western Poland till I could get 300+ there. At one point they were still relentlessly bombing me and I had 300 in two groups of 150 so I shoved the size up to 160 each and as soon as the groups began to fill out the bombers stopped.

There was plenty of activity in various zones later all in support of their attacks but no more strat bombing. You can see for yourself.
I do have some mods but just graphical ones - makes it easier to see what is going on.

I was following BitterSteel's guide for "This is Lit" and one of his jokes in the title is that Lithuania is overpowered. Strikes true since the mighty Soviet army is really shying away from engaging me. Aside from one position its just level two and three fors and the army on border is all 9/2 infantry/artillery with some support. My shore guards are just 9/0 infantry - I sort of spammed them out following the first naval invasion.

Looks nice - but I was still curious about the 300 plane limit. So I continued to try things out in the playthrough posted above (originally intended for Gibraltar-Singapore railroad), and after putting 300 fighters in air zones that the AI liked to strat bomb, but 600 in Leningrad, you can see that it still sends Pe-8's to Northern Front air zone.

 

[DukofDeth]

Here is the line from the defines

STR_BOMB_MIN_ENEMY_FIGHTERS_IN_AREA = 300, -- If amount of enemy fighters is higher than this mission won't perform

So to keep the AI from strategic bombing an air zone - pending any air changes - you just need a bit over 300 in an air zone to prompt the AI to look elsewhere. THe 300+ can be absolute junk bunched into a single air wing and set to intercept - it works.

All other air still goes through. Last night I had 4000+ and even 6000+ over various air zones as I first held off the Germans then pushed them back; they contested me in a single zone.

 

[pnt]

Well, the defines is one thing - how they are actually used in the code is another.

As you can see in the attached screen capture and save, there were 100 strat bombers in an area with 600 fighters on intercept even though there were many uncontested areas, as well as areas with 300 fighters (some of which were also hit). Still, I hope that the AI will in the future be guided by scalable rules rather than fixed numbers, so it is very good that you pointed this out!

 

[xtfoster]

Is there any plan for ading the air high commands in the field like the generals or the admirals? Especially for using more the new mechanic of upgriding highs commands and look more air commanders of WW2

If you read the DD, then you should know the answer.

The first thing I want to say is this is not Army Groups or Fleets. For now, and the foreseeable future, we are not adding Air Marshals or any sort of mechanical impact to using Air Groups. Air Groups are an organizational structure and nothing more.

or by reading the Developer responses

We were initially planning to work aces into the promotion structure but ran into issues with scope and how to make aces exactly fit into something that they currently don't. What I mean is that there are not enough types of aces compared to say relevant air advisor types. This would mean either simplifying the number of air advisor jobs or changing aces to fit the advisory roles. Beyond that, there is the aspect of leveling as an advisor which is linked to field command level in other cases. Again, this would be a lot of time spent reworking something to achieve a solution that doesn't mesh well with the game's other structures and presentation paradigms.

 

[Ariosto Marzani]

A bit late, but I am hoping that there are additional changes to be made to Air_Attack as it applies to Ground Units, as well as General AA.

• At the moment Air_Attack functions as something as an absolute, where each Division that has this capability being effectively mandated a minimum down rate of a single plane per combat tick. I understand why this is given you can't represent damaged planes, but I think it would be better if it were additive so that the SUM of the Divisions' Air_Attack is used instead, with the possibility of no aircraft being shot down unless there is sufficient AA cover amongst the Corps.
  • There should be an added stat regarding elevation effectiveness for both Air_Attack as well as General AA; by this I mean that there is an effective "ceiling" where the Air_Attack gets cut off, or the General AA is no longer able to shot down aircraft; this would be matched against an aircraft's "service ceiling". Only equipment which is capable of reaching said "service ceiling" would be counted in the SUM of the Air_Attack, which would then in turn be modified by the Agility of the Aircraft in question.
    • I'm not sure how workable this would be or the extent of coding that would be required, but I believe it will ultimately be necessary for any calculations for Air_Attack that revolve around Ground Support Missions. As it currently stands, WW1 era AA could down an AC-130 Gunship if given the chance.

 

[wagnerleung0079]

Can we have more air unit-related stats to make air units more interaction with other units?
Like:
Hard and Soft Ground Attack for CAS against hard and soft ground target (Hs 129 B3 and Ju 87 G)
Piercing for Hard Ground Attack
Strategic Bombing targeting affects the efficiency of Strategic Bombing (make player need to invest more resources (bombsight) to make Strategic Bombing worthwhile)
Different Air Attacks for dogfighting fighters and intercepting bombers (like M2 browning is very good for dogfighting fighters, but not as effective as MK 108 for intercepting bombers. On the otherhands MK 108 is very good for intercepting bombers but very bad for dogfighting fighters.)

Different loadout for planes for different missions.
Strategic Bombing, Ground Attack and Naval Attack of an aircraft is more related to the loadout (Naplam, Rocket, Tropedo, AP bomb, gun pods and etc.) of aircrafts rather than their flying characteristic and base weapons. Could the stats of Strategic Bombing, Ground Attack and Naval Attack of aircrafts depends more on something like " loadout module" which every squadrons can select what loadout to use for different missions?

 

[marcelo r. r.]

Thanks for your reply PNT

I found it difficult to fully explain the nature of the problem here and probably muddied it by referring to the Battle of Britain when I really had the Battle for France in mind.

In 1940 before the fall of France, the British split the RAF between France and the UK, keeping some fighters home to protect against air attack. In the game, I find the AI immediately sends tactical bombers against the UK as soon as war commences and in this scenario, most raids are around 400-450 bombers hitting a single region.

So my defence is to have at least 400 fighters ready to defend any targeted region and even at 400:450 it is enough to discourage the AI and call off the raids on that region. The AI then merely moves its bombers to another region without fighters, causing me to relocate the fighters from the first region to the newly contested region, until the AI realises it is contested again and withdraws moving to another region to do it all over again.

The 400 odd fighters left for home defence are sufficient to discourage the AI and get them to go off and find other targets after a month or so of probing the air defences, losses aren't a problem not least as I deliberately build high quality fighters and they are aided by AA and radar. The key is to disrupt the bombers rather than shoot them all down.

The trouble is that to cover all 4 air zones in the UK in a 'set and forget' mode to see these tactical bomber raids off, (really important if France is fighting for its life) I need 400 fighters in each region, meaning a total of 1600. This is excessive, unrealistic and not actually necessary, if the fighters can protect all of the region in the UK within their range. Placing the fighters in Yorkshire allows the range to be covered without having to spec into that in the fighter design.

Italy, France and Germany all have the same problem when they are under attack and not being able to just allocate sufficient fighters to manage home air defence, where all regions are within range of the airbase, is quite frustrating. Being able to set some fighters to home defence would then allow you to use the surplus fighters to engage in the actual land battles, very important for France in particular.

agreed to all.
also i like do the "phoney war", but need a bizarre amout of Fighters.. otherwise, just 1 month of bombing by TAC criple entire germany! these starting bombers are too OP.

 

[xtfoster]

As it currently stands, WW1 era AA could down an AC-130 Gunship if given the chance.

Not to be a nitpicker, but of the 6 AC-130 Gunships lost to enemy action, 3 were shot down by 37mm AA, 1 more was shot down by 57mm AA, only 2 were shot down by missiles (one by an SA-2, the other by a shoulder-fired SA-7). Even though the AC-130 has a service ceiling of 12km, it can't attack from there.

EDIT: I misread that. Those 6 were just those lost in Vietnam, another was lost during during Operation Desert Storm to a Strela-2 shoulder fired missile.