Unable to achieve naval supremacy even though my ships on patrol mission

[SigurdStormhand]

Bingo.

Here's what happens in a typical Germany game:

1. Shred the RAF, achieving total air supremacy in the Channel and Southern England by 1940.
2. Launch naval bombers to obliterate all ships in the Channel.
3. The Home Fleet gets damaged and runs away up north, but still stays on Strike Force.
4. Prepare naval invasion of England.
5. Launch naval invasion... but wait, that's not possible because Strike Force from the Home Fleet is still active.
6. Bait the Home Fleet to come within range of naval bombers where they'll be shredded in minutes... but wait, that's not possible either because Strike Force doesn't respond to enemy fleets patrolling in regions where they're assigned.
7. Give up on disabling the force field from the Home Fleet and just dump 600 prewar submarines in the Channel to overpower it instead through an equally magical mechanic: naval mines.

Of course there are tons of ways to circumvent this system through exploits (e.g. paratroopers, convoy shenanigans, prewar rushing, etc.) but people who want to play without exploits will be met with frustration.

Port Strike - once all the BB's are beat up the Fleet won't sally until they're repaired.

 

[Reman]

Port Strike - once all the BB's are beat up the Fleet won't sally until they're repaired.

Enemy ships are out of range of naval bombers in this scenario. They're chilling in Scotland/Orkney, so port striking isn't possible.

 

[atm]

Guys it is not anymore question about supremacy, yes it is tricky with strike mission but this issue is caused by not having enough intel in sea region even though there are radar stations around regions, have fleet of 50+ ships on missions there and have 1k + planes with full air superiority. So how can I get enough intel in region so it is not anymore marked:

We know too little about this region!

Any explanation will be welcome, screenshot is in first post.

 

[KDEstroy]

This is by far the least arbitrary solution. The whole thing stops being a problem if it's fixed this way.

The way I see it, this would cause even more problems. When the AI is invading, removing the supremacy requirement would cause the AI to continuously launch suicide naval invasions and lose hundreds of divisions to player submarines. On the other hand, this would make it far, far too easy for the player to naval invade AI-controlled countries, because they don't know how to use convoy raiding fleets effectively.

 

[SigurdStormhand]

Enemy ships are out of range of naval bombers in this scenario. They're chilling in Scotland/Orkney, so port striking isn't possible.

Use Tac Bombers - not Naval Bombers. Oh, and use Strat Bombers to knock out the AA first, then hammer the port.

 

[SigurdStormhand]

The way I see it, this would cause even more problems. When the AI is invading, removing the supremacy requirement would cause the AI to continuously launch suicide naval invasions and lose hundreds of divisions to player submarines. On the other hand, this would make it far, far too easy for the player to naval invade AI-controlled countries, because they don't know how to use convoy raiding fleets effectively.

It would also tie the player down with large numbers of divisions covering their home ports just to burn off the AI's manpower.

 

[Reman]

Use Tac Bombers - not Naval Bombers. Oh, and use Strat Bombers to knock out the AA first, then hammer the port.

At the point where Tacs+Strats are the only option left to destroy enemy ships, it's far cheaper to just overwhelm enemy superiority by spamming crappy submarines.

 

[Harin]

The way I see it, this would cause even more problems. When the AI is invading, removing the supremacy requirement would cause the AI to continuously launch suicide naval invasions and lose hundreds of divisions to player submarines. On the other hand, this would make it far, far too easy for the player to naval invade AI-controlled countries, because they don't know how to use convoy raiding fleets effectively.

I believe you are correct if the naval supremacy rule is removed without updating the AI.

Hopefully though, it would not be that difficult to have the AI notice all the naval power projected in the sea zones and make a calculated decision on whether to launch a naval invasion through them. It is what the human player should be doing, instead of having the naval supremacy mechanic remove the decision from the game. Naval supremacy is not a rule, it is earned through action. Ask the captain of the Bismark.

Supremacy is challenged, always.

The supremacy of Allied air power was challenged to the very end. The supremacy of the Tigers were challenged. The supremacy of occupiers were challenged with resistance. The supremacy of the U-boat in the early years did not go unchallenged. And lets not forget that Hitler thought is wise to challenge the obvious industrial supremacy of the UK, Commonwealth, USSR, and the USA all at the same time.

