• Crusader Kings III Available Now!

    The realm rejoices as Paradox Interactive announces the launch of Crusader Kings III, the latest entry in the publisher’s grand strategy role-playing game franchise. Advisors may now jockey for positions of influence and adversaries should save their schemes for another day, because on this day Crusader Kings III can be purchased on Steam, the Paradox Store, and other major online retailers.


    Real Strategy Requires Cunning
Status
Not open for further replies.

Jorgen_CAB

Field Marshal
53 Badges
May 2, 2002
5.012
2.562
  • Hearts of Iron IV: Death or Dishonor
  • Stellaris: Galaxy Edition
  • Hearts of Iron IV Sign-up
  • Stellaris Sign-up
  • Hearts of Iron IV: Cadet
  • Hearts of Iron IV: Colonel
  • Hearts of Iron IV: Field Marshal
  • Europa Universalis IV: Rights of Man
  • Stellaris: Digital Anniversary Edition
  • Stellaris: Leviathans Story Pack
  • Stellaris - Path to Destruction bundle
  • Cities: Skylines - Mass Transit
  • Europa Universalis IV: Mandate of Heaven
  • Stellaris: Galaxy Edition
  • Stellaris: Synthetic Dawn
  • Cities: Skylines - Green Cities
  • Hearts of Iron IV: Expansion Pass
  • Stellaris: Humanoids Species Pack
  • Stellaris: Apocalypse
  • Cities: Skylines - Parklife Pre-Order
  • Cities: Skylines - Parklife
  • Stellaris: Distant Stars
  • Cities: Skylines Industries
  • Stellaris: Megacorp
  • Hearts of Iron IV: Expansion Pass
  • Stellaris: Lithoids
  • Cities: Skylines
  • Crusader Kings II
  • Europa Universalis III
  • Europa Universalis IV: Art of War
  • Europa Universalis IV: Conquest of Paradise
  • Europa Universalis IV: Wealth of Nations
  • For the Motherland
  • Hearts of Iron III
  • Hearts of Iron III: Their Finest Hour
  • Europa Universalis IV: Res Publica
  • Semper Fi
  • Sword of the Stars II
  • 500k Club
  • Cities: Skylines Deluxe Edition
  • Europa Universalis III: Collection
  • Europa Universalis IV: El Dorado
  • Europa Universalis IV: Pre-order
  • Pillars of Eternity
  • Europa Universalis IV: Common Sense
  • Cities: Skylines - After Dark
  • Europa Universalis IV: Cossacks
  • Cities: Skylines - Snowfall
  • Europa Universalis IV: Mare Nostrum
4. Terrain maluses (not sure how the brigade ones translate to the full division stats)
My investigation have lead me to the following conclusion.

If you have two infantry brigades and one artillery brigade in a division and attack into a forest you will get an overall negative modifier of -10% because of the artillery in that division.
The artillery has a -30% attack modifier which are diluted throughout the whole division. All penalties and bonuses work like that as far as I know.
 

unmerged(148416)

Sergeant
1 Badges
Aug 5, 2009
89
0
www.egcgames.com
  • Hearts of Iron III
The open-ended nature of HoI3's division design system is astounding. However it suffers from being brigade-based. What does that mean? Its tricky to shoehorn a regiment/battalion based army to brigade-based formations. All fine and dandy when considering UK/Soviet and to some extent US OOB's. The problem lies with German formations especially if a player tries to simulate actual and historic OOB's. Assuming that a brigade consists of two regiments, going the 'historical path' of organising your army can pose a few problems.

Let me elaborate:

The Wehrmacht was more or less a regiment-based army with specialized battalions attached to division formations.

Take panzer divisions for example:

Early War

2 Pz Rgt.
1 Mot Inf Rgt.
2 Art Rgt.
1 Recce Btn
1 PzJg Btn.
1 Pio Btn.

Mid War (Barbarossa)

1 Pz Rgt.
2 Mot Inf Rgt.
1 Art Rgt.
1 PzJG Btn.
1 Pio Btn.

As for Panzer Grenadier Divisions, the plot thickens...

2 Pz Gren Rgt.
1 Pz Btn.
1 Art Rgt.
1 AA Btn.
1 Pz Jg Btn.

IN HOI3 TERMS:

Tank Brigades are larger which means that a 2 Tank brigade 'build' results in over-strength formations, more costly than necessary.

Tank Destroyer and to some extent Armored Car brigades are over-strength as well, meaning a single TD brigade carries more cost than necessary.

Artillery brigades are somewhere in between, over-strength for divisions, under-strength for larger formations i.e. attempting to simulate the Soviet system of Corps artillery formations.

To summarize: Its more beneficial to think out of the box and 'historical accuracy' terms in division design. Perhaps a given group of players attempts to stick to historical OOB's as much as possible. A somewhat problematic approach for regiment-based OOB's such as Germany.

Phew! That was a long one!

P.S. Fantastic work on those spreadsheets!
 

unmerged(152940)

Captain
14 Badges
Aug 11, 2009
317
0
  • Arsenal of Democracy
  • Hearts of Iron II: Armageddon
  • Europa Universalis III Complete
  • Divine Wind
  • Hearts of Iron III
  • Heir to the Throne
  • Magicka
  • Victoria: Revolutions
  • Rome Gold
  • Semper Fi
  • Victoria 2: A House Divided
  • Victoria 2: Heart of Darkness
  • Warlock: Master of the Arcane
Well, you have provided a pretty nice amount of information. In particular, the piece of information that is most important (to me, anyways) is your analysis of the stacking penalty. I don't remember which thread that was, or even if it was this one, but I have the formula you got from regression analysis here:

-0.0813*(x)^2 + 5.4661*(x) - 20.094, where X is the number of divisions. Great. If we apply a bit of calculus, we can determine the optimum number of divisions. First off, let us define what we have in a particular division. Let's assume that we have the Superior Firepower tech magically researched, so that we can have 5 infantry brigades per division. The maximum number of divisions possible is therefore 2, because of frontage. So, let's maximize our SA. The base infantry SA is 2.00. 5*2.00 = 10. So, we have 10 SA per division. So, multiplying 10 by the # of divisions, and then by the stacking penalty (expressed as a percent) will give us a formula for our SA.

