• 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.

theokrat

sexy shoeless god of dice
63 Badges
Feb 6, 2007
5.768
71
  • Hearts of Iron IV: Together for Victory
  • Crusader Kings II: Way of Life
  • Europa Universalis IV: Common Sense
  • Crusader Kings II: Horse Lords
  • Cities: Skylines - After Dark
  • Europa Universalis IV: Cossacks
  • Crusader Kings II: Conclave
  • Stellaris
  • Hearts of Iron IV Sign-up
  • Hearts of Iron IV: Cadet
  • Crusader Kings II: Reapers Due
  • Stellaris: Digital Anniversary Edition
  • Stellaris: Leviathans Story Pack
  • Mount & Blade: Warband
  • Crusader Kings II: Monks and Mystics
  • Stellaris - Path to Destruction bundle
  • Steel Division: Normandy 44
  • Cities: Skylines - Mass Transit
  • Hearts of Iron IV: Death or Dishonor
  • Stellaris: Synthetic Dawn
  • Steel Division: Normand 44 - Second Wave
  • Crusader Kings II: Jade Dragon
  • Hearts of Iron IV: Expansion Pass
  • Stellaris: Humanoids Species Pack
  • Stellaris: Apocalypse
  • Hearts of Iron IV: No Step Back
  • Hearts of Iron III
  • 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: Sword of Islam
  • Europa Universalis IV
  • Europa Universalis IV: Art of War
  • Europa Universalis IV: Conquest of Paradise
  • Europa Universalis IV: Wealth of Nations
  • For the Motherland
  • Crusader Kings II
  • Hearts of Iron III: Their Finest Hour
  • Europa Universalis IV: Res Publica
  • Victoria: Revolutions
  • Semper Fi
  • Supreme Ruler: Cold War
  • Victoria 2
  • Victoria 2: A House Divided
  • Victoria 2: Heart of Darkness
  • 500k Club
  • Cities: Skylines
How to use Militas.

Disclaimer: if you don't like math you should probably stop reading here.

EDIT:
Troughout the dicussion in this thread it became obvious that Defensiveness can not be neglected. I have added a calculation that takes it into account and yields an easy-to-use formular on the optimal ratio of Militia and INF for any given year, combat eff or constraint. It does not however include MTE. You can find it here.
/EDIT


Short while ago there was thread asking for usefulness of Guerilla Warfare which somehow turned into the question of the usefulness of Militas, but could not be answered in an satisfactionary manner. While the Wiki has good an article about a poor mans army it somehow lacks a analytic answer for my taste. So what i would like to do with this thread is have a look on the usefulness of Milita Units respectively when they can become usefull.
If this has been done before on the forums i am sorry. I welcome any response or further suggestions :)

When comparing Militas to INF one can easily see several differences. First of all Militas are ~1/3 the cost in ICd, 1/2 the cost in manpower, 1/5 the cost in Supplies or TCload. On the other hand the combat stats of Militas look pretty crappy when compared to INF, exspecially their toughness is virtually non exist ant. This things already point out for whom- if anyone- Militas can be usefull: a manpower rich, IC low nation with a strong defensive component. There are two examples that spring to mind immediately: China and the Soviet Union. While China fights pretty much all the time we can easily take the Soviet Scenario: If everything behaves "normally" the war between Germany and the Soviet Union is going to start in 1940 or 1941 and the Soviet Player will find himself in a position where at least parts of his front are in a purely defensive role. I think its save to consider the 41 standard as the latest INF model, while for MIL its going to be the 36 model. Now that seems to be totally unfavourable for MIL, but lets see what we can do with them:

The number we are considering to compare different setups is Battle Winning Ability (BWA) a number which describes at a basic level how many damage one can do to the enemy before loosing the battle. At least for the defender it will yield a good comparison between different setups, although it does not take into account own looses or the time one can actually hold a position. The formula is (total attack)*(total org).

Now the situation: A player has a certain amount of ICd to spend, say e.g. 2,000 ICd.
He could build 3 INF or 10 MIL from that- or anything between. Considering the low Soft Attack value of Mil one should take into account the Combat efficiency (CE) explicitly.
I calculated the BWA for different mixtures of MIL vs INF (all mixtures costing 2000 ICdays) and different combat efficiencies, one time defending against a soft target, one time defending against a tank (30% soft).

