Military Model V

[LDiCesare]

You mentioned that units fight at one range and then move closer. What if it is to their advantage not to move closer? If a unit has long-range firepower and more mobility than its opponent, it seems like it keep shooting at that range until it runs out of ammunition.

Yes, the model is a bit vague here. It supposes the attacker always wants to reach melee, which is often wrong in modern times/long-range weapons. Since there is a test to advance at next stage, the attacker, if it won the manoeuver phase (has better mobility) should be able to decide whether it wants to get closer. We should also check if both attacker and defender want to remain far range. Basically, if they have better long-range firepower than short-range, they would stay away, but that doesn't work well when you want to seize fortifications / trenches...

[Gary Thomas]

The annihilation of the legions at Carrhae has been attributed to the fact that the Parthian general, Jaselin (or some such name, I am going from memory), loaded 10,000 camels (surely an exaggeration) with spare arrows, so when the horse archers had shot off all their arrows they could go and get some more. The legions were stuck in an open plain without shelter. When the horse archers (who were armed with composite bows) attacked, all the Romans could do was hide behind their shields. The casualties and lack of water gradually wore them down.

So, generalship in the sense of providing spare ammunition. Equipment in that the horses and composite bows gave the Parthians an advantage on the specific terrain. Probably, however, the deciding factor was not good generalship, but bad generalship since Crassus should not have got himself in that position in the first place.

Crassus was the richest man in Rome. The Parthians captured him and executed him by pouring molten gold down his throat.

Cheers.

[LDiCesare]

The innards if fights

I have been preparing something for D7, which is fight at different ranges. I describe what Krenske's model originally proposed:
Fight at initial range (long/mis/short,maybe determined in actual lengths), and advance at end of a round if morale checks are made.
I coded that, so that archers can be used not only as support fire (although they better be if the opponent closes in).
The current model and code are not very deadly each round. It takes a lot of rounds to seriously harm anything. This means fighting one or 3 rounds at a long distance before closing in makes little difference on the end results (perhaps that managed to make a few opponents flee, at most). The model also fails in that long-distance armies will not want to close.
I will code this for soon:
Early Distance rounds. If all armies would rather fight at closer range, proceed as currently in the model. If a defending army is in a city, only allow it to want to get closer (they won't leave the city undefended so they can better shoot at opponents sacking it). If an army tries to close in and the other doesn't want to, check the chance of success based on manoeuver output.
I am also tweaking the morale code every way round so it is more accurate. In D6, managing flight is sometimes impossible, where I wanted it to be difficult.

[Mark_Everson]

Sounds good Laurent!

Some other thoughts for the near future:
Is defender advantage in yet, If not there should be a minimal advantage in combat outcomes from just defending, something like 30%?

In general we should soon get in some terrain defensive effects and defensive posture effects. You can see some examples in tables about 3/4 of the way down my old military proposal page.

There may already be proposals for this stuff in Krenske's model or our later discussions, but if there's anything concrete I can't remember it!

[LDiCesare]

Currently, there is a bonus for the defender in the manoeuvre phase. It is arounf 15%, and then it is scaled down since the manoeuvre phase output gives only a small bonus. It can probably be increased.
I didn't manage to reach your page. What I mostly need for terrain is figures. Terrain effects should modtly lower attack values, and lower mobility. I will try to check the page again when my internet connection feels more like it.

[LDiCesare]

Defense and terrain bonus

I managed to read the page, and checked the code and some fight results.
The current situation is the defender has 15 bonus to scout phase, which roughly translates into a +4% bonus after manoeuvre, everything else being equal, due to knowledge of terrain alone. I think a defense bonus should be added only in cities, since in the open it is more a question of manoeuvering. One should note that, if they have engineers (only legion unit has any now), the side that wins the manoeuvre phase gets a defense bonus due to the engineers (+1 per engineer - needs tuning, but that will wait).

I suggest a defense bonus be given only if there is a city (I suppose a good-sized city has defenses, defenders have better morale, etc.), and 30% as you suggest sounds good. It could be interesting to find out if there are city walls, ditches, forts, etc. to give additional bonus. These are different from field defense built during battle in the sense they take months or years to build, thus are really infrastructure. Considering just the code, I suggest a 30% bonus if in a city, and nothing else.

