-
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.
Sorry, I don't get the image, everything in it looks fine to me.
I would certainly recommend toning down the infrastructure builds, as the USSR I'd find it grossly inefficient and wasteful to speed up infra by 3 on a 4IC province, let alone one in the desert (Ashgabat), a 5IC one in the mountain (Yerevan), or a province that is an easy target for the Germans (Krivoy Rog). This just makes mention of some of the long list of infra builds in the USSR that IMO should be edited or scrapped - there are many more for other nations too. Luckily I found some great people to play with online so I don't have to deal with the stupidity of the AI as much*, but most AoDers do use single-player.
*But still, the France AI could do with some editing! The Axis minors too! I'm happy to do such editing if Pang also feels the need to edit them. I just find it silly that France doesn't just spam infantry-art or something similar. I've had it where France had 600 unused manpower by late-1939, those 40-50 divisions could make the world of difference against Germany. There's one mod where France already had 1941 infantry by mid-1940, that would be a dream to see on 'vanilla' AoD! But I know, I know, Pang prefers more 'historical' to 'practical' AI - which I could understand if players operated the same way, but 9 times out of 10 they don't! Everyone knows when this or that event will trigger, they know when war will start and they plan accordingly. I know the AI can't really plan anything in the first place, but I'd rather have it act as 'humanlike' as possible. Rant over.
Last edited:
The "ineffeciency" for Soviet union is somewhat intended. That helps soviet union not being too strong in 1941 and not being too weak after 1945. If things work out as intended most Infra will be 200% a few months before there war. The production lines allowed by that should have builded 1-2 units when the war starts. This maxes out the daily production output when barbarossa starts. The later is an approximation of what happened IRL.
French-AI & infatry
Maybe France could have switch to make more infantry if N.Spain is a Axis member? If the German-human player invades Luxenburg before Danzig or war it could also be switch for more land units.
Maybe France could have switch to make more infantry if N.Spain is a Axis member? If the German-human player invades Luxenburg before Danzig or war it could also be switch for more land units.
Franch AI must utilize higher portions of its manpower in any case. No conditions should be attached to this.
I completely disagree with that notion. Surely the AI needs improving, not deliberate weakening. I would rather operate on the notion of maximising the AI's abilities (without the need for further 'cheats' such as free research etc). If the USSR AI kepts trumping the German AI, then I would rather change the structure of the game - either editing the IC, manpower, resources or whatever is needed. But, AFAIK, the AI massively fluctuates from game-to-game anyway, ergo any idea that one AI is better then another is hard to identify in the first place and so edits deliberately weakening an AI are un-necessary.
But AoD isn't RL. A German player would've probably maximised the size of their army by June 1941, meaning they can easily defeat the USSR in the same year if has just built 1-2 units. I would rather face a clever AI, then a historical AI that is effectively blind to the changing situation (i.e. what if Barbarossa happens in 1940? What if Germany invades the UK? What if Japan invades the USSR? etc)
IMO France shouldn't need to switch to making more infantry, it should just do so. What other units does a France AI need anyway? The AI is best at land combat (I know it's still awful, but it's much better compared to it's air and naval), and France only really requires infantry anyway. Once it runs out of manpower, it should focus on upgrades (by rushing 1941 infantry and 1940 artillery) and building land forts on its border with Belgium.
EDIT: Just noticed Pang's comment below Pioniere's, I completely agree!
Last edited:
Some interceptors would be nice. And for historical reasons french A should build some armour, even if it is some cheap HA brigades for Infantry. AI pays no additional icd building divisions with brigades attached compared to building unbrigaded divisions.
Let me clarify: I mean 1-2 units per each production line. So there are 200-400 ic spend on production lines that have achieved maximum gearing bonus by 22/6/1941. I am talking about a bulk out of army unexceeded by any other AI and probably only exceeded by few human players. Therefore...
... i feel the term "weaking" somewhat unproper. It is set of prioritites not everyone might agree an. But the icd saving created by having a comparably small army before 1941 creates only slight losses in army size by 22/6/1941, but they creaty significantly increases in available tc in 1941 and later. It is possible for AI to have a greater army size and this might hamper german progress in barbarosssa. But due to a chronical tc overload soviet union might not be the real winner either. The real winner of a less risky soviet strategy as your postings might suggest are the not (axis+komintern), namely allies and a few less important parts of the world.
A realistic compensation for a stronger SOV AI controlled country at the start of Barbarossa would be to introduce a time delay before the relocated SOV factories come into production. I haven't played 1.09 SP, or SOV SP for some time, and don't know if this change has already been included.
It has not been included*. This would make sense ftrom realism point of view, but it would have the effect to hamper soviet production when it is needed most. So from a balance point of view i totally disagree with it.