If we took the mechanics and reasoning of naval supremacy and stretched it to the behavior of AI leaders, then the AI Hitler would reasonably decide that there should be no WW2.

 

[Xeonic]

The way I see it, this would cause even more problems. When the AI is invading, removing the supremacy requirement would cause the AI to continuously launch suicide naval invasions and lose hundreds of divisions to player submarines. On the other hand, this would make it far, far too easy for the player to naval invade AI-controlled countries, because they don't know how to use convoy raiding fleets effectively.

I mean the easy fix is removing the requirement for players, but not for AI. If naval supremacy cannot be challenged due to the AI only using strike fleets and no patrols, the system as is doesn't work, since it leads to unremovable blocks to naval invasions.

EDIT:
Added a suggestion in the suggestion forum for now, let's see if it gains any traction.

https://forum.paradoxplaza.com/foru...macy-requirement-for-naval-invasions.1354862/

 

[elektrizikekswerk]

The only good solution to this problem in my opinion is to remove the arbitrary 50% naval supremacy requirement to launch naval invasions. Let us determine ourselves if we want to launch stupid invasions and get caught by a huge home fleet.

This is by far the least arbitrary solution. The whole thing stops being a problem if it's fixed this way.

The problem with that would be that it works both ways. If tha AI was able to launch naval invsaions "at will" it would most probably do so. Unless, of course, it learned somehow when it's worth the risk and when not. And risk assessment is not really something in which the AI is great at all...

 

[Xeonic]

The idea is that strike forces react to naval invasions if they happen, and the 50% limit can still be imposed through other means than a hard stop for any invasion plans below 50% supremacy.

 

[TheMeInTeam]

I believe you are correct if the naval supremacy rule is removed without updating the AI.

He isn't correct. See spoiler images below for examples of naval invasions by minors against the two nations that should be most shielded by their navy in typical games. In neither instance was their navy actually sunk first. It wouldn't alter the game as much as asserted; those are consistently repeatable invasions.

The way I see it, this would cause even more problems. When the AI is invading, removing the supremacy requirement would cause the AI to continuously launch suicide naval invasions and lose hundreds of divisions to player submarines. On the other hand, this would make it far, far too easy for the player to naval invade AI-controlled countries, because they don't know how to use convoy raiding fleets effectively.

Neither of these are particularly relevant in practice.

AI naval invasions almost always end in a relatively swift wipe of AI armies *right now*. Losing 3-10 divisions instantly to subs vs landing and losing them out of supply isn't meaningful. AI naval invasion logic could use a fix as an independent goal from fixing the superiority rule, regardless of what the fixed rule looks like.

For the past 2+ years IRL, it has been possible to do stuff like this:

Spoiler

or this

So claiming "it's easy to naval invade the AI" is not a convincing refutation for making naval invasion rules more sensible. You *can* get punished from the AI killing your supply convoys. It actually happens, sometimes. The solution to this is making that more likely to happen, not fabricating magic fantasy blocks that prevent the player even launching the invasion. The latter is completely arbitrary and conflicts with the fact that you can paradrop into the same provinces you aren't allowed to naval invade, and that the player is generally allowed to take suicidal actions if desired.

The problem with that would be that it works both ways. If tha AI was able to launch naval invsaions "at will" it would most probably do so. Unless, of course, it learned somehow when it's worth the risk and when not. And risk assessment is not really something in which the AI is great at all...

As someone who often plays minors with terrible navies, this wouldn't change my experience much at all. I also fail to see how it's particularly different from USSR racking up 10M casualties w/o Barbarossa happening because the AI has a hard-on for attacking while underequipped. Let's fix the rules first, then worry about how the AI interacts with the rules once the rules are somewhat decent. The fact that HOI 4 doesn't do this is a source of a lot of its problems.

 

[elektrizikekswerk]

As someone who often plays minors with terrible navies, this wouldn't change my experience much at all. I also fail to see how it's particularly different from USSR racking up 10M casualties w/o Barbarossa happening because the AI has a hard-on for attacking while underequipped. Let's fix the rules first, then worry about how the AI interacts with the rules once the rules are somewhat decent. The fact that HOI 4 doesn't do this is a source of a lot of its problems.

