Combat mechanics seriously flawed
- Thread starter Olle8
- Start date
-
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.
4k bonus isn't achievable except in some pretty extreme situations. 2k is easily achievable though
Stats need diminishing returns, in a few of my games I've built some cheesy artillery divisions, where the inf to arty ratio was stupid, it didn't matter that org was low because the soft attack cut through anything.
HoI4's freedoms are the cause of some of her biggest problems, the AI can't handle the sandboxyness, and the player can easily cheese.
Alternatively historical only divisions could produce more "sane" battles but the underlying systems in the game haven't been built around it.
In a nutshell, difficulty sliders aside, HoI4 is just too easy, and too easy to break within it's own mechanics.
HoI4's freedoms are the cause of some of her biggest problems, the AI can't handle the sandboxyness, and the player can easily cheese.
Alternatively historical only divisions could produce more "sane" battles but the underlying systems in the game haven't been built around it.
In a nutshell, difficulty sliders aside, HoI4 is just too easy, and too easy to break within it's own mechanics.
If you know a more efficient infantry division than 7 infantry and 2 artillery I'd sure like to hear about it
It's interesting to me to look at how the Darkest Hour devs looked at some of the same problems: https://forum.paradoxplaza.com/foru...neak-peak-1-03-combat-rework-and-more.596620/
They implemented a diminishing returns mechanic, where past a certain value, additional damage dropped off because only so much concentration of firepower was possible:
The saturation of attack also meant that there was a certain minimum butcher's bill to pay when wiping a defender's ORG. If you can't wipe a division in an hour or two, then they actually get some time to shoot back.
They implemented a doctrine-based "dispersion" mechanic that reduces damage (to offset increases in firepower):
Now, the HoI4 combat engine is different, but I think these two ideas could be developed.
They implemented a diminishing returns mechanic, where past a certain value, additional damage dropped off because only so much concentration of firepower was possible:
The saturation of attack also meant that there was a certain minimum butcher's bill to pay when wiping a defender's ORG. If you can't wipe a division in an hour or two, then they actually get some time to shoot back.
They implemented a doctrine-based "dispersion" mechanic that reduces damage (to offset increases in firepower):
Now, the HoI4 combat engine is different, but I think these two ideas could be developed.
That's all well and good but it doesn't get to the root of the issue at hand which is casualties taken on the offensive. To fix this central issue, negative modifiers need to be toyed with.
To Make the game more fun though, we definitely out to rebalance all modifiers eventually.
To Make the game more fun though, we definitely out to rebalance all modifiers eventually.
High casualties when on the offensive with infantry is historical though and an aspect I at least want to keep in the game if possible.
The problem is the AI not realizing the dangers of front-line infantry Banzai charges and is committing to them like it is Okinawa 1945 all the time constantly, not the mechanics themself.
I do agree that 100:0 casualties doesn't make sense, but if you charge well prepared enemy lines with just infantry I would expect to lose at least 10:1 or 5:1 casualties if not even more.
Very interesting to see how DH approached it. Does anybody remember how it played with diminishing returns?
My thoughts more widely is that 1:1 ratio, then Defence, on the whole, really should be stringer than attack. Attack success should come from operational concentration of force.
My thoughts more widely is that 1:1 ratio, then Defence, on the whole, really should be stringer than attack. Attack success should come from operational concentration of force.
I wonder if you could tweak the mechanics could be adjusted to represent more reasonable behavior and leave modelling the Banzai charges to special rules. Something like slowing down or automatic early cancelling of attacks that are going terribly, representing the degree of willingness to throw troops and equipment away for no gain.
Alex, yes the AI attacks are an issue. But it really looks to me like the battle stats are way off too. I check the casualty window, and I often get crazy 0 results when I'm attacking. Most other times I get absurdly low casualties.
The strength of defense is already represented by the defense and breakthrough modifiers of most units. The problem right now is the AI is incapable of playing the game it was meant to be played. With the modifiers they wind up getting come wartime (- supply, - river crossing, - missing equipment) The AI cases zero damage to defending units. And that's at the start of the war! As time goes on, the problems become more and more drastic. I hope confidence in the development team though. Things get ironed put peices by peice. Hopefully by 1.5 the AI will be able to keep its divisions in good condition.
We need an expansion dedicated to ironing out the land war though
I never said the battle stats were not way off, they indeed are. I agree fully.
What I said was that high casualties when attacking with infantry is historical.
Oh yes understood. And agreed.
I'm also mindful that many successful attacks were actually defensive tactically, attack operationally. If memory serves, eg El Alamein?
And of course armour isn't immune to it, Mortain.
Just throw some figures in here:
During the Battle of the Bulge, the Germans suffered more than 100,000 casualties; the Americans approximately 81,000.
Total Allied losses NWE 44-45:
US 4500 M4 & 1500 Stuart (25%).
UK 4000 Medium & 440 Stuart (10%)
Lifted this:
US Division Theater Casualties
3rd Infantry Division Mediterranean & European 25,977
9th Infantry Division Mediterranean & European 23,277
4th Infantry Division European 22,660
45th Infantry Division Mediterranean & European 20,993
1st Infantry Division Mediterranean & European 20,659
29th Infantry Division European 20,620
36th Infantry Division Mediterranean & European 19,466
90th Infantry Division European 19,200
30th Infantry Division European 18,446
80th Infantry Division European 17,087
Source: Army Battle Casualties and Nonbattle Deaths in World War II, Final Report, 1 December 1941 - 31 December 1946.
And US were winning, had mass fire, air supremacy, best supply, medivac, etc. We need this sort of casualties in HoI.
