How far is HoI3 from a WW2-Simulation?

[X_MasterDave_X]

Before we start this thread, i want to remeber everyone, what Johan said, as HoI3 was still in development, and also in sooner HoI games:

"HoI is a game and no simulation". I even remeber he talked something about pink elephants, but i cant find the exact post. Anyhow...


I want to show the USSR as an example here. In HoI3 many Players said, and even after Semper Fi still say: Its way to easy to win as Germany against an AI USSR. Why is that so? IS the AI flawed? Full of Bugs? Sure HoI3 still have some problems and bugs still, but one of the major problems, why for example the USSR is way easier to conquer, is the simple fact, that the AI cant build the Numbers, the real WW2 USSR built at the war.

Have you ever checked how many Units the USSR-AI can build from June 1941 until 31 Dez 1941, with the amount of IC it has in HoI3? Its only a couple of Divisions....a small fraction of the real historic deployment of new USSR-Divisions in 1941. If you ever asked yourself, why it is so easy to beat the USSR-AI in HoI3, even with Semper Fi and newest patches...its because USSR in HoI3 is a toothless tiger compared to History. In HoI3 the USSR-AI can build around 20-30 Divisions in the half year of 1941. In History they built around 360 new Divisions from June to December 1941. That was more as they had all together in the country (even other borders) as the war broke out against Germany.

Pls check this links for more Infos of the historic WW2 numbers.
http://www.operationbarbarossa.net/
http://www.operationbarbarossa.net/Fast-Facts/Soviet-fast-facts.html#an_8

The Soviets ‘Mobilised and Deployed (MD)’ the following (principal) combat units from 23rd June to 31st December 1941:

* 182 rifle divisions, 43 militia rifle divisions, 8 tank divisions, 3 mechanised divisions, 62 tank brigades, 55 rifle brigades, 50 cavalry divisions, 1 mountain cavalry division, 1 mountain rifle division, 31 airborne brigades, 11 naval infantry brigades, 21 naval rifle brigades, 3 NKVD rifle divisions, 3 NKVD motorised rifle divisions, 20 NKVD border regiments, 2 NKVD security (rifle) brigades, approximately 600 militia fighter battalions, 73 anti-tank regiments, 38 RVGK artillery regiments, 34 RVGK rocket artillery regiments and battalions, 38 engineer and construction battalions, 57 anti-aircraft battalions.
* 41 new Army HQs and 11 new Front HQs.

If you count this together, this are around 361 USSR Divisions build, only in the 6 months after the start of Barbarossa in 1941. The USSR hat according to this source above around 304 Divisons all around the USSR (all frontiers). So the USSR built more as they had at time of the 22. June 1941 in around 6 months. Thats a awesome production ability.


Now, if we want to simulate this in HoI3/SF, how high would we need to set the IC of the USSR to get the AI to produce around 361 Divisons with this compilation above? In my last handsoff, Russia had around 250 IC at June 1941 (2.03b and Lothos 09July Lua). Then after the war starts, they get down around 200 IC at Dec 41, as they lose land and IC capacity. But even at the start the USSR AI cant use all 250 IC for Ground Unit production. In my handsoff, i loaded as Russia in June and they had only 115 of the availiable 250 IC in Unit production. And only 75 IC were into Ground units, the rest was Planes/ships. In Dec41 the Production of Ground units was down around 60 IC, for a couple of Armor and Inf-Units.