Using basic Calculus, we can maximize our SA function by taking the derivative, setting it to zero, and analyzing the critical points. However, the maximum domain we have (because of frontage) is x is less than or equal to 2. If there are no critical points below this, 2 will be our maximum value. So, let's get started.

Take the derivative of this:

10x(-0.0813*(x)^2 + 5.4661*(x) - 20.094)/100 (the 1/100 converts to percent)

d/dx f(x) = -0.02439 (x-42.9021) (x-1.92033)

Now we set this equal to 0, and solve.

-0.02439 (x-42.9021) (x-1.92033) = 0

Our candidate points are X = 42.9021, and X = 1.92. We can use the second derivative test to determine which is a maximum or minimum. For negative values of X, there isn't a maximum. You can make the function infinitely huge by going negative. However, thankfully for us, there's a local min/max.

So we take the second derivative, and we get this lovely expression:

-0.04878 (x-22.4112).

We know there's a root at about 22. If we plug in the critical point and get a positive answer, it's a minimum. If we plug in the critical point and get a negative answer, it's a maximum. Thus, The optimum number of divisions in a province, if frontage were to not apply, is ~ 43 divisions. Which is a massive 215 brigades! And, with the penalty, attacks, etc. etc. the highest SA would be around 250! If resources are not a problem, it might make sense to produce 2 x 5 infantry brigades, plus umpteen million artillery brigades bundled up into divisions.

Without artillery brigades, the most effective number of divisions with 5 infantry brigades is 2. Which is the maximum possible. There is no penalty with less than 4 divisions, IIRC. So the 2 divisions will get an attack of 20.
 
Last edited:

unmerged(152940)

Captain
14 Badges
Aug 11, 2009
317
0
  • Arsenal of Democracy
  • Hearts of Iron II: Armageddon
  • Europa Universalis III Complete
  • Divine Wind
  • Hearts of Iron III
  • Heir to the Throne
  • Magicka
  • Victoria: Revolutions
  • Rome Gold
  • Semper Fi
  • Victoria 2: A House Divided
  • Victoria 2: Heart of Darkness
  • Warlock: Master of the Arcane
I apologize for the double posting, but I have found (also in the Proving the Manual Wrong thread) another formula for the stacking penalty. The original formula I used was the second regression that dhelmet99 made in his Proving the Manual Wrong thread; this provided an optimum number of divisions of about 43 or so. The second calculation comes from Peekee, also in the same thread. It goes like this: 0.95^(X-4), where X is the number of divisions. This one is applicable for X greater than or equal to 4, just like the last one. However, it's much easier to remember. It also has a different optimum value, as we shall shortly see.

The Setup

We'll not be using any sort of standard division this time. What we will do is introduce a new constant, S, which will stand for the value (per division) of the stat you wish to maximize. So, let's set up an equation to maximize.

Sx(0.95^(X-4)) = Total Stat

The Calculation

Let's jump into the calculus of maximizing the function.

d/dx Sx(0.95^(X-4)) =
Sx*ln(0.95)(0.95^(X-4)) + S(0.95^(X-4))

Now we have to set it equal to 0. I'll factor, first, to make it easier to see what to do.

(S * (0.95)^(x-4))(1 + ln(0.95)x) = 0

Okay, now we want to solve for X. We know that the exponential portion can never be zero. So we can safely ignore it, and concentrate on the other, linear, part.

(1 + ln(0.95)x) = 0
1 = -ln(0.95)x
-1/ln(0.95) = x

Plugging this into a calculator, we get ~19.5. I'll skip the second derivative test, having already graphed the function and determining, graphically, that this is, in fact, the maximum.

So, we get ~19.5 as the best number of divisions. We clearly cannot have half a division, so we'll try both 19 and 20 and see what results we get.

For 20 Divisions
S(20)(0.95^(20-4)) = 20*S*(0.95^(15)) = S * 9.26582.

For 19 Divisions
S(19)(0.95^(19-4)) = 19*S*(0.95^(14)) = S * 9.26582 (The same number!)

The Conclusion

19 divisions is better than 20 divisions. Both have stacking penalties above 50%, but the added firepower of the 20th division is only enough to offset the increased stacking penalty. The 20th division provides no extra firepower, yet it requires more production. I suppose 19 divisions is possible to deploy with either attacking from multiple provinces (to get extra frontage) or by making use of militia with half frontage (doesn't a tech give them that?). You can then overload each of the divisions with support brigades to your heart's content.
 

dhelmet99

First Lieutenant
38 Badges
Mar 5, 2004
281
0
  • Crusader Kings II: Reapers Due
  • Victoria 2
  • Victoria 2: A House Divided
  • Victoria 2: Heart of Darkness
  • 500k Club
  • Europa Universalis IV: Pre-order
  • Crusader Kings II: Way of Life
  • Crusader Kings II: Conclave
  • Stellaris
  • Hearts of Iron IV: Cadet
  • Supreme Ruler 2020
  • Crusader Kings II: Monks and Mystics
  • Steel Division: Normandy 44
  • Hearts of Iron IV: Death or Dishonor
  • Hearts of Iron IV: Expansion Pass
  • Hearts of Iron IV: Expansion Pass
  • Crusader Kings II: Holy Fury
  • Imperator: Rome
  • Hearts of Iron IV: Expansion Pass
  • Crusader Kings II
  • Crusader Kings II: Charlemagne
  • Crusader Kings II: Legacy of Rome
  • Crusader Kings II: The Old Gods
  • Crusader Kings II: The Republic
  • Crusader Kings II: Sons of Abraham
  • Crusader Kings II: Sword of Islam
  • Deus Vult
  • Europa Universalis III
  • Hearts of Iron II: Armageddon
  • For the Motherland
  • Hearts of Iron III
  • Hearts of Iron III: Their Finest Hour
  • Europa Universalis III Complete
  • Europa Universalis III Complete
  • Victoria: Revolutions
  • Europa Universalis: Rome
  • Semper Fi
Not a bad analysis Zack, but it is perhaps too general. The only way to get 19 or 20 divisions is to have militia with the width reducing tech (analyzed above) or to attack from multiple provinces, in which case the formula changes (only half the number of divisions count from additional provinces). I haven't looked at attacking from multiple provinces yet, but I don't think it fundamentally changes anything because the stacking penalty scales with the frontage. For regular infantry formations then it would seem that attacking from three provinces would make sense, as this arrives at the 19-20 division optimum. All of this is contingent on the logistical system being able to supply your troops and having enough troops to pack so many in so small a spot.
 

unmerged(157014)

Corporal
1 Badges
Aug 21, 2009
43
0
www.bizarrecreations.com
  • Hearts of Iron III
Interesting stats, but it leaves out one important aspect of the combat system.

