Combat mechanics seriously flawed

[Olle8]

I am talking about attacker vs defender losses here.. if you just dug in and wait for the enemy to attack you then you wont loose barely anything.. in previous hearts of iron games you would still have some losses even how well dug in you were (alot smaller losses than the attacker thou). But in Hoi iv it is so common that you have 0 losses in most defending battles if you are dug in and the attacker might loose thousands....... i mean okay if you have 10-20% of the attackers losses cause ur so dug in but ZERO????????

Are the developers even aware of this?

 

[mursolini]

I am talking about attacker vs defender losses here.. if you just dug in and wait for the enemy to attack you then you wont loose barely anything.. in previous hearts of iron games you would still have some losses even how well dug in you were (alot smaller losses than the attacker thou). But in Hoi iv it is so common that you have 0 losses in most defending battles if you are dug in and the attacker might loose thousands....... i mean okay if you have 10-20% of the attackers losses cause ur so dug in but ZERO????????

Are the developers even aware of this?

In previous HOI games situation was absolutely the same, with a tiny difference that it was far harder to build a bad division. In fact, in HOI4 attacking is generally much easier.

Issue is with AI that is very suicidal and has really bad templates.

 

[KevinG]

Tanks and battleplanner bonus makes attacking easier in HOI4. In HOI3 everything had close to a 1:1 ratio of breakthrough/defense. Tanks had slightly more breakthrough than they had defense but in HOI4 you can design a division with 1k+ base breakthrough and 300 defense. You can also see absurd things like 4k soft attack due to max planning bonus + veterans + country bonus that wasn't possible in HOI3 either.

 

[Gort11]

It definitely feels like casualties are too low when things are going well. Americans in Vietnam were touting their ten-to-one kill ratio at places like Ia Drang, and that's where they were defending, had howitzer support, and had over 800 bomber sorties to help them. You can easily exceed that kind of kill ratio in Hearts of Iron 4 just by stacking your line with 7 inf / 2 art divisions and waiting for the enemy to attack you.

 

[Dan1109]

10:1 was a bad day....but yeah, fully agree with OP. Very good observation. Podcat did say at PdxCon that 1.5 will 'make defense more fun' which I think means giving a method for the AI to better handle defense and some new mechanic to counter-act planning bonus, as entrenchment is really just good for humans ATM. Maybe something in 1.5 will change the current rofl-defense behavior.

 

[--Yigito123--]

I remember a game where AI Poland and Romania actually WON (Well to be fair, Iran and Finland jumped in too later on, but too late to do much, just enabled the Poles to advance) against the AI Soviet Union in a war where they lost 2 million men together while the Soviets lost 17 million! So... I don't think it can be attributed to just "AI builds bad templates, player builds good, player takes less casualties".

 

[fabius]

OP is right. But the issue is much wider.

I've posted a thread about the buffs causing far too many completely lopsided battles. Really is an immersion breaker, and its so out of whack that I see it as an elephant in the room.

How come the forum isn't in uproar about this, or at least revolting a little. Ground combat is the core of the whole game and at present the 0 losses to Thousands of enemy is just silly.

Also, balance this and I reckon it would really help the AI.

Come on devs, can we please have an acknowledgement that you'll look at 'fixing' or balancing the movie style casualty stats ?

 

[mursolini]

It definitely feels like casualties are too low when things are going well. Americans in Vietnam were touting their ten-to-one kill ratio at places like Ia Drang, and that's where they were defending, had howitzer support, and had over 800 bomber sorties to help them. You can easily exceed that kind of kill ratio in Hearts of Iron 4 just by stacking your line with 7 inf / 2 art divisions and waiting for the enemy to attack you.

That is because AI choses battles poorly. If you try playing Germany, you can complete invasions of Poland, Benelux and France on less than 15k cassualties, earning you something like 1:20 cassualty ration while attacking. When you attack infantry with tanks, while having air superiority, you can expect absolutely insane cassualty ratio as well.

I remember a game where AI Poland and Romania actually WON (Well to be fair, Iran and Finland jumped in too later on, but too late to do much, just enabled the Poles to advance) against the AI Soviet Union in a war where they lost 2 million men together while the Soviets lost 17 million! So... I don't think it can be attributed to just "AI builds bad templates, player builds good, player takes less casualties".

Terrain also matters, Romania has river to improve their odds notably. Problem is AI is suicidal and will attack at disadvantage.

 

[Olle8]

That is because AI choses battles poorly. If you try playing Germany, you can complete invasions of Poland, Benelux and France on less than 15k cassualties, earning you something like 1:20 cassualty ration while attacking. When you attack infantry with tanks, while having air superiority, you can expect absolutely insane cassualty ratio as well.

Terrain also matters, Romania has river to improve their odds notably. Problem is AI is suicidal and will attack at disadvantage.