Terrain is divided in several subcategories. I'll add a defenseFactor: Defense is multiplied by (1+defFactor), and mobility divided by (1+defFactor) for infantry, (1+2*defFactor) for other units. Mobility is only useful for scouting, manoeuvre phases and morale/fleeing so it is probably not enough a penalty for mounted units. I may also decrease attack for non infantry by dividing by (1+defFactor), based on test results.
Here are suggested defFactor values, additive:
Rolling, Broken, Massif: 0.25,0.5,0.75
Forest, Jungle, Dune, Swamp: 0.15,0.3,0.3,0.5
I don't think soil type is important.

[Mark_Everson]

Hi Laurent:

I think a defensive bonus to combat power is assumed to be important by virtually everyone in the military gaming industry. Such a bonus simulates that the tactical defender chooses the terrain on which to fight, and can dig in within some limitations. A defensive combat power bonus has also been corroborated by the regression analyses carried out by Dupuy and co. that are referred to on that web page. I believe it is a Really Important thing that we can't miss.

Perhaps using Krenske's system it should be a bonus to the defense characteristic of the tactical defenders in a battle. Anyway, I believe a power multiplier of about 1.25 or so for a defender Anywhere is appropriate, and it should be larger if the defender has substantial time to dig in, is as you say in a city, and clearly in a fortification.

Since Gary was foolish enough to sign up as mil model lead, I guess we should see what he has to say on the topic .

The terrain stuff looks reasonable to me for now. We can always tweak the exact values...

Cya,

Mark

[LDiCesare]

I decided that it would be easier to code if, to simulate defense bonus, I decreased attack by the ratio between optimal mobility and actual mobility: So f.e. a cavalry with attack 80 def 30 against infantry attack 80 def 30 we get:
flat: 80/30 vs 80/30
swamp: 40/30 vs 53.3/30: The infantry gets a bonus in rugged terrain.

Using KRenske's model, I think I'll propagate the defense bonus of 15% in scout phase to the manoeuvre phase and test what the results are. Should become a final figure around 20%, more or less depending on the troops mobility.

[LDiCesare]

I put a few features in the code for D7. Here they are (beyond bug fixes):
Fight at different ranges:
Krenske's model provided fight at various ranges (5 or 6, I don't remember). I have a few rounds at long, medium and short range before fighting melee. Advance between ranges depends on whether the armies want to advance (archers usually don't), and their mobility, defenders getting a small bonus. This allows to simulate properly javelineers who send their weapons before closing in.
Here is how it works: Fight starts at long range. At the end of one round of fight, morale is checked, and armies who want to advance (based on whether they have better firepower at shorter range) advance by a step equal to mobility. When armies have advanced enough, fight gets on at medium range, then short, and last melee. For now, the attack strength is binary: distance/melee. The number of rounds at each range and attack strengths will have to be playtested. When adding more elements with short or long range capability (javelineers, slingers, catapults), I'll cut down the distance strength into more categories.

I also added defense bonus for terrain. Currently, all units benefit from it. Its main effect is to reduce offensive power of mobile units more than that of infantry. I can square the penalty for attackers to increase the defensive value of mountains, hills etc. The figures are as given in my previous post.

I also put in a bonus for defenders based on ownership of terrain. It gives an 80% chance to win a fight for the defenders if they have the same numbers as their attackers. Additional bonuses based on the orders (sentry, forced march, etc.) are already in the model, although there is currently no interactive way to give such orders to the units.

Last, the tech/military hookup seems to work, as I gave myself an exponential tech progression in Military Tactics, resulting in monstrous light cavalry units. I still have to be able to say I can build units based on tech requirements. I don't know if the tech model itself is interested in knowing it, so right now I'll code it as methods in UnitArchetype/ElementArchetype.

[Mark_Everson]

Hi Laurent:

Thanks for the info! This is good stuff. I tried out your recent mods and they seem to work pretty well. The defensive bonus is especially good in terms of gameplay since it weakens the current AI strategy of just sending a stream of single units. Its now more like IMO it is in the real world. Two units on defense can now hold fairly well against a string of single-unit attacks. Of course the AI will get smarter soon too!

I think at some point it would be really great to get some simple descriptive info on what won the battle. That way the player can get the feel for what is going on under the hood of the model! I'm not sure about a good way to implement it... Something that said that "Withering arrow fire from our archers" won the battle would come up if the battle was half won before troops even reached melee range would be a good start. I wouldn't put this before 'hard' features like ship boarding/ fighting and tech, but when you've gone as far with those as you can (you may be there already) a simple "battle description" seems a good feature to me. What do others think?