Each "round" of combat, each DIVISION involved has a chance to fire... thus having 4 divisions against your enemies 3, gives you a clear advantage.

I've taken it a step further, and created divisions of TWO Combat Brigades, followed by 2 or 3 support Brigades (depending on what tecks I'm working on). Thus having FIVE shots per round as there are 10 frontline 'slots' per province front, and my unit frontage per division is TWO.

This can add to the amount of Brigades involved in the combat... but even 5 divisions of 5 Brigades is only a 25% decrease in combat power.... vice 4 divisions of 5 brigades which would give you a 20% decrease...

I'll take a 20% increase in firepower over a net 5% penalty anyday.

I also often add 2 Combat Brigades to HQ units... it decreases the firepower of that unit... but makes it a very good Second Line unit used to clean up the "leakers" (enemies who manage to retreat back into your lines).
Sorry, but you've misread how the 'shots' work: "Thus having FIVE shots per round as there are 10 frontline 'slots' per province front, and my unit frontage per division is TWO." This isn't correct, you don't get 1 shot per division per round. The number of shots is based on the divisions total soft/hard attack values, from the manual:
"The Firing Phase consists of a series of Shots, the number of which may vary from hour to hour, depending on the capability of the units involved. These are numbered as “Shot 1,” ... The Firing Phase will continue until there are no more Divisions eligible to fire a Shot in the one-hour Round. ... Each Division has a number of shots it may be able to fire during a single Firing Phase. This number is largely dependent on its Soft and Hard Attack values, modified by its Effectiveness. It is also lightly affected by random factors. ... Every round, each Division will randomly select an enemy Division upon which to fire."

On average, low combat width divisions are at a disadvantage versus high combat width divisions in combat, even when you've got the same total number of identical brigades on each side. This is entirely due to the fact that divisions select a target division for the entire round. The larger division gets more shots and can exceed the smaller divisions lower defensiveness/toughness, hence getting guaranteed hits for later shots. The smaller divisions get less shots and can't exceed their opponents higher defensiveness/toughness so all of their shots have a chance to miss. The high width divisions, even when attacked by multiple low width divisions, receive a smaller percentage of total shots as guaranteed hits.

A good case to consider is 2 divisions of 2xinf versus 1 division of 4xinf, where the 4xinf division always concentrates it's firepower on one division of 2xinf, overwhelming it's defensiveness/toughness whereas the 2 2xinf divisions combined can't overwhelm the 4xinf's defensiveness/toughness. It's the guaranteed hits versus the (80%?) chance to miss that make all the difference. The greater the disparity between width on each side, the worse this effect becomes.

Personally I think this is a major flaw, since attacking divisions aren't even going to know exactly where the defending divisions boundaries are anyhow, so concentrating all of their firepower onto just one division is unrealistic. But that's the game mechanic and we're probably stuck with it :( Obviously, you can improve the odds by packing in more support brigades, but that's avoiding the point of a like-for-like comparison in total composition. It's also heading towards the realm of gamey mega-stacks :D
 
Last edited:

Peekee

Field Marshal
36 Badges
Dec 11, 2008
4.601
273
  • For the Motherland
  • Crusader Kings II: Monks and Mystics
  • 500k Club
  • Victoria 2: Heart of Darkness
  • Victoria 2: A House Divided
  • Crusader Kings II
  • Semper Fi
  • Victoria: Revolutions
  • Europa Universalis IV: Res Publica
  • Hearts of Iron III: Their Finest Hour
  • Hearts of Iron III
  • Europa Universalis IV: Call to arms event
  • Europa Universalis IV: Wealth of Nations
  • Europa Universalis IV: Conquest of Paradise
  • Crusader Kings II: Sword of Islam
  • Crusader Kings II: Sons of Abraham
  • Crusader Kings II: The Republic
  • Crusader Kings II: Rajas of India
  • Crusader Kings II: The Old Gods
  • Crusader Kings II: Legacy of Rome
  • Hearts of Iron IV: Colonel
  • Hearts of Iron IV: Field Marshal
  • Victoria 2
  • Hearts of Iron IV: Death or Dishonor
  • Hearts of Iron IV: Expansion Pass
  • Cities: Skylines - Parklife
  • Prison Architect
  • Hearts of Iron IV: Cadet
  • Hearts of Iron IV Sign-up
  • Crusader Kings II: Conclave
  • Victoria 2 Beta
  • 200k Club
  • Crusader Kings II: Charlemagne
  • Cities: Skylines
  • Cities: Skylines - Parklife Pre-Order
In light of the changes made in the spreadsheet due to my vastly increased understanding of the stacking penalty I have seen the need to re-analyze the super-stacks of doom problem. I chose to fix the technology year to 1940, which is roughly the early war period and includes the militia width reducing tech.

The pic below shows all of the mega-stack combinations I analyzed. The first is the comparison of the Swiss Army knife-like stack of 4 different 3 Inf divisions to the super-stack of militia doom (1 Mil + 3 Art times 20 divisions). The stacking penalty is ~55% on the militia stack. Even so the militia stack of doom has much greater combat stats, excepting toughness. It is not cost effective in anything but manpower. In fact it requires massive IC, officers and supply to get that firepower. The supply in particular could cause some serious issues with actually using such a formation (it's a whole frickin' army in one province). Things look better when the number of divisions is reduced to 10 (not shown), but at that point you might as well go to infantry over militia for the better toughness.