So, if we want, that the USSR AI produces the historic 361 Divisions, how much IC did they need? Well this is some complicated calculation. It starts with the talk about, whats a HoI3 Division compared to a WW2 Division. According to the sources above, the USSR had around 4,5 Mio Men in the complete Red Army (all frontiers) at 22. June 1941. This 4,5 Mio were in 304 soviet Divisions. Same time the Wehrmacht had 5 Mio Men in 208 Divisions all together (Russian Border, France, Norway...) at same time. Now let us define a "standard Division" for our comparisons. Lets say 5 Mio Germans serve in 208 Divs, thats a average of around 24,000 per Division (thats fighting personal and all Servants outside the Divisons) So a standard german Div had around 24k. Now, when 4,5 Mio Russians serve in 304 Divs, that would be a average 14,800 per Div. 24k compared to 14,8k is a multiplier of 1,62. Meaning, the "standard" german Division had around 162% of a USSR Div...or with other word 62% more. Now to do the calculation with a standard HoI3 Div, we convert now a WW2 russian Div into a german Div and take this as a HoI3 Div for all other calculations we do later. So from the real 304 USSR Divisions we only have to build 222 HoI3 Divisions, to simmulate the Manpower that USSR brought into war after June 41 until Dez.

Ok, how much IC we need daily to build all that until the 6 months are over?
I did some calculation in Excel, and show you a Foto here:

As you can see, in the upper box i do the calculation how much we need to build the units. Look at the Armor Div, as i loaded in my handsoff in june 1941 i needed 18 IC per day to build one Arm Unit, and it would be for 116 days. So around 2088 IC/days per Armor Division. In WW2 they build 40 in that 6 months, but as we take the "standard Division" thats only 22 Armor Divs (40/1,62). All is rounded. But at the end i get 222 Divs.

What we see there in the upper box means, we would need to have alone 52200 IC/days to produce the Arm Units. Now look at the bottom box. Even in the best month (june) we only have 250 IC per day, meaning only 45,000 IC/days until end of Dezember all together as Industry Capacity. And thats for all....also supply, repair, ships and planes. So all together only 45k and the 22 Tank Divisions would need alone 52k to produce.

And when you calculate all Divisions together we would need the gigantic amount of 151,920 IC/days to build our historic USSR replacement Army. And all the AI has for this use are 75 IC...meaning 13,500 IC/days. Thats a multiplier of 11,25. We would need more then eleven times this IC to produce what the USSR did. And in Dez it looks even worse, as in my handsoff the AI only built with 60 IC ground units. We would need more then 13 times the capacity in Dez for the AI.


How much is now the complete IC we would need for the USSR AI daily?
Well, lets take the 151,920 divided with 180 days, thats around 844 IC per day. We add to that the differnce of the starting 250 IC -minus the 75 IC for Ground Unit production, thats 175 IC. This are the IC the Ai would need to supply, repair build ships/planes and so on. Now we have all together 1019 IC. So, yes the USSR AI would need roughly around 4times as much IC as it has now to produce the historic numbers of Men.

This calculation was made also with much simplicity...i am aware, that the AI would built even more ships/planes when thry had so much Power, but i assumed, Master Lothos would tell the AI, it should only use a smaller %tage to built Ships/planes. Also not in..in this calculation is the higher need for supply and repair. So maybe we would be not around 1000 IC, maybe we would need around 1200-1500 IC. And as i didnt looked at all about historic Plane production, maybe we would even get up until 2000 IC. Oh and the MP we need alone for the Ground Units are 2000 MP until Dez 41.


Now, this is all to show you, how far we are from historic building in HoI3. How far we are from simulating WW2. If a modder wants to use this calculations, he is free to use all, or parts of it. Maybe it would be fun, one day to play a mod, with realistic Unit production. We would also need to find out, how much smaller is the IC for Germany and the other Majors, because, maybe they alos produce much less as in History. Only then this game would be balanced again, and same time a simulation...or at least closer to one. There are so many other multipliers...like loss counts per battle.....and and and.


Thread is open for discussion: What you think about all this calculations? Do you think it would be interesting to play against a WW2 simulated USSR AI?




here is the excel file for download.

 

[Gaizokubanou]

If the AI can build much more than they can now, wouldn't it suffer even more from bad officer ratio because of USSR's poor leadership in this game?

And I think it has a lot to do with the AI. It just cannot manage mass construction well as players can do.

