Military Model VI

[Mark_Everson]

Hi Laurent:

Option 1: Have the archer element be able to split its actions in the code. Essentially the archers would have a part of them defined as in-line and a part as in-support. The support ones would use their weapons as always, and the in-line ones would fight with whatever melee weapons were available when the enemy closed. In case its useful, here is a link to discussion on arbitrary-sized elements. It only talks about frontage, but I think the principles can, I believe, be generally applied to your issue.

Option 2: Of course if you want to make your life easier with the existing code, you can create extra elements just for the duration of the fight. Then the "front" archer element can be of exactly the size needed to plug the gap in the line. This at least should take care of model artefacts. Of course the archers would need to be put back together once the fight is over. But I'd wait to hear what Gary says on this idea first, since I think it not unlikely he'll object on some arcane coding principle or another

[Mark_Everson]

Hi Laurent:

A while ago it was true IIRC that Army Support was the same for an element whether it was full-strength or almost eliminated. Has the code yet been updated so that everthing is proportional?

[LDiCesare]

Mark,
if you use the CombatData object to retrieve the cost (which should be the case, but there is a rogue method in Element), then the upkeep cost is proportional to the health of the element (unit).

[Nadexander]

Re: Should manpower for Units come from whole Province?

Originally posted by Mark_Everson
Right now we require 5000 population in the relevant square to go into building a military unit. (Although build points are harvested across the province, the build still happens in a singel square.) I think we're going to have trouble with this in the Dawn scenario. Even if your starting square has enough population to build the unit, it could be severly depopulated. I think we are ok for now if there is Just the starting square, since the build would go so slowly the population would likely get to 12k before it was finished.

A realistic and playable solution would be to allow a disbanding of the unit to return the population to the square. Most military units werent full time proffesional soldiers anyways. Military units were raised in times of war and then dispersed so that soldiers could return for the planting/harveest. You probably want to cap at no more than 1/2 the population to be able to be part of a military unit at any given time (to account for women, children, elderly, etc. etc. who wouldnt fight)

[LDiCesare]

We have implemented a limit on population able to fight. This is more than 1/2 because women, even if they don't fight, can be part of the army (as cooks, whores, or even fighters sometimes). Disbanding a unit will also add population to the square, although it is currently not the full population but only 1/2 to take into account stragglers, deserters (?). I also think it should be the whole figure, though.

[Mark_Everson]

IIRC the proportion of militarily-capable members of the populace is simply a variable that's settable by scenario. At some point this will be replaced by a better model so that if you use up all the combat-capable people replacements will only build back again with time. I don't recall what thread we've discussed that in, but it was fairly recently. The part-time soldier aspects you're talking about Nadexander can IMO be covered by the militia model. There will also be a mobilization/demobilization model, but that's further in the future.

I agree with Laurent that all military manpower should be added to the local economy if the unit is disbanded, rather than the current 50%. But to change it, all that's required is to change a single value in the program, so its nothing we need to settle now.

[Nadexander]

I read some really good articles at www.killology.com about the essentially pschological nature of warfare. Definitely a must read. I made me think that we might want a more morale based combat system.

Due to morale faliure, units will usually break before they lose all their soldiers. When a unit breaks it takes a high number of casualties (killed as they are fleeing) and then the unit either escapes or is disbanded. This allows small high morale high tech troops to defeat large units without killing them off. Also allows two large armies without decisive advantage to stalemate against each other. This is based on the morale system in europa universalis.

[Nadexander]

random other suggestions (just off the top of my head in response to the article. I read the website but im not sure how much of this is already included in the whole CEV thing) :

1) a base moral defined by lots of factors (pay, training, patriotism, effectiveness of command and control)

2) certain tactical actions reduce morale more than others, such as flanking or encirclement

3) certain weapon technologies should be more damaging to moral (scarier) to those that dont have them esp. horses, cannons, machine guns, air power

4) it should be more costly in terms of morale to attack and cheaper to defend esp. towns, forts, and otherwise defensible positions (hill tops)

[Mark_Everson]

Hi Nadexander:

I'm generally in agreement with your comments. The current model already has some morale aspects in it, although they are IIRC rather simple at this point. Units do already break as opposed to fighting to the death. I'll let Laurent comment on exactly what is there right now. The Military Model has some mention of morale, but I don't remember how detailed it is on the topic.

[LDiCesare]

There is a Military coding thread somewhere down there which details how morale is implemented.
Currently what happens is mostly:
When a unit loses a certain %age of its elements (manpower), it decides to flee. It may or may not manage to. When it flees, it changes its orders to flee to the square it came from. This flight is not instantaneous, so it takes some time, during which the unit will probably be attacked again, flee again etc.

Although I agree morale is very important, and things like sarin affect morale more than bullets, I want to stress out the gameplay effects. You do not want an army to flee one square back and chase them without damaging it more than 5% for 20 turns because they keep fleeing. So, an army which flees a lot should disband itself to avoid such tedium. This may be modelled by an additional threshold: Flee if 25% loss, disband if 50% loss for instance. Right now, gameplay feels OK for me, I wait for additional feedback or comments before changing anything here.

[Nadexander]

You do not want an army to flee one square back and chase them without damaging it more than 5% for 20 turns because they keep fleeing.

This would be prevented if:
1) units take heavier casualities as they try to flee
2) units that retreat should have many more desertions and lower morale
3) some hardware (i.e. catapults, seige equipment, supplies, etc.) should be lost while retreating

[LDiCesare]

Your points are good. 1 is already taken into account as units need time to flee from the square. 2 and 3 amount to desertion/loss of equipment. Desertion is not modelled, but it is probably worth adding. That and death because of disease so that armies will slowly disppear all by themselves if you let them out in the open.

[Mark_Everson]

Originally posted by LDiCesare
2 and 3 amount to desertion/loss of equipment. Desertion is not modelled, but it is probably worth adding. That and death because of disease so that armies will slowly disppear all by themselves if you let them out in the open.

Sounds good to me long-term. Perhaps we could do something crude in terms of desertion that can last until there's a good model for desertion and disease. Something involving the terrain type, its capability for generating food, and also distance if any from the civ of the unit.

On another subject Laurent, what is your preference to tackle, naval or siege combat? I think I can work a bit on roughing out a model of sieges relatively quickly if lack of models is a bottleneck at this point. Gary had mentioned IIRC that he wanted to do the naval combat one. Please correct me if I'm wrong.

[LDiCesare]

I'd rather go for sieges than naval combat because I am more knowledgeable about the former, so I can discuss and interpret the model and fit it to a codable form if needed without doing silly things, whereas I don't know naval warfare so I would have to keep asking for clarifications coding my way through the model, which should wait until I am done with my ISP problems (which are well on their way to last one more month the way things are going).
I would also like to have the input I need from the social model in order to make militia more realistic (currently militia will always help the defender, whether or not it is of the same ethny as the people, because I have no way of knowing which civ the population wants to support).

[Mark_Everson]

Originally posted by LDiCesare
I'd rather go for sieges than naval combat because I am more knowledgeable about the former, so I can discuss and interpret the model and fit it to a codable form if needed without doing silly things, whereas I don't know naval warfare so I would have to keep asking for clarifications coding my way through the model, which should wait until I am done with my ISP problems (which are well on their way to last one more month the way things are going).

Sounds good. I'll see what I can do about roughing out a model in collaboration with you this weekend.

I would also like to have the input I need from the social model in order to make militia more realistic (currently militia will always help the defender, whether or not it is of the same ethny as the people, because I have no way of knowing which civ the population wants to support).

Have to wait for Gary's reappearance on that one...