For the soft target i find the following:

infgc5.jpg


The black numbers are the CEs, the x-axis is the number of INF divisions in the setup (non-integer numbers, because one can of cause have more than 2000 IC days).

Or in the 3d plot:
77029768ig5.jpg



One can clearly see that for low CE the setup with 3 INF per 10 MIL (or 50% of ICdays for Mil, 50% for INF) is optimal, while at higher CE Militas get even more important.

Now Militas have a considerably bad hard attack (0), so whats against a tank?
Here is what i find:
tankum2.jpg


As expected this situation favours INF a little more, but even now we find that a considerable investment of >40% of once ICdays into Militas yields the best results. However one still should note that tanks usually do not make up the main enemy force.

As stated before this takes into account only battles in the defence and completly unbrigaded units, as well as IC as the limiting factors of ones force. I am convinced that a purely INF army is much better in attacking. But even if conducting major offensive actions one might find parts of the front or other areas such as borders (Japanese) or beaches that should only be defended for the moment.


On a concluding remark i want to suggest a little tweak in the usual SOV strategy of IC-whoring in the early years and building INF41 once available. One can start building 2 series of MIL right from the start (they do not need to be upgraded untill in the war), in the first year each will produce ~9 divisions and every year after that ~12 summing up to ~130 Militas at the start of the war. Those can be mixed with INF at a rate of ~1INF/3MIL and deployed along the Japanese boarder or scattered at the German front. One way or the other they free up additional resources for the player. Additionally such a large number of units will speed up autopromotion while the Militas themselves do not really need leaders. In the long run they wont become worthless either, because the 43 standard will make pure MIL forces the best setup for a long time...


Baseline: for IC low nations: use 3 Milita + 1 INF for Defence
 
Last edited:
Very good analysis. I do use militias quite a bit when playing GER. I build 2 stacks of 6 with ART brigades and put them in Finland after I annex that country. Thier purpose is to draw off SOV infantry from the main host. 12 brigaded militia seems to be about the right force to keep SOVs bogged down up north while I send in the tanks, infantry, and TACs in mainland USSR. It's too expensive to rely on infantry to defend Finland and garrisons just get overrun.

After Barbarrosa, those experienced militia groups are good for light defense of the western European coasts. I just put them in strategic provinces that allow them to quickly react to random amphib landings - cheaper than building coastal defenses and more "movable". Besides, I'd rather allow the landing and defeat the invader than prevent the landing in the first place.

jwh
 
I also use miltia for mainly defensif roles, beachead protection or on fronts sometimes even with 1 old mech and 2 miltiias with old slow brigade(artilery orso) good def that way , expierenced+def combo bonus. I also like to think that when mixing them in they draw fire away from your "good" troops.
 
theokrat said:
The number we are considering to compare different setups is Battle Winning Ability (BWA) a number which describes at a basic level how many damage one can do to the enemy before loosing the battle. At least for the defender it will yield a good comparison between different setups, although it does not take into account own looses or the time one can actually hold a position. The formula is (total attack)*(total org).

But shouldn´t we consider the time it takes to loose the org instead of the org itself. Because no matter how much org a unit has, if it looses that org within one hour, there´s hardly any sense in using that unit. Keep in mind that the amout of Org a unit has and the time it can fight in battle is not direct proportional. It depends on the Defensiveness/Toughness value.

Inf41 has a DEF of 24 compared to the 8 of the Mil36. This results in a way faster drop in Org. I don´t know, if the combat system of HoI keeps firing on units out of Org, but if it doesn´t, it is very relevant to keep the DEF/TOU values in mind.
On the other side, if the combat system keeps on firing on units out of Org, the most efficient way might be building lots of militias only to absorb shots that hardly do damage so that your units with most DEF/TOU can keep on fighting for almost ever. This way a defensive line of 30 Mil and 1 Inf would keep on fighting almost as long one of 31 Inf, because the 1 Inf unit would statistically be damaged by about the same amount of shots per minute.
The purpose if this would only be holding the front for as long as possible and not winning the battle. But for SU, this might be a great defensive strategy to win time in the early phase of the war.
 
General Failure said:
I don´t know, if the combat system of HoI keeps firing on units out of Org, but if it doesn´t, it is very relevant to keep the DEF/TOU values in mind.

It most certainly does. Put a couple Arm in the same stack as Militia and send them into battle. The tanks will resist org loss for quite a while, while the Militia will keep getting shelled, org or no org, and their strength will drop accordingly.