*: The current provincial effeciency is not immidiatly changed. The maximum provincial effeciency is changed according to the new concentration bonus, so there already is a small loss.
*: The current provincial effeciency is not immidiatly changed. The maximum provincial effeciency is changed according to the new concentration bonus, so there already is a small loss.
Influence pupet nations
Maybe it should be easier to influence your puppet nations: like if you was paternal N.Spain and have Portugal as a puppet, but have gone fascist gone or social-conservative the puppets government will change if you influence it.
Maybe it should be easier to influence your puppet nations: like if you was paternal N.Spain and have Portugal as a puppet, but have gone fascist gone or social-conservative the puppets government will change if you influence it.
Last edited:
Agreed, it's infra building in any place bar Paris is effective a bonus to Germany more then anyone else. AS to building interceptors, I'd see that as a waste as the AI is awful at air combat and France shouldn't waste research time on air doctrines or techs.
I highly doubt any player in their right mind would stick to the infra builds you have for the USSR (and most other countries). Much of it is a great waste that, even for the USSR, I highly doubt the AI can afford to do, especially when faced with a human opponent. The infra builds in the Ukraine would perhaps be acceptable if they ended in 1938 (but preferably didn't exist at all), rather then have the AI build them upto the maximum possible before thankful German soldiers capture them.
If the AI Red Army does not have, at least 300 divisions (but preferably 400-500) by June 1941, then the human German player will most likely achieve Bitter Peace by 1941, if not the spring of 1942.
The UK is also significantly disadvantaged by this over-focus on infrastructure builds - if you want it to build something, make it IC in the British isles until 1938/39, not infra all over it's empire. I doubt the UK AI can even properly defend India and Malaysia.
Last edited:
Nor would i. Those infra builds are for AI. The rules for AI and the behavoir of AI is different from human.
AI is optimized for something in the scale of 300 so that german AI can beat soviet AI. I suppose those numbers scale up if you increase the difficulty setting.
I completely agree, no human would player like the AI as the AI is, for lack of a better word, retarded. Ergo it needs all the help it can get, not these arbitrary limitations.
Is that so the German AI can beat the Soviet AI severely in 1941 or do you mean so it can achieve Bitter Peace?
Either way, I don't understand the logic. Arsenal of Democracy, like all computer games, is played and enjoyed by humans (not the AI). I'm guessing (and hoping) most players prefer a challenging AI over a retarded one - IMO the AI should act as human-like as possible, we should be maximising its chances of winning, not hampering it for the sake of another AI. Unless the USSR AI is defeating the Germany AI in 1941/42, then I see virtually no reason why it should waste its IC on superfluous infrastructure builds - and if it is doing so, then perhaps Germany's infra builds need to be edited, in addition to the AI itself. I am happy to do such editing if it stands a reasonable chance of being included in the proposed 1.09 edits.
Is is so that chances are high that either soviet union or germany will achieve a sort of victory that makes them a challange to beat for a human player of USA etc..
If i believed they were superfluous i would (have) remove(d) them. Some choices may be ambivalent, true. But the intensity that you fight those infrabuild makes me suspect that you donnot sufficiently understand the underlying mechanics of AI building Infrastructure. If you ask me they are the least of the existing problems and changing them comes with the risk to change things for the worse.
Pray tell, is it because the AI doesn't factor in the issues of TC (and manpower) when building an army? Or because the AI struggles to manually put infrastructure builds into production, or that it cannot increase the speed of production itself? Or...
Of course there are much bigger issues, but most of them are hardcoded. Or are there still significant softcoded issues?
Last edited:
All you list up and some more. Also maintainance and upgrades are rather relevant from the simple icd perspective.
Much is related to the combination of both.
That's true, but maybe a more effective use of ICD, at least compared to building infra all over the place, would building land forts in Moscow, and perhaps even a defensive line somewhere? But then again, I guess the AI wouldn't prioritise defending those land forts other other provinces, would it? But still, land forts in Moscow, Leningrad, Sverdlovsk might be ideal.
Can't we make trigger events, so that the AI only upgrades, say, 1918 infantry, once the 1939 infantry tech is researched?
That's good, and can't there be switch events for manpower, say if a major country has less then 100 manpower in reserve, they should stop building units?
Regarding those Soviet AI events, did you make the proviso that if at war early, they start upgrading immediately, regardless if tech?
Regarding those Soviet AI events, did you make the proviso that if at war early, they start upgrading immediately, regardless if tech?
I donnot think that this is possible, but there are some AI events to change building schemes from high manpower/icd to low manpower/icd. The tresholds however are more like 1000 manpower, depening on army size. AI does not delete serials from production, but long serials are needed for an efficient build up. There is quite a tradeoff between those 2 objectives.
Yes.