Next is the 10 division stack of 1 Inf + 3 Art divisions. This looks like a stack that might actually be usable. The third item in the table shows that this has almost the same combat stats as the giant militia stack of doom at much lower cost.

The last two show how the two main armored division designs stack up (get it?) against the mega infantry-artillery stack. Not that well. The armored division combos cost more IC for inferior combat stats except softness. At some point I may try to convert a softness level into an effective increase in defensiveness and toughness to make the comparison a little more concrete. I still think armored divisions make sense to use due to their high speed and the resulting encirclement tactics, but it looks like super-stacks of infantry and artillery are better for pure firepower.

Note that I could also look at stacks of armor and SP artillery for the most expensive and firepower laden mega-stacks, but I have not yet done so.

Looks like the min width for any division in 1.2 is 1. Thus you can not fit 20 Mil+Art brigades on a single province only 10. Also the support brigades look more expensive.
 

dhelmet99

First Lieutenant
38 Badges
Mar 5, 2004
281
0
  • Crusader Kings II: Reapers Due
  • Victoria 2
  • Victoria 2: A House Divided
  • Victoria 2: Heart of Darkness
  • 500k Club
  • Europa Universalis IV: Pre-order
  • Crusader Kings II: Way of Life
  • Crusader Kings II: Conclave
  • Stellaris
  • Hearts of Iron IV: Cadet
  • Supreme Ruler 2020
  • Crusader Kings II: Monks and Mystics
  • Steel Division: Normandy 44
  • Hearts of Iron IV: Death or Dishonor
  • Hearts of Iron IV: Expansion Pass
  • Hearts of Iron IV: Expansion Pass
  • Crusader Kings II: Holy Fury
  • Imperator: Rome
  • Hearts of Iron IV: Expansion Pass
  • Crusader Kings II
  • Crusader Kings II: Charlemagne
  • Crusader Kings II: Legacy of Rome
  • Crusader Kings II: The Old Gods
  • Crusader Kings II: The Republic
  • Crusader Kings II: Sons of Abraham
  • Crusader Kings II: Sword of Islam
  • Deus Vult
  • Europa Universalis III
  • Hearts of Iron II: Armageddon
  • For the Motherland
  • Hearts of Iron III
  • Hearts of Iron III: Their Finest Hour
  • Europa Universalis III Complete
  • Europa Universalis III Complete
  • Victoria: Revolutions
  • Europa Universalis: Rome
  • Semper Fi
Obviously 1.2 will necessitate an update. For all I know this thread might have influenced the devs to make support brigades more expensive. I'll compile the new stats when I have the time.

Note that it is possible to attack with 20 divisions; you just have to do it from three provinces.
 

Romeo13

Sergeant
32 Badges
Aug 10, 2009
53
0
  • Hearts of Iron III
  • BATTLETECH
  • Europa Universalis IV: Pre-order
  • Stellaris
  • Hearts of Iron IV: Cadet
  • Hearts of Iron IV: Colonel
  • Stellaris: Leviathans Story Pack
  • Hearts of Iron IV: Together for Victory
  • Stellaris - Path to Destruction bundle
  • Steel Division: Normandy 44
  • 500k Club
  • Hearts of Iron IV: Death or Dishonor
  • Age of Wonders III
  • Steel Division: Normand 44 - Second Wave
  • Hearts of Iron IV: Expansion Pass
  • Steel Division: Normandy 44 -  Back to Hell
  • Hearts of Iron IV: Expansion Pass
  • Europa Universalis III
  • Sword of the Stars II
  • Sword of the Stars
  • Semper Fi
  • Europa Universalis IV: Res Publica
  • Europa Universalis III Complete
  • March of the Eagles
  • Europa Universalis III Complete
  • Hearts of Iron III: Their Finest Hour
  • For the Motherland
  • Europa Universalis IV: Call to arms event
  • Europa Universalis IV: Wealth of Nations
  • Europa Universalis IV: Conquest of Paradise
  • Europa Universalis III Complete
Sorry, but you've misread how the 'shots' work: "Thus having FIVE shots per round as there are 10 frontline 'slots' per province front, and my unit frontage per division is TWO." This isn't correct, you don't get 1 shot per division per round. :D
Hmmm.... could be right, but it depends on how the game sequences those shots between the Divisions...


"Keep in mind that even if a Division “sits out” a
round because its attack values didn’t exceed the shot number," From the manual.

Looks like the engine starts off with "shot" one... each division in the combat sees if its Hard Attack, or Soft Attack, allows them to fire.

It then goes to "shot" two... where each division in the combat once again has a chance to fire...

It sequences through the "shots" until everyone is done...

So, each "shot" each DIVISION gets a chance to fire.

Thus 5 Divisions in a combat will put out more initial weight of fire, than 2 divisions... although the Total number of Hard attack or Soft Attack is the same.

Now, this only really matters in the round in which you Break an opposing division... as it will then limit their firepower (they won't get as many shots off as you).... in any round that goes to completion, its immaterial.
 

unmerged(157014)

Corporal
1 Badges
Aug 21, 2009
43
0
www.bizarrecreations.com
  • Hearts of Iron III
Hmmm.... could be right, but it depends on how the game sequences those shots between the Divisions...


"Keep in mind that even if a Division “sits out” a
round because its attack values didn’t exceed the shot number," From the manual.

Looks like the engine starts off with "shot" one... each division in the combat sees if its Hard Attack, or Soft Attack, allows them to fire.

It then goes to "shot" two... where each division in the combat once again has a chance to fire...

It sequences through the "shots" until everyone is done...

So, each "shot" each DIVISION gets a chance to fire.

Thus 5 Divisions in a combat will put out more initial weight of fire, than 2 divisions... although the Total number of Hard attack or Soft Attack is the same.

