HOI4 Oak 1.4.2 BETA patch [checksum: a48d]

  • 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.
Status
Not open for further replies.
Unfortunately in almost every case this is due to the straitjacket of the focus code. It over-writes everything without regard to the situation on the ground. This might not be able to be changed without a re-write of the engine or diplomatic code. Probably going to have to go focus by focus.

Focuses should just give you wargoals and not already declare war for you. In this way a generic code based sanity check can do its job. Alternatively the AI should understand the war decision implied by a focus, apply its sanity check and then bypass the focus, if the result is negative. Not exactly a rewrite of the engine, I would think.
 
Last edited:
Focuses should just give you wargoals and not already declare war for you. In this way a generic code based sanity check can do its job. Alternatively the AI should understand the war decision implied by a focus, apply its sanity check and then bypass the focus, if the result is negative. Not exactly a rewrite of the engine, I would think.
...they do? Have you been playing Vanilla+? Vanilla foci only ever give wargoals.
 
Hmm, don't think that will stop Around Maginot. IIRC the focus for going after Denmark is after Around Maginot in the ai focus tree. The ai will still fire AM and declare on BeNeLux which will cause the AI front code to pull troops away from the East.

If you remember, Around Maginot was originally 70 days like all other focuses. @podcat changed it to 28 days and moved it in front of Weserburung to avoid the Germans getting locked in Denmark and dying on the Maginot. Now PDS has added new code that improves the Polish ability to defend itself, and for the German ai to not stupidly attack fortresses. In the games I have played in 1.4.... Poland usually either dies or is about to between 30-40 days as long as Germany isn't distracted. That is why I just went KISS and moved the time to complete AM to 42 in my home mod.

As far as I understand it, it's an OR, so if GER is either at war with POL and it exists, or at war with DEN and it exists, it shouldn't take the focus - so DEN in there shouldn't stop being at war with POL blocking the AI from taking it. We've tinkered with some of the orders of our focuses as well, which may well be why DEN is there (so it's possible to have Weserebung pre-Maginot, but not have Maginot fire until Germany has Copenhagen).

Nothing wrong with a 42 days thing either though - was just throwing ideas around in case it helped :).

The real solution to this is always going to have been to get the AI to consider its war performance when thinking about starting a new war. There are lots of situations where the AI starts wars it shouldn't, and joins factions that it shouldn't - it would be good to get them all at once rather than fixing each one with national focuses.

Aye, long-term I'm sure this is what the devs are working towards. Probably a devilishly tricky thing though - I expect a bit of nudging one way or the other through things like NFs will help the situation (particularly from a 'historical game' perspective).

Focuses should just give you wargoals and not already declare war for you. In this way a generic code based sanity check can do its job. Alternatively the AI should understand the war decision implied by a focus, apply its sanity check and then bypass the focus, if the result is negative. Not exactly a rewrite of the engine, I would think.

...they do? Have you been playing Vanilla+? Vanilla foci only ever give wargoals.

At the very least, "Danzig or War" gives an "at war" outcome if Poland doesn't give up Danzig, but in the vast majority of cases it's just a war goal, and the devs have noted in the past (iirc, but I'm fairly sure about this), that the AI has it's own checks before going to war, even if it has a wargoal (although there's obviously room for improvement - noting it's a very complicated thing to get right, and all other Paradox games (and pretty much every 'free form' strategy game ever) have issues with the AI knowing when and were to declare, and not declare, war).
 
In my runs Germany usually brute forces through the Maginot during its offensives and the French has a tendency to surrender parts of the Maginot through withdrawing.

I had same thing. I played Germany and the french did not defend Maginot a few months after the war began and my troops took it. it was no brute force. I didn´t even attack Maginot. and I declared war on Belgium and Netherlands.

Bug or work as designed? I guess it is a bug.

and Italy is still sending things through the english channel. I have a border to italy. why don´t they bring the stuff from Africa to Italy and from there to Germany by land transport?
 
Will there be any fix to the overly aggressive Chinese? They will just attack over and over until they're at less than 5% equipment. It's almost impossible to lose to as the Japanese, in fact the only thing you need to do is place soldiers on their border and declare war and the Chinese will just do the work for you by assisted suicide.
 
Unfortunately in almost every case this is due to the straitjacket of the focus code. It over-writes everything without regard to the situation on the ground. This might not be able to be changed without a re-write of the engine or diplomatic code. Probably going to have to go focus by focus.

Yes, the game has no strategic AI whatsoever. There are no victory conditions the AI could pursue.
 
Hmm, don't think that will stop Around Maginot. IIRC the focus for going after Denmark is after Around Maginot in the ai focus tree. The ai will still fire AM and declare on BeNeLux which will cause the AI front code to pull troops away from the East.