No.... its not about it being poorly.. i am okay with an attack being poorly done and resulting in terrible kill death ratios but when it is 0 and the other side looses 7353 manpower it is something flawed, say some odd number 353 losses vs 4446 in a terrible attack would be ok, but when it is 0 it is just so unrealistic and bad.

 

[Dan1109]

Sounds like their formula needs an inverse log somewhere...

 

[jaredstanko]

Positive modifiers could be lowered slightly. Slightly. Maybe a stacking penalty past 200%. But positive modifiers have nothing to do with casualty ratios. What really needs to happen is penalties need to be lowered signifigantly. We need a floor on combat modifiers, so you can expect a certain type of offensive to never push, but still do org damage and cause the offensive that matters to push forward. Maybe 10%, and under that it just counts as 10%? 10000 to 1 casualty ratios are insanely stupid. On top of that negative modifiers need a slight nerf. Just around 20% on average, so a modifier that was previously50% might be 40%. With some exceptions, like the amphibious penalty to tanks.

The thing is, if you dive into the division battle system, its actually an incredible thing of beauty, but it desperately needs tweaks, and for all intents and purposes combat hasn't changed since 1.0. It certainly needs a numbers pass, and maybe a free patch dedicated to some ironing(fix combat width, fix org on small divisions, fix combat modifiers by adding a small scaling penalty, and a combat modifiers floor of 5-10%.)

The way combat works, if you have less attack than your enemy has defense(defense or breakthrough), you do 10% damage, any attack overflow does 40% damage, but the combat system assumes that in most cases you wont have more attack than your enemy defense. That is only for special cases like a very lopsided fights. The MAIN PROBLEM with casualty ratios are attackers having literally 0 defense(taking insane casualties) what having literally 0 attack. Nobody would attempt such a hopeless assault.

 

[impspy]

Would making combat-width more fluid help?

It hurts my immersion to see 40-with divisions being the best strategy for a lot of builds and makes me wonder why Pdox moved from regiments to battalions.

 

[Dan1109]

Would making combat-width more fluid help?

It hurts my immersion to see 40-with divisions being the best strategy for a lot of builds and makes me wonder why Pdox moved from regiments to battalions.

I'd like to see the combat simply happen on a per battalion basis - that would be the only fair way with the current system. But...too many units.

It's like they need to divide the stats of the division by its combat width to obtain a single width stat. Then 40w stats would be more comparable to a 10 or 20w div. Of course a 40w div should still get 4x the attacks and dodges of 10w div.

40w should be for lazy fockers, not the best armored attack division. And FM/CC exp should be given for the amount of troops in combat, not the amount of divisions.

 

[fabius]

Positive modifiers could be lowered slightly. Slightly. Maybe a stacking penalty past 200%. But positive modifiers have nothing to do with casualty ratios. What really needs to happen is penalties need to be lowered signifigantly. We need a floor on combat modifiers, so you can expect a certain type of offensive to never push, but still do org damage and cause the offensive that matters to push forward. Maybe 10%, and under that it just counts as 10%? 10000 to 1 casualty ratios are insanely stupid. On top of that negative modifiers need a slight nerf. Just around 20% on average, so a modifier that was previously50% might be 40%. With some exceptions, like the amphibious penalty to tanks.

The thing is, if you dive into the division battle system, its actually an incredible thing of beauty, but it desperately needs tweaks, and for all intents and purposes combat hasn't changed since 1.0. It certainly needs a numbers pass, and maybe a free patch dedicated to some ironing(fix combat width, fix org on small divisions, fix combat modifiers by adding a small scaling penalty, and a combat modifiers floor of 5-10%.)

The way combat works, if you have less attack than your enemy has defense(defense or breakthrough), you do 10% damage, any attack overflow does 40% damage, but the combat system assumes that in most cases you wont have more attack than your enemy defense. That is only for special cases like a very lopsided fights. The MAIN PROBLEM with casualty ratios are attackers having literally 0 defense(taking insane casualties) what having literally 0 attack. Nobody would attempt such a hopeless assault.

This is a very interesting analysis, that on the face of it sounds like you're on to something with the negative modifiers. I'm going to look at the attack/defence modifiers again with this in mind later.

I'd add: Podcat previously said, ground combat hasn't changed radically or that much since Hoi3.

I don't think this needs a radical change, like you said looks like tweaks and balancing would fix most of it. But maybe another look at the core ground combat mechanics at some point wouldn't be a bad thing.