Now, this only really matters in the round in which you Break an opposing division... as it will then limit their firepower (they won't get as many shots off as you).... in any round that goes to completion, its immaterial.
Ok, 5 divisions of 2xinf vs 2 divisions of 5xinf: the 2xinf divisions will have soft/hard attack 2x of a single inf brigade, the 5xinf divisions will have soft/hard attack 5x of a single inf brigade. Hence the 5xinf divisions will get 2.5 times the shots per round of a 2xinf division. The crititcal bit is that the 5xinf divisions will always apply ALL of those shots to just one of the 2xinf divisions each round. That many shots will exceed the 2xinf's def/tough by a large amount and will therefore always result in a lot of automatic hits.

Even if all the 2xinf divisions attack one 5xinf division (low probability) that's still only 2 times the shots per round (10 vs 5), hence less auto hits. And don't forget that you've got 2 5xinf divisions getting that 2.5 times shots vs def/tough advantage. For all other cases you're mostly just slogging it out with a low probability of hits.

It's all about the probability of overwhelming the opponents def/tough with shots and, most importantly, by how much. The auto hits really are critical.
 

dhelmet99

First Lieutenant
38 Badges
Mar 5, 2004
281
0
  • Crusader Kings II: Reapers Due
  • Victoria 2
  • Victoria 2: A House Divided
  • Victoria 2: Heart of Darkness
  • 500k Club
  • Europa Universalis IV: Pre-order
  • Crusader Kings II: Way of Life
  • Crusader Kings II: Conclave
  • Stellaris
  • Hearts of Iron IV: Cadet
  • Supreme Ruler 2020
  • Crusader Kings II: Monks and Mystics
  • Steel Division: Normandy 44
  • Hearts of Iron IV: Death or Dishonor
  • Hearts of Iron IV: Expansion Pass
  • Hearts of Iron IV: Expansion Pass
  • Crusader Kings II: Holy Fury
  • Imperator: Rome
  • Hearts of Iron IV: Expansion Pass
  • Crusader Kings II
  • Crusader Kings II: Charlemagne
  • Crusader Kings II: Legacy of Rome
  • Crusader Kings II: The Old Gods
  • Crusader Kings II: The Republic
  • Crusader Kings II: Sons of Abraham
  • Crusader Kings II: Sword of Islam
  • Deus Vult
  • Europa Universalis III
  • Hearts of Iron II: Armageddon
  • For the Motherland
  • Hearts of Iron III
  • Hearts of Iron III: Their Finest Hour
  • Europa Universalis III Complete
  • Europa Universalis III Complete
  • Victoria: Revolutions
  • Europa Universalis: Rome
  • Semper Fi
I have thought about this probabilistically and come to the conclusion that the number of divisions will NOT have a significant effect independent of the combat power of the stack. Over time (ie. many rounds of combat) the various factors will average themselves out.

For example, the hypothetical 2 X 5 Inf combat versus 5 X 2 Inf. While it is true that on each round the 5 brigade division will overwhelm whatever division it targets, it is equally true that the 5 divisions will overwhelm one of the 5 brigade divisions. The luck of the draw may favor one of these over the other for any given combat round, but it will eventually even out (or at least come close). I will remain unconvinced until someone sets up a rough numerical simulation to definitively prove otherwise.
 
Sep 1, 2009
8
0
www.johnhafner.info
A good case to consider is 2 divisions of 2xinf versus 1 division of 4xinf, where the 4xinf division always concentrates it's firepower on one division of 2xinf, overwhelming it's defensiveness/toughness whereas the 2 2xinf divisions combined can't overwhelm the 4xinf's defensiveness/toughness. It's the guaranteed hits versus the (80%?) chance to miss that make all the difference. The greater the disparity between width on each side, the worse this effect becomes.

D
I think you forgot division targets for attacking shots are re-assigned randomly each 1 hour round. So it would be spread out over both over time.

Combat is COMPLEX in this game but one important thing to remember is that combat brigade heavy divisons mean less total divisions. More total divisions is usually very useful in defense actions. Since defense is typically much higher than attack value, 2-3inf inf divisions are very cost effective. This is because of oppurtunity cost. Eg. If this unit were larger 3inf of 4inf it would gain defensiveness (but we are already have more than enough to not be overwhelmed) and soft attack (which isn't useful anyway against a typical hardness level of an attacking force) and very little hard attack. The extra cost isn't worth the benefit most likely. Most front lines will crumble against a powerful focused combined arms attack so we are just delaying till we can get our mobiles reserves that have enough firepower to dent the armour to plug the hole.

So the extra cost for a 2-3inf to a 4inf to beef up these stats is a very poor decision if your opponent is attacking with combined arms and 66% or lower softness The stats increased don't offer much benefit against this attack. Now for units used for attacking it is completely different. The main point there is no ONE good unit. 1 inf divisions can MORE than enough def against weak neighbors, but faced with a strong armour heavy german front may need 4 divs of 3inf 1td in every province. It's very situational and complex. Which is why this game is so awesome.
 

Romeo13

Sergeant
32 Badges
Aug 10, 2009
53
0
  • Hearts of Iron III
  • BATTLETECH
  • Europa Universalis IV: Pre-order
  • Stellaris
  • Hearts of Iron IV: Cadet
  • Hearts of Iron IV: Colonel
  • Stellaris: Leviathans Story Pack
  • Hearts of Iron IV: Together for Victory
  • Stellaris - Path to Destruction bundle
  • Steel Division: Normandy 44
  • 500k Club
  • Hearts of Iron IV: Death or Dishonor
  • Age of Wonders III
  • Steel Division: Normand 44 - Second Wave
  • Hearts of Iron IV: Expansion Pass
  • Steel Division: Normandy 44 -  Back to Hell
  • Hearts of Iron IV: Expansion Pass
  • Europa Universalis III
  • Sword of the Stars II
  • Sword of the Stars
  • Semper Fi
  • Europa Universalis IV: Res Publica
  • Europa Universalis III Complete
  • March of the Eagles
  • Europa Universalis III Complete
  • Hearts of Iron III: Their Finest Hour
  • For the Motherland
  • Europa Universalis IV: Call to arms event
  • Europa Universalis IV: Wealth of Nations
  • Europa Universalis IV: Conquest of Paradise
  • Europa Universalis III Complete
Ok, 5 divisions of 2xinf vs 2 divisions of 5xinf: the 2xinf divisions will have soft/hard attack 2x of a single inf brigade, the 5xinf divisions will have soft/hard attack 5x of a single inf brigade. Hence the 5xinf divisions will get 2.5 times the shots per round of a 2xinf division. The crititcal bit is that the 5xinf divisions will always apply ALL of those shots to just one of the 2xinf divisions each round. That many shots will exceed the 2xinf's def/tough by a large amount and will therefore always result in a lot of automatic hits.