This way a defensive line of 30 Mil and 1 Inf would keep on fighting almost as long one of 31 Inf, because the 1 Inf unit would statistically be damaged by about the same amount of shots per minute.

That's a bit of an exaggeration. Remember that once the stack's org drops below 5, it will retreat. If you have eight militia and an infantry in the same stack, and all the militia are out of org, the Infantry's org needs to only reach 45 before the stack's average hits 5/unit.

I actually like it when the AI puts armor in the same stack as militia or cavalry, and they get locked with one of my infantry armies. The battles will sometimes go on forever, and while that one armor will hold out for days, the other units will run out of org and will slowly start to wilt, unable to flee. Given the AI's difficulties in properly re-enforcing, a couple months of this kind of warfare can destroy a whole army group's ability to fight over the long term.
 
BeBro said:
This one could have reached easily 500 replies if the title was "Are militia useless?" ;)

I know, but tough i du trust my calculation, i fealt it would be a bit blasphemy to call it in a blue emu way ;)
I hope i could show something though, i show that 3Militas+1INF can defend just as good as 3 INF- and you still have the IC left to build another INF to send elsewhere. Hopefully a boost in efficiency by 50% should be something i guess.



General Failure said:
But shouldn´t we consider the time it takes to loose the org instead of the org itself. Because no matter how much org a unit has, if it looses that org within one hour, there´s hardly any sense in using that unit. Keep in mind that the amout of Org a unit has and the time it can fight in battle is not direct proportional. It depends on the Defensiveness/Toughness value.

Inf41 has a DEF of 24 compared to the 8 of the Mil36. This results in a way faster drop in Org. I don´t know, if the combat system of HoI keeps firing on units out of Org, but if it doesn´t, it is very relevant to keep the DEF/TOU values in mind.

Generally you are right general, it can be important to compare the time the org drops. Now that time is of cause a direct function of the number of hits a unit suffers: An undefended shot is double as likely to hit as a defended shot. But as blue emu pointed out to me a while ago each shot propably is targeted randomly, so i just asume that they will somehow evenly distribute among the defenders. With that in mind we have to compare the defensiveness of one INF41 (DEF=24) to the DEF of three Militas (the same IC costs). The Militas will all be targeted more or less equally so we just have to add up their Def and we find 3*8=24. So there is no difference at this level. Of cause there might be a little tweaks when rounding, i just had a look at them, green graph is one INF, blue 3 Militas:
defez6.jpg


Note that the baseline is not Zero, but 18. So the difference is there but i would say small.


Not that this is exactly the reason why i despite Militas in an attack, here Toughness is the important parameter and as Militas hardly have any their looses in an attack will be doubled...
 
theokrat said:
With that in mind we have to compare the defensiveness of one INF41 (DEF=24) to the DEF of three Militas (the same IC costs). The Militas will all be targeted more or less equally so we just have to add up their Def and we find 3*8=24. So there is no difference at this level. Of cause there might be a little tweaks when rounding, i just had a look at them, green graph is one INF, blue 3 Militas:
I completely agree with this, but the basic calculation relied on the assumption that the BWA would be (total attack)*(total org). And I think there´s an error. It should be (total attack)*(total_org_Infantry + total_org_Militia/3) as the militias will loose their Org about 3 times faster. Even better should be (total_attack_militia)*(total_org_militia)/3 + (total_attack_infantry)*(total_org_infantry).
P.S. If Imageshack would finally send my account confirmation e-mail I could upload these formulas as images that would not require an entire line of text.

P.P.S. Grüße nach München :)
 
General Failure said:
I completely agree with this, but the basic calculation relied on the assumption that the BWA would be (total attack)*(total org). And I think there´s an error. It should be (total attack)*(total_org_Infantry + total_org_Militia/3) as the militias will loose their Org about 3 times faster. Even better should be (total_attack_militia)*(total_org_militia)/3 + (total_attack_infantry)*(total_org_infantry).
P.S. If Imageshack would finally send my account confirmation e-mail I could upload these formulas as images that would not require an entire line of text.

P.P.S. Grüße nach München :)


The later formular has one flaw, it does not take into account that the Militas keep the INF at battle longer, whihc is kind of their puropse in my calculation. Acting as sort of a punchbag. The other thing is that the MIl will not loose their Org faster than the INF- the attacks are diveded between them, so their DEF is not beeing exeded if the INFs DEF is not exided. So they will loose their org at the same rate (in defence, not in attack).

