Mil Model discussion thread

[Harlikwin]

Ok. This is to be the "chat room" for the military dicussion.

[Mark_Everson]

Kewl:
I am working up a first post

[Harlikwin]

Ok kewl, Here is my agenda of stuff to discuss...

Range effects modeling in the combat resolution model. (Other than what I have posted already (Cuz I really don't like it))

Supplies and forage

Local recruiting of levies from enemy territory (Kinda ties into supply)

[Mark_Everson]

M1

Warning... I am using my dictation software to produce these posts. This can result in really weird typo is. If I say something you don't understand trying substituting a "sounds like" phrase for the thing that looks really bizarre, and sometimes you can figure out the meaning.

There are several big issues that I would like to address. I'm sure you all have your own lists... As I said before mine are more the things leading to combat in a specific square, and related to it, rather than the actual combat that happens in the single square. However, I think getting basic agreement on these things leading to combat is Essential. I think the actual combat/battle model can be polished in a lot of ways after these decisions are laid down. Before we've reached complete agreement on them, I think it's possible to end up with a combat system that makes a lot of sense in the single square, but doesn't connect well to the rest of the game.

The basic areas I think are important are:
1. Establishing a link between movement outside of battle and any battles that may be happening.
2. The basic way units are built, maintain, and supplied (not a lot of detail, just a general outlines)

1a. Strategic movement rates... the player should have the choice of moving their units at a variety of rates. The maximum rate will be set by things like unit composition, road and bridges infrastructure, prepositions supplies, etc. However, ratcheting down from the maximum possible movement rate there should be a variety of choices (we have talked about this before) from forced march (maximum) down too cautious advance. Whether a unit can be surprised, or put in a disadvantage in a coming battle should be affected by these movement choices. We also need to figure out the connection between the movement rate on the main map, and Harli's scouting phase. If we end up using phases (ticks) the distance traveled per tick will depend on the basic movement speeds selected.

1b. Connection between outside units and battles. Basically what I'm talking about here is the logical extension of the simple support mechanism in demo 4. For one thing, a real modern battle will sometimes take place in an area significantly larger than a single square. We need to have rules for how a TF can join a contest in a single square. For instance in a large fluid engagement like the battle of Kursk... overtime during the engagements (during a single turn IMO) units should be able to enter the contest in adjacent squares. For instance, I can see nothing so irritating to the the player as having a large task force that annihilated the enemy in its square (say, during the scouting phase) being unable to Ever make it to help in the closely-fought battle sequence immediately adjacent until a whole other turned goes by. Such a series of battles, according to Harli is model could go for many battles, clearly plenty of time for some armor that is only 50 miles away to enter the fight at some point later than the initial part of the first engagement.

1c. Another thing we need to nail down is whether task force movement is simply square-by-square, or more continuous. A continuous movement system, or you keep track of Where the unit is within each square is clearly more complicated, but has some advantages. The main advantage is that you can capture a different movement rates much better. For instance in civ, you can only have strategic movement differences that are very large in terms of movement rates, because you're either move a whole square or you don't. So we need to discuss whether we want to use continuous or by-square movement (before battle)

2. We've already discussed building units, and supplied somewhat. I don't think the issues here are as big as in the movement regime. It seems we're agreed that building units requires real amounts of population, and combining them with weapons, training, and possibly doctrine. In terms of maintenance and supply, I think Harli's simplified system presented before should be given a shot. At some point we will have the merchant coding functional, and see how difficult it would be to do Real supply. But that may be too computationally expensive. There will also be a variety of mobilization levels I think, and we can work out the details on exactly what these are. There has already been some mention of this in the economics model in the miscellaneous section. But I'm sure we can work up something better and a few quick ideas that are there.

Okay, I just read the things you want to discuss, and am waiting for your initial positions.

[Harlikwin]

Ok, For my benefit I NEED to know how big one square is. I do agree on your points in terms of movment in terms of getting there effects how a battle is fought. Howver I want to clarfy a few things

No force could get aniahliated during the Scout phase, technicaly all battles would go up to the the assulat phase, or if a withdrawl is ordered to the witdrawl phase or if that fails to the pursuit phase.