If you remember, Around Maginot was originally 70 days like all other focuses. @podcat changed it to 28 days and moved it in front of Weserburung to avoid the Germans getting locked in Denmark and dying on the Maginot. Now PDS has added new code that improves the Polish ability to defend itself, and for the German ai to not stupidly attack fortresses. In the games I have played in 1.4.... Poland usually either dies or is about to between 30-40 days as long as Germany isn't distracted. That is why I just went KISS and moved the time to complete AM to 42 in my home mod.

I never have played historical AI but seems logical to me that Around Maginot is placed after Weserübung and gets completed in 28 days since the real Fall Gelb took place also around a month after Weserübung.

Biggest problem with AM in non-historical is that it is limited by having war with a neighbour (good idea) but usually war is declared after the next focus is chosen so most of the time it does not work. I have modded an additional requirement in so that the AI can't take it as long as Danzig is controlled by Poland. This way Germany will not go Around Maginot right after Danzig Or War, maybe only if Danzig is ceded (but i think ownership change also takes place slower than focus choice).
 
Last edited:
Will there be any fix to the overly aggressive Chinese? They will just attack over and over until they're at less than 5% equipment. It's almost impossible to lose to as the Japanese, in fact the only thing you need to do is place soldiers on their border and declare war and the Chinese will just do the work for you by assisted suicide.

This is not China-specific. You see it a lot with many different countries.
 
It seems like the AI is trying to launch holding attacks and conduct attritional war with that crazy aggressive attack stuff, but they're not stopping for entrenchment or planning bonus so it doesn't work at all.

When you're weaker than your enemy you do kind of have to attack even when the battle is costing you more than it costs your opponent so that they can't move to an area where you've got overwhelming superiority in numbers. And it can also be a good idea to attack even into a losing situation if your kit is cheaper and the enemy is struggling to replace theirs in the local area. But before you do it you've got to stack every bonus you can. And you've got to be able to put the costs into the context of the impact on both the local area and the wider war. A stupid attack in one front might enable a successful attack in another if it draws in enemy units. Intuitive for HUMANS, AI still needs help with that stuff.
 
I had same thing. I played Germany and the french did not defend Maginot a few months after the war began and my troops took it. it was no brute force. I didn´t even attack Maginot. and I declared war on Belgium and Netherlands.

Bug or work as designed? I guess it is a bug.

and Italy is still sending things through the english channel. I have a border to italy. why don´t they bring the stuff from Africa to Italy and from there to Germany by land transport?

Agree for Italy to sent units via the English Chanel instead of over the border to Germany via Austria is just an example of Stupid AI and poor game development...Did the programmers & play testers miss this?
 
Hopefully that shorter sea routes fixes most of the stupid trips AI takes. For exampleItalian troops jumping to transports from France, not Italy and sea transports from Japan to Mongolia won't go to Korea or Siam, but around Singapore to Burma.
 
Also what occurs under the hood when you save and load? I have noticed that the AI sometimes loses their sh** after I load up a previous game. Most recently I was playing Romania, and after loading, I saw my ally France move 100% of their troops from the alps and ship them to turkey and yugo. Meanwhile, the German and Italian soldiers stormed into France and ended their game. . .

On another note, the AI is too aggressive when attacking mountains and fortified positions. I had thought it was just Germany and Italy, but now playing as other nations I see this is a common, fatal flaw. No war leader would attack continuously until his forces were down to 5% strength-- under any circumstance. Meanwhile, the actual circumstance is often an offensive taking 50 casualties for every casualty inflicted.

I have seen France take 150,000 casualties a month attacking Spain through the Pyrenees, with no attack having more than a 7 score. Meanwhile, an amphibious assault could have worked. Or, you know, they could always make the decision to not attack.
 
The AI is a bit too aggressive. Playing as Japan it is way too easy to bleed AI China and AI SU by digging in behind rivers and mountains. I also saw the Germans break through the Maginot line and take Paris while still fighting in Belgium. In that case they won, but still doesn't look good.

Edit: the game is still a lot better than before.
 
So when will this hot fix go live? Really can't wait for that!

In a few minutes....

The AI is a bit too aggressive. Playing as Japan it is way too easy to bleed AI China and AI SU by digging in behind rivers and mountains. I also saw the Germans break through the Maginot line and take Paris while still fighting in Belgium. In that case they won, but still doesn't look good.

Edit: the game is still a lot better than before.

We are looking into AI defending for 1.5 actually. Its an area where Ai is weak. its especially visible for China.
 
closing as patch is going up live now
 
Status
Not open for further replies.