Even if all the 2xinf divisions attack one 5xinf division (low probability) that's still only 2 times the shots per round (10 vs 5), hence less auto hits. And don't forget that you've got 2 5xinf divisions getting that 2.5 times shots vs def/tough advantage. For all other cases you're mostly just slogging it out with a low probability of hits.

It's all about the probability of overwhelming the opponents def/tough with shots and, most importantly, by how much. The auto hits really are critical.

I see what you are saying, but you also have to take into account that Each of those 5 division WILL be firing back, at one of those two divisions. The total amount of Firepower is identical. As is the total amount of toughness/defensivness... its just split into larger or smaller groups.

What is not explained in the manual, is how total defensivness works... ie... if two divisions are firing at me, does half go to One Div, and half go to the other? or do I get total defensive stats against each division (hard to believe they'd do that though, pretty unrealisitc).

Ifn I remember.... standard early Inf have S/A of 2, Toughness of 3, Def of 5 (think so, doing this off the top of my head....)...

So... a 5 Brigade Div would have a S/A of 10.... a 2 Brigade Div would have a Def of 10 (would have to see how this scales with Teck increases)... pretty even fight there... all other things equal no clean shots get through.

Now, if the 2 brigade div was on attack? its the 10 S/A of the Big Div against his Toughness of 6... 4 get through clean...

But this example leaves out the biggest part of my Paradigm... which is using 2 Art Brigades (I've always got manpower problems) to increase the firepower of the 2 Inf Brigade Divisions... it tilts their Att/Def type stats towards a more Attack balance... ie, much more S/A, less defensivness... especialy later in the game when tecks start to add up.
 

rejax

Recruit
28 Badges
Apr 28, 2008
4
0
  • Hearts of Iron III
  • Hearts of Iron III: Their Finest Hour
  • Hearts of Iron III Collection
  • Semper Fi
  • For the Motherland
  • Pride of Nations
  • Hearts of Iron IV: Expansion Pass
  • BATTLETECH: Heavy Metal
  • Stellaris: Humanoids Species Pack
  • Stellaris: Apocalypse
  • BATTLETECH - Digital Deluxe Edition
  • Stellaris: Distant Stars
  • BATTLETECH: Flashpoint
  • Stellaris: Megacorp
  • Stellaris: Ancient Relics
  • BATTLETECH: Season pass
  • Stellaris: Lithoids
  • Stellaris: Synthetic Dawn
  • Hearts of Iron IV: Death or Dishonor
  • BATTLETECH
  • Stellaris - Path to Destruction bundle
  • Hearts of Iron IV: Together for Victory
  • Stellaris: Leviathans Story Pack
  • Stellaris: Digital Anniversary Edition
  • Hearts of Iron IV: Cadet
  • Stellaris
  • Magicka: Wizard Wars Founder Wizard
  • Cities: Skylines
If you would take a look at defines.lua, you will find:

BASE_CHANCE_TO_AVOID_HIT = 0.8,
-- Base chance to avoid hit if defences left.
CHANCE_TO_AVOID_HIT_AT_NO_DEF = 0.6,
-- chance to avoid hit if no defences left.

In other words, anyone could still get hit with defense points, and that anyone could still avoid hits without defense points.

Concluding from this, the only benefit for having more divisions is to allow the inclusion of more support brigades to end a fight quicker before random factors average out.
 

pnt

Colonel
53 Badges
May 23, 2008
975
10
  • Semper Fi
  • Europa Universalis IV: Art of War
  • Europa Universalis IV: Conquest of Paradise
  • Europa Universalis IV: Wealth of Nations
  • For the Motherland
  • Hearts of Iron III
  • Heir to the Throne
  • Europa Universalis III Complete
  • Europa Universalis III Complete
  • Europa Universalis IV: Res Publica
  • Imperator: Rome - Magna Graecia
  • Crusader Kings II: Horse Lords
  • Victoria 2
  • Europa Universalis IV: Common Sense
  • Crusader Kings II: Way of Life
  • Mount & Blade: With Fire and Sword
  • 500k Club
  • Europa Universalis IV: El Dorado
  • Hearts of Iron IV: La Resistance
  • Achtung Panzer
  • Hearts of Iron IV Sign-up
  • Hearts of Iron IV: Cadet
  • Hearts of Iron IV: Colonel
  • Crusader Kings II: Reapers Due
  • Hearts of Iron IV: Together for Victory
  • Crusader Kings II: Monks and Mystics
  • Hearts of Iron IV: Death or Dishonor
  • Hearts of Iron IV: Expansion Pass
  • Hearts of Iron IV: Expansion Pass
  • Crusader Kings II: Holy Fury
  • Imperator: Rome
  • Hearts of Iron IV: Expansion Pass
  • Europa Universalis: Rome
  • Europa Universalis III Complete
  • Hearts of Iron Anthology
  • Crusader Kings II
  • Crusader Kings II: Charlemagne
  • Crusader Kings II: Legacy of Rome
  • Crusader Kings II: The Old Gods
  • Crusader Kings II: Rajas of India
  • Crusader Kings II: The Republic
  • Crusader Kings II: Sons of Abraham
  • Crusader Kings II: Sunset Invasion
  • Crusader Kings II: Sword of Islam
  • Europa Universalis III
  • Darkest Hour
  • Europa Universalis IV: Call to arms event
  • Victoria 2: A House Divided
  • Victoria 2: Heart of Darkness
I apologize for the double posting, but I have found (also in the Proving the Manual Wrong thread) another formula for the stacking penalty. The original formula I used was the second regression that dhelmet99 made in his Proving the Manual Wrong thread; this provided an optimum number of divisions of about 43 or so. The second calculation comes from Peekee, also in the same thread. It goes like this: 0.95^(X-4), where X is the number of divisions. This one is applicable for X greater than or equal to 4, just like the last one. However, it's much easier to remember. It also has a different optimum value, as we shall shortly see.