No, its' exactly the other way around: First step would be to improve the mentioned risk assessment. The problem with the underequipped suicidal attacks is a result of bad risk assessment. Opening up another way to kill yourself easily would make the whole thing even worse.
Once the AI's risk assessment is in order you can start to relax the rules.

Assuming, of course, this is a rule which needs to be relaxed. Which it isn't. The problem here is the Strike Force mechanic which gets accidently exploited by the AI. I say "accidently" because I don't think that this is intentional. Relaxing the rule is like taking pain killers without caring for the cause of the pain.

 

[TheMeInTeam]

No, its' exactly the other way around: First step would be to improve the mentioned risk assessment. The problem with the underequipped suicidal attacks is a result of bad risk assessment. Opening up another way to kill yourself easily would make the whole thing even worse.

If the rules don't work, making an AI that operates in the non-functional rules is senseless. The rules need to work first.

Once the AI's risk assessment is in order you can start to relax the rules.

No. The game's design comes first, then the agents that operate in the game can play it. It's extremely backwards to leave the rules gutted and flounder with an ever-struggling AI. This nonsense is why the devs still can't manage a competent or even self-consistent algorithm for war score. In fact the naval invasion restriction is self-inconsistent in its own right. If the stated logic for preventing naval invasion is risk to invading divisions and concern over supply, it should not be possible to paradrop at all, and the game should more consistently block units moving to poorly supplied fronts.

Assuming, of course, this is a rule which needs to be relaxed. Which it isn't.

It's a self-inconsistent implementation without coherent justification. I don't see how that isn't grounds for improvement. The rule is arbitrary regardless of "strike force".

 

[elektrizikekswerk]

If the rules don't work, making an AI that operates in the non-functional rules is senseless. The rules need to work first.

The point is, the AI already works with the "non-functional" rules - no work needed

No. The game's design comes first, then the agents that operate in the game can play it. It's extremely backwards to leave the rules gutted and flounder with an ever-struggling AI. This nonsense is why the devs still can't manage a competent or even self-consistent algorithm for war score. In fact the naval invasion restriction is self-inconsistent in its own right. If the stated logic for preventing naval invasion is risk to invading divisions and concern over supply, it should not be possible to paradrop at all, and the game should more consistently block units moving to poorly supplied fronts.

While I agree in theory here I disagree in practical terms.
The design/rules can be as elaborated as you want. If the agent (AI) cannot handle them the overall game experience is worse than with "non-functional" rules. This is almost or completely a non-issue in MP games, true, but in SP it matters if the AI can or cannot handle the rules.

 

[TheMeInTeam]

The point is, the AI already works with the "non-functional" rules - no work needed

You're mistaking HOI 4 for something else. I already demonstrated that the reasons the change would be "harmful to the AI" are actually present in HOI 4 right now. The AI suiciding millions of manpower in naval invasions is something that happens already for example and can be freely baited. Using that as a refutation to the rule change is non-sequitur.

While I agree in theory here I disagree in practical terms.
The design/rules can be as elaborated as you want. If the agent (AI) cannot handle them the overall game experience is worse than with "non-functional" rules. This is almost or completely a non-issue in MP games, true, but in SP it matters if the AI can or cannot handle the rules.

It matters, but it is not valid reasoning to make incoherent or internally inconsistent rules. This consideration is more in line with making a game less complicated so that making an AI for it is within the abilities of the programmers. That's still not a basis for arbitrary rules.

 

[elektrizikekswerk]

I already demonstrated that the reasons the change would be "harmful to the AI" are actually present in HOI 4 right now.

That bad things are already present doesn't justify making it worse.

It matters, but it is not valid reasoning to make incoherent or internally inconsistent rules. This consideration is more in line with making a game less complicated so that making an AI for it is within the abilities of the programmers. That's still not a basis for arbitrary rules.

The rule is neither incoherent nor internally inconsistent, though. 50% are arbitrary, true, but 50% are 50%.
It's even absolutely logical that strike forces in port contribute to the 50% - "Fleet in Being" is/was a thing IRL. The only problem is that you can set up Strike Forces without (sufficient) Patrols. THAT rule had to be changed. You shouldn't be able to create a "Strike Force" without patrols and the STrike Force should be disabled automatically if all patrols are sunk/destroyed/in port for repair.