But that's speaking for 1936 scenario where human player using USSR can do massive buildup... otherwise for later scenario yeah their IC is definitely underwhelming to build up a counterattack with. Maybe not 800 IC short underwhelming, but it does feel like they can use 50 or so more IC from central Asia that can be protected and be used to build a fresh army against German advances.

Edit: Also keep in mind that PI won't just implement historically accurate IC/Manpower for the nations in all scenarios because of how undermining it could be to multiplayer balancing.

 

[X_MasterDave_X]

If the AI can build much more than they can now, wouldn't it suffer even more from bad officer ratio because of USSR's poor leadership in this game?

And I think it has a lot to do with the AI. It just cannot manage mass construction well as players can do.

But that's speaking for 1936 scenario where human player using USSR can do massive buildup... otherwise for later scenario yeah their IC is definitely underwhelming to build up a counterattack with. Maybe not 800 IC short underwhelming, but it does feel like they can use 50 or so more IC from central Asia that can be protected and be used to build a fresh army against German advances.

Edit: Also keep in mind that PI won't just implement historically accurate IC/Manpower for the nations in all scenarios because of how undermining it could be to multiplayer balancing.

Sure the Officer ratio need be multiplied also.

 

[ulrichomega]

I think a modifier for the USSR that triggers when at war with Germany should do the trick. Something like: "For the Motherland!:Units take 25% less time to construct, and require 25% less supplies."

 

[unmerged(197132)]

I think something like this is a good idea to make the game more challenging, but I'm not sure that increasing soviet IC is the best way to go about it.

Just because you give the USSR AI the resources to build a historical red army doesn't mean it will use them correctly. Might be better, imo, just to make events that give an AI USSR huge lumps of units/officers/supplies at certain points (start of war, loss of certain provinces, etc), when there's a human Germany. Otherwise you'll likely see the production AI doing something like churning out 2inf/2aa divisions to counter your panzer thrust .

If it's possible, you might also make a modifier for soviets that makes their occupied provinces hugely more likely to spawn partisans.

AI Germany already gets pushed back and defeated due to lack of fuel within 6 months of war against USSR in pretty much every game since the tank fuel use increase was implemented, so I don't find they need help of any kind in that situation.

 

[unmerged(187441)]

Some kind of trickery is used in pretty much every WWII game I've ever seen to manage to mimic the USSR's humongous troop production. When a totalitarian regime tells pretty much every able man to join the army, raising such an army is feasible, but doing so in a way that allows an invaded USSR to do this but doesn't break the game in half is tough indeed.

The closest thing I can think of is a 'desperation economy' mode that can only be triggered when a country's mainland is invaded. At this point, IC goes up x4, and manpower reaches theoretical maximums, at the cost of a dramatic fall in national unity that speeds up the longer the economic system is in this setting.

Something like that would allow the USSR to be fearsome, and make the USA impossible to conquer, turning the invasion of a major country into a race against time: Any large country in this economic setting will maul you to pieces if you give them enough time to put that IC into good use.

 

[lee1026]

Another approach is to have realistic numbers for consumer goods - it should consume at least 50%, even in a very desperate war. (That was the historical amount of production given over to civilians in 1943 for the USSR) Then, make it possible to shaft your population in times of true desperation at the cost of long term production. That will simulate the 1941 USSR situation pretty well, I think.

 

[unmerged(17767)]

I think we need to remember also that throughout the war a German division was roughly equivalent to a Soviet Corps.

That being said I agree that something ought to be done to simulate the Soviets ability to mass produce divisions. The way things stand now a basically historical war in the east is impossible to replicate.

I'm not sure that bumping up their IC that high is the way to go. Possibly a series of events would work better?

It's definitely a tough one.

 

[Karlburg]

One of the primary problems with the game that no other wargame has is that it allows buildup from 1936 that will heavily distort any historical WW2 you will try to attempt. The player will re-arm as the Allies at a break-neck pace for an inevitable war, when war was not inevitable on the minds of the people at the time.