The Setup

We'll not be using any sort of standard division this time. What we will do is introduce a new constant, S, which will stand for the value (per division) of the stat you wish to maximize. So, let's set up an equation to maximize.

Sx(0.95^(X-4)) = Total Stat

The Calculation

Let's jump into the calculus of maximizing the function.

d/dx Sx(0.95^(X-4)) =
Sx*ln(0.95)(0.95^(X-4)) + S(0.95^(X-4))

Now we have to set it equal to 0. I'll factor, first, to make it easier to see what to do.

(S * (0.95)^(x-4))(1 + ln(0.95)x) = 0

Okay, now we want to solve for X. We know that the exponential portion can never be zero. So we can safely ignore it, and concentrate on the other, linear, part.

(1 + ln(0.95)x) = 0
1 = -ln(0.95)x
-1/ln(0.95) = x

Plugging this into a calculator, we get ~19.5. I'll skip the second derivative test, having already graphed the function and determining, graphically, that this is, in fact, the maximum.

So, we get ~19.5 as the best number of divisions. We clearly cannot have half a division, so we'll try both 19 and 20 and see what results we get.

For 20 Divisions
S(20)(0.95^(20-4)) = 20*S*(0.95^(15)) = S * 9.26582.

For 19 Divisions
S(19)(0.95^(19-4)) = 19*S*(0.95^(14)) = S * 9.26582 (The same number!)

The Conclusion

19 divisions is better than 20 divisions. Both have stacking penalties above 50%, but the added firepower of the 20th division is only enough to offset the increased stacking penalty. The 20th division provides no extra firepower, yet it requires more production. I suppose 19 divisions is possible to deploy with either attacking from multiple provinces (to get extra frontage) or by making use of militia with half frontage (doesn't a tech give them that?). You can then overload each of the divisions with support brigades to your heart's content.
While a simple calculation is always nice, given the variety of people in this forum, perhaps it would be better to simply plot the strength function? The optimum will be immediately visible, and it will give a much better feeling for the higher derivatives.
 

unmerged(152940)

Captain
14 Badges
Aug 11, 2009
317
0
  • Arsenal of Democracy
  • Hearts of Iron II: Armageddon
  • Europa Universalis III Complete
  • Divine Wind
  • Hearts of Iron III
  • Heir to the Throne
  • Magicka
  • Victoria: Revolutions
  • Rome Gold
  • Semper Fi
  • Victoria 2: A House Divided
  • Victoria 2: Heart of Darkness
  • Warlock: Master of the Arcane
While a simple calculation is always nice, given the variety of people in this forum, perhaps it would be better to simply plot the strength function? The optimum will be immediately visible, and it will give a much better feeling for the higher derivatives.
While plotting the function might be good for those who don't know anything about calculus (and there are many), I didn't have the time to generate a nice fancy graph. I had just reinstalled Windows and had no program to graph with. Plus, I can easily modify the calculation to take into account other negatives. Have a -25% penalty from something (and no other negative/positive modifiers)? Well then, we change it a bit, and there's our number. In this case it'd be 12 divisions. But then again, I believe patch 1.2 changed it so that the minimum width a division can have is 1, so debating whether or not 12 or 19 divisions is best is kind of a moot point anyways.
 

pnt

Colonel
53 Badges
May 23, 2008
975
10
  • Semper Fi
  • Europa Universalis IV: Art of War
  • Europa Universalis IV: Conquest of Paradise
  • Europa Universalis IV: Wealth of Nations
  • For the Motherland
  • Hearts of Iron III
  • Heir to the Throne
  • Europa Universalis III Complete
  • Europa Universalis III Complete
  • Europa Universalis IV: Res Publica
  • Imperator: Rome - Magna Graecia
  • Crusader Kings II: Horse Lords
  • Victoria 2
  • Europa Universalis IV: Common Sense
  • Crusader Kings II: Way of Life
  • Mount & Blade: With Fire and Sword
  • 500k Club
  • Europa Universalis IV: El Dorado
  • Hearts of Iron IV: La Resistance
  • Achtung Panzer
  • Hearts of Iron IV Sign-up
  • Hearts of Iron IV: Cadet
  • Hearts of Iron IV: Colonel
  • Crusader Kings II: Reapers Due
  • Hearts of Iron IV: Together for Victory
  • Crusader Kings II: Monks and Mystics
  • Hearts of Iron IV: Death or Dishonor
  • Hearts of Iron IV: Expansion Pass
  • Hearts of Iron IV: Expansion Pass
  • Crusader Kings II: Holy Fury
  • Imperator: Rome
  • Hearts of Iron IV: Expansion Pass
  • Europa Universalis: Rome
  • Europa Universalis III Complete
  • Hearts of Iron Anthology
  • Crusader Kings II
  • Crusader Kings II: Charlemagne
  • Crusader Kings II: Legacy of Rome
  • Crusader Kings II: The Old Gods
  • Crusader Kings II: Rajas of India
  • Crusader Kings II: The Republic
  • Crusader Kings II: Sons of Abraham
  • Crusader Kings II: Sunset Invasion
  • Crusader Kings II: Sword of Islam
  • Europa Universalis III
  • Darkest Hour
  • Europa Universalis IV: Call to arms event
  • Victoria 2: A House Divided
  • Victoria 2: Heart of Darkness
I am not saying that your calculation is not useful. :) And regardless of the game, it is nice to demonstrate how simple certain problems become if you know a little math - so hopefully can stimulate an interest among players who are in high-school or college.

However, from a game point of view, looking for stationary points does not quite tell the full story. You really want to know the gain of adding one more division. From this point of view, having plots of the strength function and its first derivative would be very useful. It would be very nice if you could update your calculation, and also find some way to make the plots. And since the strength varies with tech level, and depends on the softness of your opponent, in fact one could calculate several optima.