 

[TheMeInTeam]

That bad things are already present doesn't justify making it worse.

Could you highlight how, precisely, it's actually worse?

The rule is neither incoherent nor internally inconsistent, though. 50% are arbitrary, true, but 50% are 50%.

It is both.

You can do all sorts of suicidal actions freely, but not this "suicidal" action. Blocking naval invasions and to a lesser extend paradrops are a special cases. Even in this special case, the game prevents naval invasions in situations where the convoys would not be attacked and allows naval invasions in situations where the convoys are extremely unlikely to make it, which is necessarily internally inconsistent for any reasoning presented for blocking naval invasions so far.

"Fleet in Being" is/was a thing IRL. The only problem is that you can set up Strike Forces without (sufficient) Patrols. THAT rule had to be changed. You shouldn't be able to create a "Strike Force" without patrols and the STrike Force should be disabled automatically if all patrols are sunk/destroyed/in port for repair.

"IRL" is a very poor rationale in this context. IRL, nothing technically prevented Germany from attempting the invasion. It was a bad idea, just like artillery only was a bad idea, so they didn't do it.

I could get behind strike forces having a scaling superiority penalty based on strength of patrol fleet detection, going to 0 with no patrol fleet presence. But this is exactly the kind of complicating factor you pointed out an AI would struggle with handling.

 

[elektrizikekswerk]

Could you highlight how, precisely, it's actually worse?

I already did:

The problem with the underequipped suicidal attacks is a result of bad risk assessment. Opening up another way to kill yourself easily would make the whole thing even worse.

If you need an explanation of why for an "AI"(!) having <n+1> ways of more or less erratically making bad decisions is worse than having <n> ways of more or less erratically making bad decisions I cannot help you.

It is both.

You can do all sorts of suicidal actions freely, but not this "suicidal" action. Blocking naval invasions and to a lesser extend paradrops are a special cases. Even in this special case, the game prevents naval invasions in situations where the convoys would not be attacked and allows naval invasions in situations where the convoys are extremely unlikely to make it, which is necessarily internally inconsistent for any reasoning presented for blocking naval invasions so far.

Ah, when you said "incoherent or internally inconsistent rules" you actually meant "incoherent or internally inconsistent with other rules". Yes, in that I agree. Wasn't the point, though.

But this is exactly the kind of complicating factor you pointed out an AI would struggle with handling.

I don't disagree with that. As you might have noticed in my opinion a rule change should imply that the AI can handle it. That's why I am advocating for FIRST giving the AI some proper risk assessment capability of which it would benefit already now in all those scenarios mentioned above by you and only THEN change the rules. So changing the requirements of creating Strike Forces (or their contribution to naval superiority) should of course only be made if the AI can handle it.

 

[TheMeInTeam]

If you need an explanation of why for an "AI"(!) having <n+1> ways of more or less erratically making bad decisions is worse than having <n> ways of more or less erratically making bad decisions I cannot help you.

It's making literally the same decision though (suicide invasion that loses all of the troops). Evaluation is identical too, player can even activate/deactivate naval missions to take advantage of the poor threat assessment right now, if you want to be attentive enough. But you don't need that because the AI will just land and lose all its units over and over.

Ah, when you said "incoherent or internally inconsistent rules" you actually meant "incoherent or internally inconsistent with other rules". Yes, in that I agree. Wasn't the point, though.

I meant within the context of the game as a whole, yes. Naval invasions are also internally inconsistent with themselves (example about being able to launch invasions that are trivially intercepted, but blocked when they would not be). If the ostensible reason for stopping the invasion is that it's too dangerous/suicidal, this is the opposite of the behavior we should expect.

I don't disagree with that. As you might have noticed in my opinion a rule change should imply that the AI can handle it. That's why I am advocating for FIRST giving the AI some proper risk assessment capability of which it would benefit already now in all those scenarios mentioned above by you and only THEN change the rules.

I've already demonstrated that the risk assessment failure fully impacts the AI right now. The risk assessment should be improved, but lack of it does not meaningfully alter the gameplay implications of having sensible naval invasion rules.