 

[Claremont Waltz]

Its weird that people reached for modern warfare rather than the obvious example :

https://en.wikipedia.org/wiki/Battle_of_France

As you can see Axis casualties were ~5% of allied ones, or about 20:1. Most axis casualties in BOF were sustained in the little known Case Red (assuming my memory hasn't fucked me), not Case Blue.

https://en.wikipedia.org/wiki/Fall_Rot

(this particular campaign is poorly studied and contains fun stuff like the second BEF evacuation out of Cherbourg)

The other super obvious example is the opening weeks of barbarossa. This is from The Viazma Catastrophe by Lev Lopukhovsky. As with any discussion of casualty figures in the eastern front there will be disagreements, so just take these as convenient approximations rather than hard figures.

Thus, the Red Army over 15-18 days of combat operations lost 747,870 personnel, of which 588,598 were killed or missing and 159,272 wounded or sick. In comparison, by 6 July 1941, the Wehrmacht’s army losses on the Eastern Front (not including the sick) amounted to 13,869 (829) killed, 42,755 (1,588) wounded, and 5,010 (81) missing-in-action (the figures in parenthesis show officer losses, which were not included in the totals).4 Thus over the period in question, the Germans lost 64,132 soldiers and officers, of which 19,798 were killed or missing-in-action. The correlation of total personnel losses was therefore 11.7 to 1 against the Soviet side. The irrecoverable losses of Soviet military personnel proved to be almost 30 times larger than the Germans’, while so-called medical casualties [wounded and sick] were 3.6 times greater.

Paradox is trying to model these incredibly lopsided results. They are not typical, and arose out of very specific and difficult to replicate historical situations. Things like the 55th ID (FRA) literally disintegrating with many personnel fleeing having never even seen a German (see The Breaking Point by Robert Doughty) simply cannot arise out of normal game play.

It would be much better to create a more balanced combat system less prone to lopsided results and instead give event based penalties if players have done roughly historical things as GER, ENG, FRA or SOV. It'd make most sense to gate various goody focuses behind such an event. Ex if France is Allies and has gone with extend Maginot and Czech annexed and other parameters ex anus, big penalty.

 

[Carlyle1]

Attacking is hard actually since defense stat is extremely cheap compared to breakthrough. Also harsh terrain can also be a pain in the a**. %50 max planning bonus doesn't cut it. Sometimes as Germany you want to go GBP doctrine.

 

[Gort11]

Attacking is hard actually since defense stat is extremely cheap compared to breakthrough. Also harsh terrain can also be a pain in the a**. %50 max planning bonus doesn't cut it. Sometimes as Germany you want to go GBP doctrine.

I actually feel like Grand Battleplan is a really good doctrine for Germany. You can build up a huge planning bonus in peacetime, then blitz irresistably across small countries.

 

[Claremont Waltz]

I actually feel like Grand Battleplan is a really good doctrine for Germany. You can build up a huge planning bonus in peacetime, then blitz irresistably across small countries.

GBP is expensive in wartime because you need signal companies to get planning bonus in sane time frame.

You can do the same thing by slamming into the enemy with concentrated attacks from multiple provinces to a single province while fixing enemy divisons in place with single division holding attacks along the line.

If against humans you probably want to have multiple points where you that so that they can't just dump all reserves in one spot.

Once that's done send high speed exploitation through and clean up.

 

[Alex_brunius]

Its weird that people reached for modern warfare rather than the obvious example :

https://en.wikipedia.org/wiki/Battle_of_France

As you can see Axis casualties were ~5% of allied ones, or about 20:1. Most axis casualties in BOF were sustained in the little known Case Red (assuming my memory hasn't fucked me), not Case Blue.

https://en.wikipedia.org/wiki/Fall_Rot

(this particular campaign is poorly studied and contains fun stuff like the second BEF evacuation out of Cherbourg)

The other super obvious example is the opening weeks of barbarossa. This is from The Viazma Catastrophe by Lev Lopukhovsky. As with any discussion of casualty figures in the eastern front there will be disagreements, so just take these as convenient approximations rather than hard figures.

The difference is that those casualties are as a result of large encirclements / mass surrender... Not from combat.

When Germans and French forces fought in France the Germans didn't inflict an average of 20:1 casualties in the individual battles.


I have no problem with being able to inflict 20:1 or more casualties if you encircle the enemy and get an entire army to surrender ( and I don't think anyone else have either ). However from normal combat it just shouldn't be possible to achieve such results unless we are talking about under-equipped militia attacking against maginot line level fortifications without support.

 

[Killerrabbit]

Would making combat-width more fluid help?

It hurts my immersion to see 40-with divisions being the best strategy for a lot of builds and makes me wonder why Pdox moved from regiments to battalions.

The 40 width "minigame" has been acknowledged by devs that it's something they are looking at getting rid of - but this and the fact that 1 large Div > 2 smaller divs (due to armor, breaktrough, piercing values) has been said to be a hard thing to fix properly since averaging all forces on the same side in a combat will lead to other problems.