 

[Oerdin]

It's extremely poor as a simulation. While no one wants an exact simulation most of us do want a game where it is possible to recreate WW2 as it happened just so we know it is realistic. Unfortunately, that is not possible with an unmodded HOI3. Maybe a mod fixes the imbalance issues but I haven't seen yet if it does.

 

[Timmy41]

One of the primary problems with the game that no other wargame has is that it allows buildup from 1936 that will heavily distort any historical WW2 you will try to attempt. The player will re-arm as the Allies at a break-neck pace for an inevitable war, when war was not inevitable on the minds of the people at the time.

this is why a strategic event that only triggers under certain conditions would be the answer. there are several already proposed in this thread.

 

[LaDoncella]

The problem lies in that the rules of production, industry, manpower and leadership are static trough the whole game. If you start the game in 1941, the problem is still there, the Soviets cannot rebuild fast enough. And once you break their org you will push them to the pacific without opposition, unless your supply lines implode before.

 

[lee1026]

Well, making it a series of events just for the USSR is kind of a cop out, as a number of nations had the same effect. Germany under speer is a very good example.

 

[dayrinni]

Some kind of trickery is used in pretty much every WWII game I've ever seen to manage to mimic the USSR's humongous troop production. When a totalitarian regime tells pretty much every able man to join the army, raising such an army is feasible, but doing so in a way that allows an invaded USSR to do this but doesn't break the game in half is tough indeed.

The closest thing I can think of is a 'desperation economy' mode that can only be triggered when a country's mainland is invaded. At this point, IC goes up x4, and manpower reaches theoretical maximums, at the cost of a dramatic fall in national unity that speeds up the longer the economic system is in this setting.

Something like that would allow the USSR to be fearsome, and make the USA impossible to conquer, turning the invasion of a major country into a race against time: Any large country in this economic setting will maul you to pieces if you give them enough time to put that IC into good use.

I like the sound of this greatly. Quick, someone make a mod!

 

[aquilarossa]

Well, making it a series of events just for the USSR is kind of a cop out, as a number of nations had the same effect. Germany under speer is a very good example.

I agree with this. If you mod one country, then you also have to mod in all the other country's unique war parameters to keep things even.
Modders can fix stuff, but really the vanilla situation needs some sorting out above all.

This is a massive game and I think I read that PI states it is still in development, which I presume means a WIP. If so, then this is going to be an epic title once most of these fundamental problems are resolved.
If they can fix everything important, I can say goodbye to my life and study, because it will be be very hard to not play this game too much.

 

[Gaizokubanou]

Some kind of trickery is used in pretty much every WWII game I've ever seen to manage to mimic the USSR's humongous troop production. When a totalitarian regime tells pretty much every able man to join the army, raising such an army is feasible, but doing so in a way that allows an invaded USSR to do this but doesn't break the game in half is tough indeed.

The closest thing I can think of is a 'desperation economy' mode that can only be triggered when a country's mainland is invaded. At this point, IC goes up x4, and manpower reaches theoretical maximums, at the cost of a dramatic fall in national unity that speeds up the longer the economic system is in this setting.

Something like that would allow the USSR to be fearsome, and make the USA impossible to conquer, turning the invasion of a major country into a race against time: Any large country in this economic setting will maul you to pieces if you give them enough time to put that IC into good use.

Your suggestion of IC/Manpower boost at national unity hit could actually murder USSR even harder, as Germany could reach the lower national unity to trigger the surrender even faster before all that IC/MP is put to use.

And USSR/USA is already definitive super power in this game for player to use. If you want to have USSR/USA hard to beat as GER on default game setting... keep in mind that would make it trivial to play USSR/USA.

And events are easy to mod... in fact with SF USSR already has this sort of boost in their strategic effect (I think it's called General Winter or something along that line), just go add any modifier you would like for that strategic effect to make USSR more powerful.