Also I really don't buy the idea of armies abmusing each other. On a smaller scale it happens but not a big army level. The scout phase reflects the commander trying to pin down exactly where the enemy is so that they can deply their forces in the best manner to deal with the enemy (the manouver phase)

I'm not sure how to handle Kursk style engagments other than to say that they should just be fough as individual battles if they are in seperate squares... Perhaps we could use something like in EIA where there is a possiblity based on the commander of reinfrcing an adjecent battle or something...

as for the continous moment idea it is cool, and so far I have assumed square by square movment. It would force me to reevalute the comat model in terms of it but I don't think that it would be a huge proble. Maily I would have to design a pre-battle section

-Harli

[Harlikwin]

Issues to address


Resolve range issue


Problem: don't want the range thing to turn into a tactical combat system.
Solutions:
1. Roll it into one thing
Use tactics, mobility rating and ranged firepower rating rolled into one to determine casualties, also use commitment/aggressiveness level

Range effects,
Depends on several factors, local tactics used and terrain mainly

Local tactics
Defend (dig in/hold position and shoot as much as possible)
Assault (Charge the enemy and close to hand to hand range)
Maneuver (Try to maintain fixed distance from enemy while firing)

So if you are defending infantry versus an assaulting cavalry charge you get X number of shots
If you are a Mongol horse archer attacking Knights, you would use your advantage of mobility to use indirect fire on them, and get X shots before you withdraw

These things are modified by terrain.
Also mobility is also affected by terrain

Casualties per element per phase should be on the order of ~10% max for the 1st 2 phases in modern times.
It should be noted that almost all heavy casualties and destruction of armies historically occurs in what I term the pursuit phase, i.e. the phase where the enemy's morale is broken and they are retreating. If the battle goes to this phase (rather than the withdrawal phase) then heavy casualties should be expected. The main difference in modern times is that usually the battle will proceed to the withdrawal phase and thus avoid heavy casualties

This system should be able to model things such as: The steppe horse archers, which no one really found a counter to, only other horse archers. The advent of the long range rifle in the civil war that put the rifle as the premier defensive weapon. The reverse of this in the first world in terms of artillery (note: using bombardment at the end of the scout and maneuver phases reflects this IMO, just as long as you have enough arty to do it)

Ex: In the civil war, primitive entrenchment's put the defensive infantry in a position of primacy on the battlefield. This effect would be exacerbated stuff like bolt action rifles.


2. Use my pre proposed system of movement and fire along the 1D range band with multiple resolutions.
Problems: This seems awfully tactical to me, and in those terms really unrealistic since it is 1D…

Supply
We need to also model foraging as a function of civs? Or during different time periods or as af unction of tech. (well certainly as a function of tech)
Also I want to take a look at recruiting local troops to fight for an army that is far from home, Historically this happened during roman times (2nd Punic war; Hannibal recruited Italians in Italy to fight for him against Rome.), and medieval times.

Harli

[Mark_Everson]

M2 (an initial and number will make it easier for us to reference which post of the other we are talking about)

I'm trying to look over the old models for range...

I just saw you're newer post, but haven't really had chance to read or respond to them. In terms of square size, for now a square is assumed to be 60 miles (100 kilometers) on the side. It's possible they might be twice as big is that were so if it turns out there are just too many squares for us to handle in the map, which is currently 320x 200.

While doing that one more important thing occurred to me. We need to lock in some agreement on the relative movement rates of "military time scale" vs. the "chronological time scale". As I said before, I don't think we can really match them do to gameplay considerations. Is this okay by you Harli? I can certainly explain it in more detail, but I think I've already gone into it, and you may perhaps agree already, in which case there's no point in discussing further. I think it would be attractive to have the difference between the military and chronological timescales get more realistic as things become more modern. So we don't have a WWII-style conflict lasting hundred years, but 20 might be acceptable.

[Harlikwin]

RE: M2

Um as to the time scale thing, It has been dicussed but I'm not sure what the actual concensus was. From what I heard it all sounded reasonable... I was knd of partial to dividing each yearinto 12 months and letting the computer resolve that kind of stuff but the other things I heard were pretty good. Actually in terms of that, 12 months would be fine for modern wars, but I think that a 6 month turn would be fine for ancient and something like 4 3 month turns would be ok for say napoleonic... Again this would all be handeled by the computer so that the player wouldn't really be involved.