High casualties on the attack were the norm, unless supported by artillery, armor, air power, and/or other force multipliers, and unless done with a strong numerical advantage. If you're proportionally taking massive casualties despite those advantages (or the AI is doing so), then something isn't working as intended. If the AI is attacking at better than 3:1 odds, with support, then casualties shouldn't be all that one sided without other factors heavily skewing the results.
As far as I can see, the combat mechanics in HOI3 worked fairly well, at least after they fixed the Defensiveness and Toughness issue in patches for the first expansion (someone mixed up their percentages (5%) and decimal fractions (0.05), causing the game to give units only small fractions of a percent (0.05%) as a bonus). You could use combinations of terrain, leader bonuses, special forces, air support, armor breakthroughs, and other advantages to get heavily one-sided results under the right circumstances, but you still took some losses even in the most extreme situations. Inflicting 100:1 casualties or greater falls into the realm of fantasy, not counting the encirclements and armor overruns that force a surrender and remove units completely without actually killing them.
As far as I can see, the combat mechanics in HOI3 worked fairly well, at least after they fixed the Defensiveness and Toughness issue in patches for the first expansion (someone mixed up their percentages (5%) and decimal fractions (0.05), causing the game to give units only small fractions of a percent (0.05%) as a bonus). You could use combinations of terrain, leader bonuses, special forces, air support, armor breakthroughs, and other advantages to get heavily one-sided results under the right circumstances, but you still took some losses even in the most extreme situations. Inflicting 100:1 casualties or greater falls into the realm of fantasy, not counting the encirclements and armor overruns that force a surrender and remove units completely without actually killing them.
I think that the 4 main reasons for this are:
- HoI3 AI was not quite as suicidal.
- HoI3s division design was much more restricted with 3-4 brigade units being the only option ( 5 was possible with latewar doctrine ). Compare that to the HoI divisions that can differ from 3 battalions on some AI units to 30 battalions (including support).
- HoI3s unit stats (mainly soft attack) did not spiral out of control like they do in HoI4. Early game HoI3 Soft attack would be ~2 for infantry and highest possible latewar ~8. In HoI4 we got a range from 3 to 140+ soft attack from WW1 infantry to 1943 Heavy SPART.
- HoI3s bonus to surpassing enemy defense/breakthrough was x2 damage ( or lower IIRC ), not x4 like in HoI4.
When AI attacks.with division with less than 100 soft attack, the observation does hold true. Sure, eventually AI could build ok templates, but usually it never gets the chance if matched against player.
At some point in game, AI has to fight with an army it has, not one it will have in 5 years. And fighting against human, 1st month usually is where war is won, or lost.
I mostly play with mods that 'fix' the poor AI templates, to make it more competitive.
Heck I even play with AI buffs, and AI getting huge equipment buffs.
Yet still many and massive lop sided battles without casualties to me. so templates and low equipment are a factor, but they are by far not the sole issue either.
Heck I even play with AI buffs, and AI getting huge equipment buffs.
Yet still many and massive lop sided battles without casualties to me. so templates and low equipment are a factor, but they are by far not the sole issue either.
Last edited:
My response to AI offensives in SP:
I think it's worth mentioning that I don't get these rates against human players. (Except in pinning attacks or attacks that get called off quickly.)
I see the AI being overaggressive in poor conditions as a cause for most of this. The catch is that if the Axis isn'y hyper aggressive, it's not going to be WWII, so I don't know what the right answer is. Would anyone be happy if Germany sat quietly along the border of the Soviet Union waiting for a good time to attack once Barbarossa is launched?
Ha, good meme
Secret Master, interesting , very interesting.
Your MP experience points towards my first suspicion of a key component of SP lopsided results:
Experience level Minimum experience Modifier
Green/Fresh 0% −25%
Trained 10% 0%
Regular 30% +25%
Seasoned 75% +50%
Veterans 90% +75%
Player- trains divs to regular, deploys and updates Field Hospital, and I even often rotate divs out of the line if casualties drop them to trained level.
My tanks get seasoned, and even Vet fairly quickly.
The AI does not seem to care about training level, so a -25% vs +25 and even up to +75% could be a big chunk of the SP issue guys?
My thoughts are that WW2 combat, it could be argues that these rates are too much. But for me they are if they gimp the AI so much. Think I will experiment role playing without training everybody, not updating Field Hosp etc.
SM, can you think of anything else that would cause the lopsided results in SP that is highlighted by not happening in MP?
Secret Master, interesting , very interesting.
Your MP experience points towards my first suspicion of a key component of SP lopsided results:
Experience level Minimum experience Modifier
Player- trains divs to regular, deploys and updates Field Hospital, and I even often rotate divs out of the line if casualties drop them to trained level.
My tanks get seasoned, and even Vet fairly quickly.
The AI does not seem to care about training level, so a -25% vs +25 and even up to +75% could be a big chunk of the SP issue guys?
My thoughts are that WW2 combat, it could be argues that these rates are too much. But for me they are if they gimp the AI so much. Think I will experiment role playing without training everybody, not updating Field Hosp etc.
SM, can you think of anything else that would cause the lopsided results in SP that is highlighted by not happening in MP?
A willingness to commit to attacks when modifiers change to be unfavorable (i.e. air superiority changed, resulting in lower defense/breakthrough than when the attack was first launched).
A willingness to commit to attacks when ground attack is in play on the opposing side.
A willingness to commit to attacks across straits and rivers into forts and other silliness without overwhelming firepower.
Misreading where tanks would be better employed versus where infantry would be better employed.
I can probably think of a few others.