 

[plasticpanzers]

simple math solution based upon history is that prewar production is almost always one shift of 8-10 hours per day for production. In wartime its 24/7
which should increase production by up to 300% if you factor in "defend the
motherland" for Russia (building armor in blown out factories in Stalingrad to
drive straight onto the battlefield comes to mind. Wartime production vastly increases output beyond peacetime production especially in Soviet factories. Somewhat unlike the more relaxed conditions US or British factories had to work under.

 

[loki100]

Interesting ideas but to be somewhat contrary, not exactly convinced either by the solution or the underlying analysis.

There is one major problem with all the post-1936 scenarios, your build queue is empty. So if you start with the USSR in the 1941 Barbarossa scenario, your first new rifle divisions aren't ready till about October, when you should have a steady flow of them.

In the actual campaign, by the time the Germans swung south to encircle Kiev, the Soviets were left with no coherent front, they were reduced to disconnected segments guarding the approaches to the key cities, post Kiev, briefly there was almost no front left in the Ukraine. So they didn't have masses of troops in say October.

The Soviet build up to the battle of Moscow came from 3 sources - new formations, the Siberian troops and that for about a month the Germans stopped putting any pressure on. A fair number of the new formations (identified in the original post) were DNO (militia) divisions, which most commanders kept out of the main combat area and in any case had the undesired impact of taking manpower out of the factories. The Soviet OOB by then looks ok, but thats mainly as they very rarely reinforced their rifle divisions, all new manpower went into new formations.

Also, Soviet industry was on what someone calls 24/7 well before the start of the war - I'm doing some reading around Soviet industrial policy in the build up to the GPW and the 1940 Party Conference is full of speeches etc about the need to improve work efficiency. Also the game does not model the time lost when the factories moved east. The Soviets did an amazing organisational job in this respect but still lost 45-60 days production, in game the IC transfers and operates immediately.

So where am I going? Well, I think a human USSR player can generate near to the number of formations the Soviets did - in the main by directing manpower to new units and ignoring reinforcements. The problem in terms of the game is what does the AI do, and how best to sort it. There you either have scripted events or train the AI to improve both its industrial policy and combat performance. From using the betas, and the new lua scripts, I think we're seeing real gains in this respect and I'd rather that was the approach, as scripted events are a great fix but, as in HOI2, they can operate out of context if the game develops differently.

 

[wstein]

Urss

hi all,

in all french mp game, we have for each major countries (as example germany or russia), create historic rules. this rules are supported by a large & objectives historics sources.
to simulate barbarossa and avoid the non historic effort production for the russia, this is our rules :

1st july : activating reserve (from the military district and the mp mobilization plan june 41). the Mp soviet plan mobilization (july 41) is the plan writted in janvier 41 by soviet military officiers.
for this : 30 infantry divisions (2 brigades not 3 + 1 art brigades)

1st septembre : second wave from the reserve unit production :
30 infantry divisions (2 brigades + 1 art brigade) + 10 armoured divisions (1 T34 brigades + 1 BT brigades + 1 support art brigades)

1st november : the role about siberian units : 20 inf divisions (3 brigades + 1 art brigade) + 6 cavalery division (2 brigades cav + 1 brigades arm (BT)

1st janvier 1942 : -25% for all cost of production -40% for armoured. (the use of the factorie's transfert in the Ural & siberie)

in mp game, the challenge is high for german and high for the german. i precise that we play with scenario 38, the most realistic scenario in mp game to avoid all problems with the new factories production etc...)
we have realize 4 MP game : 2 victories for axis & 2 victories for allies.
in this case, the rules show are really well balanced.

 

[wstein]

barbarossa

sorry

i forgot too for the russia, all modifications about the sliders.

(diminution for the CG, for example)

and the reduction for the cost of unit production is not a purcentage but fixed.
example : 10 PP for a arm unit division.

sorry again.