-Harli

[Mark_Everson]

M3

Army's Ambushing...
it's not exactly army's ambushing each other, but obtaining relative advantage leading into the scouting phase. If an Army screams in at maximum rate not knowing much about the terrain it is going to be at a severe disadvantage to someone who has had a chance to already scout the ground etc.. But anyway, I don't think this is a critical thing for us to worry about right at the moment.

Kursk...
I think armies can have fairly straightforward pre-programmed orders like: first priority is achieve objective of taking x square. If that succeeds, support friendly troops as much as possible... this sounds kind of like the system you are citing which I'm not familiar with.

Continuous movement...
I think it has a lot of potential advantages, but also some disadvantages in terms of complexity. After you thought about it a bit, let me know which way you think we should go on it.

Range effects for combat within a square...
your model number one looks better to me to at the moment. I think it is Quite sufficiently complex. If we can get the interaction of terrain, mobility, and range more or less correct, we will already be going Way beyond what is available in the genre. Also I think a system that doesn't produce huge surprises (which range proposal 2 might do depending on troop placement) will be a big advantage for the AI.

Supply...
yes I think we definitely need to model foraging, especially in earlier time periods. Probably there will just be a movement penalty associated with it. Recruiting local troops is good, but I think that's almost more diplomacy than military. I think we should clearly have the possibility of that.

[korn469]

is anyone still here? i had problems getting on the internet this morning

[Harlikwin]

Re M3

Ok, I can easily insert a modifier based on the amount of leftover movment (in using soemthing like the tic system) into the scouting phase (one of the beauties of the system, as many modifiers as you want ) So thats no biggie...

Kursk I agree that the AI should be intellignet enough to deal with large battles like that. Also the time scale for one battle in one hex I currently have is about a day for ancient -- napoleonic times and maybe half a day or less for modern times. Like I pointeed out eralier more than one battle per square is possible, it would take a few days. I think this is pretty realisitc considering the square scale. (since modern warfare is pretty mobile forces would give ground more readily wheras napoleonic armies would fight for a few days in the same square)

Well the thing is I'm trying to avoid a 1D tactical system like I initally proposed. Actually I think a divsion between direct fire and bombardment weapons also needs to be made, it is refleced in the 1-10 scale that kerense uses but Ithink we can just mash it down to 3 catagories like I had originally and just try to model the effects of longer range weapons.

As for the local troops: I agree that is diplomacy but I think it should be readily avialble in the premodern era, since it happened pretty regularly. Esp into the napoloenoic era using the idea of using "wild" troops i.e. zouaves and the like.

Also are there any ideas about mercnearies currently?

I'll let you know about the continous move model though to be conservative at this point and say that it might be better to stay with a more traditional model....

-Harli

[Harlikwin]

yeah korn just hop on in...

[korn469]

ok harli

i'm typing up some responces right now

[Mark_Everson]

M3:

On timescales...

I am going to be one of the guys doing the AI. Your and Krenske's alternative of having the computer run things for 12 military mini-turns during a year is simply not going to work IMO. Even a brilliant computer AI will not be able to do what the player would want to do it each juncture, it's just impossible. Way too much can happen in the year. The player can't even give orders that far enough ahead because so much can happen in the year. General orders based on the success of an initial operational level flight, can easily be invalidated if the fight turns out to be a disaster, or a great success. To model where the AI handles it would trundle along with the same old orders for a whole year before the player got a chance to respond. IMO, in order to get this one real world detail right, you'll end up with endless player frustration. I think we just have to sacrifice reality when reality conflict with fun.

So, could we let the players actually run all the turns that you envision?
I think your proposal for how long the military turns might be it is reasonable, but if you just add up the number of turns that would require, it is Enormous.
Ancient (six months turns) if we do about 6000 years over ancient history, that's 12,000 turns
Napoleonic (4 turns per year) over a 100 year period That's 400 more turns
modern (12 turns per year) over 100 year period That's 1200 more turns.
All of civ is only 400 or 500 turns!

IMO you need to accept the fact that military actions are Not going to take the right amount of time. It's an abstraction that makes the whole idea of a civ-type game work.

Now what you say is Very doable four scenarios. In that case we could easily have seasonal or monthly turns without any problem, and the player could direct each monthly turn if necessary.

[korn469]

ok about the 12 mini-turns, with ticks it's not 12 mini-turns it's just that each unit have 12 "action points in a turn"

different actions cost a different amount of action points...and each unit has an order queue, like a waypoint list in starcraft