On terrain defensive bonuses, I don't think it should apply Nearly equally. Usually the defender gets much more of the terrain bonus. If, as defender, I'm already on Top of the hill, I don't Need to climb it in combat as the attacker will have to ... I'm not sure about "squaring" the modifier since I don't know what the numbers you're using are. Maybe something like 1/3 of the current terrain bonus could go to the attacker, and the whole amount to the defender. Traditionally in most wargames the defender gets the entire bonus, and attacker gets nothing. Since Gary is the current model lead we should see what he's got to say...

Great Job!

-Mark

[Richard Bruns]

Originally posted by Mark_Everson
I think at some point it would be really great to get some simple descriptive info on what won the battle. That way the player can get the feel for what is going on under the hood of the model! I'm not sure about a good way to implement it... Something that said that "Withering arrow fire from our archers" won the battle would come up if the battle was half won before troops even reached melee range would be a good start. I wouldn't put this before 'hard' features like ship boarding/ fighting and tech, but when you've gone as far with those as you can (you may be there already) a simple "battle description" seems a good feature to me. What do others think?

I think that this should be included and given top priority. I think there should be, at least for now, a popup with a detailed accvount of every battle. I can already tell that the game mechanics will lead to a few monumental battles. Players will build up huge task forces, which will meet in decisive engagements. This is a good thing, and accurate to the anciant world. But if no account of a massive battle is given other than "The romans won" it will be really anticlimactic.

The user interface is the first impression someone gets, and first impressions of games are vitally important. When I played I had no clue what was going on with the military. The game just didn't tell me. I just made heaps of legions and bowmen and sent them to enemy cities when the power circle looked big enough.

I think that there should be some kind of in-game advisor or tutorial, at least to tell the player what units are best for what uses. I had no clue what was best to build.

Originally posted by LDiCesare
Additional bonuses based on the orders (sentry, forced march, etc.) are already in the model, although there is currently no interactive way to give such orders to the units.

If they are already on the model, why don't we add them to the interface? It would make the game a lot better.

It seems that the interface is really lagging behind the rest of the game. We have great models, but the player can't interact with them well. This is just one example.

On a different note, I think we should limit the size of the task forces. In order to prevent abuse of stacking, the Logistics technology level should limit the size of task forces. As a very simple model, can we say that the top size of a TF is limited to the level of logistics tech? So if you have a tech level of 12 you can have a TF of 12 units. Does that sound good? It should prevent mega-armies form being built, which is good for play balance and more realistic.

[Mark_Everson]

You've got some good points, Richard. The tutorial idea has already been put forward by Gary, and is in the D7 & 8 Planning list. I need to think about some of them a bit more. But there is one response I can give right away...

Originally posted by Richard Bruns
On a different note, I think we should limit the size of the task forces. In order to prevent abuse of stacking, the Logistics technology level should limit the size of task forces. As a very simple model, can we say that the top size of a TF is limited to the level of logistics tech? So if you have a tech level of 12 you can have a TF of 12 units. Does that sound good? It should prevent mega-armies form being built, which is good for play balance and more realistic.

TF size will be limited by Game behavior, not arbitrary rules! Fairly soon, when military quartermasters come into the game, there will be a Natural limit on how many units you can keep in the field. It will depend on infrastructure and Logistics techniques. It won't be fixed though, but just depend on features like how much money you're willing to spend, where you want to go, etc. I would like to get that part working first and see how it goes. If needed we could throw in an arbitrary rule, but I don't want to confuse people by putting in too many temporary things.

[Richard Bruns]

There is nothing arbitrary about the rule I suggested. Keeping big armies in the field is a constant logistical headache, and there was a limit to how biug armies could be. The logistics tech would be a natural way to reflect this.

But I didn't know you had a more detailed system on the way. Where can I read about it?

[sas]

I don't think we should limit the size of taskforces. Although it should be some sort of "penalty" for making really big forces. An increased cost for armies above your tech level is perhaps one way to meet the problem. There are historically basis for not limiting the size of armies, but the consequences should be just as damaging. Hint: The Huns (As in Atilla)

battle description sounds great. It would really make the game feel more realistic and fun at the same time. Perhaps a small report for small battles and a longer report for bigger battles ?

[axi]

As a very simple model, can we say that the top size of a TF is limited to the level of logistics tech? So if you have a tech level of 12 you can have a TF of 12 units. Does that sound good? It should prevent mega-armies form being built, which is good for play balance and more realistic.

On this issue, see how Europa Universalis handles it. Attrition is the key. Army health should drop, as army passes through inhospitable lands. Armies used to be very expandable in the past.