P.S. You knoww you can host images without subsrcibing?

P.P.S.:
Gruesse nach Erlangen, werde ich in ein paar Stunden auf der A9 dran vorbeifahren ;)
 
theokrat said:
The other thing is that the MIl will not loose their Org faster than the INF- the attacks are diveded between them, so their DEF is not beeing exeded if the INFs DEF is not exided. So they will loose their org at the same rate (in defence, not in attack).

That goes against two years of observations. He who gets hit harder losing org faster. Cavalry lose org faster than armor, unbrigaded lose faster than brigaded, and militia lose faster than infantry. :confused:
 
I use militia a lot for beach defense. With ART they have a good chance of fighting of an invasion in stacks of 3 but often the AI will not even attack the beaches if they see lots of troops stacked on it.

In MP however i don't think this trick would work.
 
Very nice!

I always knew milita was worth building them...
...now i got a mathematic proof.


Yet, there are still things in my mind:

1) Militia need way less supply for being set on offense, thats a easy way to improve mixed stacks CE (and even easier for pure MIL stacks!).
2) I need a detailed explaination of the combat system (again), because it really feels like MIL looses much more org - maybe because you have a lower threshold for "lucky" mass-hits on a unit?
3) What about MIL as "brigade carriers"? (ART/AT)


However, let me say it again:
Great work!


PS: Grüße aus Friedrichsdorf (auch wenn ich nix mit Erlangen am Hut hab :p )
 
Last edited:
General Failure said:
... the basic calculation relied on the assumption that the BWA would be (total attack)*(total org). And I think there´s an error. It should be (total attack)*(total_org_Infantry + total_org_Militia/3) as the militias will loose their Org about 3 times faster. Even better should be (total_attack_militia)*(total_org_militia)/3 + (total_attack_infantry)*(total_org_infantry).
Not true.

If the defending units are receiving exactly 24 incoming shots per turn each, the Militia will lose ORG 1.67 times as fast as Infantry (not three times as fast), a +67% difference in casualty rates, not +200%, between Infantry and Militia... and this is the MAXIMUM differential of loss.

If each defending unit is receiving only 8 shots per turn (or less), there will be no difference at all in the rate of casualties between Infantry and Militia.

If each defending unit is receiving 40 shots per turn, there will be only a +29% difference in casualty rates between Infantry and Militia.

This is because excess (unblocked) shots have double the normal chance of causing damage... so a unit with zero defense would only take twice the casualties as a unit with 1,000,000 defense. All realistic examples will give casualty ratios in between 1.0 and 2.0 (between +0% and +100%)
 
e905433k said:
so does that mean militia is much better for defense than for attack?
Militia is far more useful in Defense than in Attack... even if you ignore the slow walking speed.

Another point that should be mentioned is that Militia only costs 5 Manpower but still has full ORG and 100 Strength points. This means that battle casualties will be replaced at half Manpower-cost, making Militia very economical to keep up to full strength.

1943-model Militia are probably the best defensive unit in the game... in terms of "bangs-per-buck".
 
blue emu said:
Not true.

If the defending units are receiving exactly 24 incoming shots per turn each, the Militia will lose ORG 1.67 times as fast as Infantry (not three times as fast), a +67% difference in casualty rates, not +200%, between Infantry and Militia... and this is the MAXIMUM differential of loss.

If each defending unit is receiving only 8 shots per turn (or less), there will be no difference at all in the rate of casualties between Infantry and Militia.

If each defending unit is receiving 40 shots per turn, there will be only a +29% difference in casualty rates between Infantry and Militia.

This is because excess (unblocked) shots have double the normal chance of causing damage... so a unit with zero defense would only take twice the casualties as a unit with 1,000,000 defense. All realistic examples will give casualty ratios in between 1.0 and 2.0 (between +0% and +100%)

Moreover i compared three Militas vs one INF, as they are ~ the same cost (not counting in the faster gearing bonus of milita). If we compare three MIL vs 1 INF, there will be hardly any difference, the only sources beeing statistical fluctuations in the distribution of the shots and the a little unfavourible rounding when not fighing at 100%, but i strongly suppose that this is very small correction, leading to the result that 3 Mil will loose Org at the same rate as 1 INF against any enemy...