What you might also want to consider, however, is not to limit yourself to studying the effective strength. The 1.2 patch was a major improvement over 1.1 in the sense that now, for instance, Art has a higher supply usage per attack than Inf. Thus, if you are supply limited (which you will be in a stack of doom), you want to max out your inf first. When transitioning from 12 Inf + 8 Art per province (with no stacking penalties) to 10 Inf and 11 Art you will see a bigger kink than otherwise. In any case, it would be nice to also analyze the behavior of functions like strength/supply to get a more nuanced view of the problem.
 

unmerged(150728)

Corporal
1 Badges
Aug 9, 2009
47
0
  • Hearts of Iron III
The updated link is below and I'll edit the earlier post to reflect this too.

http://www.mediafire.com/file/z5iermrgnro/Division_designer3__8-24.xls
I think there's a problem calculating the frontage of armour brigades. It's showing zero frontage for LA, Arm and HA regardless of year. The problem seems to be in Division_database!AS125 etc.

Nice work though, this is really handy. One suggestion, excel's vlookup() function could save you some work - just make sure to set the fourth parameter to FALSE.

Edit: this problem also affects the softness calculation on armoured cars.
 
Last edited:

Buh the Ghost

Second Lieutenant
8 Badges
Sep 4, 2009
194
0
  • Europa Universalis III Complete
  • Hearts of Iron III
  • Heir to the Throne
  • Europa Universalis III Complete
  • Europa Universalis III Complete
  • Victoria 2
  • 500k Club
  • Pride of Nations
hmm... ok. very interesting thread! it took me hours to read it and all the links, but i think it was worth its time. :) I like these "what if we have enough IC and time to produce a million mules. Can they beat an inf division?" - discussions. In HoI 2 I started with one and ended up teching to mech divisions and starting war in late 1943 :D.


Go on with this guys, though it was frustrating to see, that - according to these information - my lovely "2 Armor + 2 Mech + SP R Art" Divisons seems to be quite... ineffective. I am one of the speed people and will always be one. :cool: I owned france with 16 early versions (2LArmor, 2Mot) of them and the starting inf holding the border at the magginot line. but i'm getting off topic.


It is like u already said, speed can be a huge factor and not be calculated. what is the need of a super milita stack if the enemy just drives around it?

But if the task is holding a single province, then it is a shame for HoI, that a militia/art super stack seems to be the "best" way. This should not be the case.


Greetings,
BtG

(oh and btw: lower the aircraft stacking penalty!! ;))
 

pnt

Colonel
53 Badges
May 23, 2008
975
10
  • Semper Fi
  • Europa Universalis IV: Art of War
  • Europa Universalis IV: Conquest of Paradise
  • Europa Universalis IV: Wealth of Nations
  • For the Motherland
  • Hearts of Iron III
  • Heir to the Throne
  • Europa Universalis III Complete
  • Europa Universalis III Complete
  • Europa Universalis IV: Res Publica
  • Imperator: Rome - Magna Graecia
  • Crusader Kings II: Horse Lords
  • Victoria 2
  • Europa Universalis IV: Common Sense
  • Crusader Kings II: Way of Life
  • Mount & Blade: With Fire and Sword
  • 500k Club
  • Europa Universalis IV: El Dorado
  • Hearts of Iron IV: La Resistance
  • Achtung Panzer
  • Hearts of Iron IV Sign-up
  • Hearts of Iron IV: Cadet
  • Hearts of Iron IV: Colonel
  • Crusader Kings II: Reapers Due
  • Hearts of Iron IV: Together for Victory
  • Crusader Kings II: Monks and Mystics
  • Hearts of Iron IV: Death or Dishonor
  • Hearts of Iron IV: Expansion Pass
  • Hearts of Iron IV: Expansion Pass
  • Crusader Kings II: Holy Fury
  • Imperator: Rome
  • Hearts of Iron IV: Expansion Pass
  • Europa Universalis: Rome
  • Europa Universalis III Complete
  • Hearts of Iron Anthology
  • Crusader Kings II
  • Crusader Kings II: Charlemagne
  • Crusader Kings II: Legacy of Rome
  • Crusader Kings II: The Old Gods
  • Crusader Kings II: Rajas of India
  • Crusader Kings II: The Republic
  • Crusader Kings II: Sons of Abraham
  • Crusader Kings II: Sunset Invasion
  • Crusader Kings II: Sword of Islam
  • Europa Universalis III
  • Darkest Hour
  • Europa Universalis IV: Call to arms event
  • Victoria 2: A House Divided
  • Victoria 2: Heart of Darkness
Go on with this guys, though it was frustrating to see, that - according to these information - my lovely "2 Armor + 2 Mech + SP R Art" Divisons seems to be quite... ineffective. I am one of the speed people and will always be one. :cool: I owned france with 16 early versions (2LArmor, 2Mot) of them and the starting inf holding the border at the magginot line. but i'm getting off topic.
As a major, a human player can always win the early scenarios with relative ease (try Germany in 1944 for more fun), so variety can be entertaining. Squeezing the most out of your research and production is most important for minors. For instance, as Italy I managed to build an extremely powerful '36 scenario army based on divisions with

3 Arm + 1 AC, supported by pure Inf, Mtn, and Para (I scrapped the Cav and Mot, as well as the Mil after Ethiopia), a navy based on BB and DD, and an air force with Int, Tac, and Nav. This gave an enormous punch despite limited resources.

In any case, I tried to summarize everything on the wiki.

http://www.paradoxian.org/hoi3wiki/Division_composition


But if the task is holding a single province, then it is a shame for HoI, that a militia/art super stack seems to be the "best" way. This should not be the case.
I absolutely agree. The good thing, however, is that due to the change in attack/supply ratio, artillery is no longer a relevant alternative for defense of anything except individual provinces (which is historically correct). As for its use on the attack, it only makes sense if you already have and can supply 12 Inf brigades per province, at least in that area, but need more! :)

It still makes sense in combination with armor in some cases, though (see link above), and in particular rocket art can be used in some fun combos.
 
Status
Not open for further replies.