Military Model V

[Mark_Everson]

Hi Laurent:

Thanks for putting the odds back in!

Originally posted by LDiCesare
Sentry means you don't attack

Why? If it is our territory, has an enemy in it, and there is no order to avoid combat I think default sentry activity should be to attack if opportunity presents itself, and the attack would be at reasonable odds.

I can either change all values so that for the moment, fights happen at strength ratios between 0.5 and 2, or we can plug in orders for units now. The first one requires much less code, of course. It can be interesting, though, as it would allow fortify orders, and giving various levels of attack ratios (as in D4). What do you think?

I favor the first one for now. Its simpler, so not much is lost if we decide to do it differently in the future. We should really discuss how orders are going to work in more detail, and come up with a coherent plan before going much futher on orders IMO.

[LDiCesare]

quote:
Originally posted by LDiCesare
Sentry means you don't attack
Why? If it is our territory, has an enemy in it, and there is no order to avoid combat I think default sentry activity should be to attack if opportunity presents itself, and the attack would be at reasonable odds.

To clarify a bit, I consider "attack" when you are not in your territory, and "defend" when you are in your territory.
So currently, a sentry order means you would attack only at 9 vs 1, but defend at 1 vs 1 (I don't remember the figures), and a fortify order would mean attack at 9 vs 1, defend at 1 vs 2.
This means that a unit on sentry, seeing a trespasser, will attack it.

I'll tweak things so that odds are predetermined like 3 vs 2 to attack and 1 vs 2 to defend. Since attackers may not be able to retreat, that should allow fights where defenders have an advantage.

I still have to code the go back to last square stuff. Note that if previous square is occupied by enemies, I suggest the army being surrounded should surrender or disband or something.

[Mark_Everson]

Bonjour Laurent:

Originally posted by LDiCesare
To clarify a bit, I consider "attack" when you are not in your territory, and "defend" when you are in your territory.

I think we will need to modify that definition as things go forward. The relevant factor for figuring a defender bonus for terrain or whatever, is the tactical defender, which is not necessarily the same as the strategic defender. One can be the strategic attacker while the tactical defender by heading for targets that the enemy must prevent you from reaching. Say, important cities. Even though you are strategically on the attack, you will have choice locally of good defensive ground when the enemy does engage. Eventually the military-general equivalent AI will make a decision whether they want to attack or not. Based on what the generals from two sides pick there can be a determination if there is a tactical defender (usually the one with the weaker army, but not always).

So currently, a sentry order means you would attack only at 9 vs 1, but defend at 1 vs 1 (I don't remember the figures), and a fortify order would mean attack at 9 vs 1, defend at 1 vs 2.
This means that a unit on sentry, seeing a trespasser, will attack it.

I think seems to be a reasonable approach in general, but 9:1 seems extremely long odds to use as a requirement to attack. But as you say after the quote it should be smaller.

I still have to code the go back to last square stuff. Note that if previous square is occupied by enemies, I suggest the army being surrounded should surrender or disband or something.

For the short term that sounds ok. But the relative sizes of the forces need to be taken into account. And perhaps making them go back to the exact square they came from is too constrataining. Maybe it should be the square they came from or the two 'flanking squares' of the origin square. The details can be worked out when we get some playtesting in...

[Gary Thomas]

Having settled (lost?) the dispute with Mark over cross-country chariots, I was about to change the terrain xml file to allow chariots to cross mountains at infantry rates, when I realized that the category for chariots is "wheel". But so is the category for truck, bicycle, motor bike, Conestoga, motor car and bullock cart.

A light chariot can be carried, as can a bicycle. But none of the others can.

It rather seems to me that we need a new category of "lightwheel" or "portable" or some such.

Cheers

[Mark_Everson]

Hey Gary, sounds good to me. I'd prefer light wheel or some similar name. Portable sounds confusing to me...

[Gary Thomas]

Unfortunately xml tags have to be single words (because anything after a space might be a parameter), so "light wheel" is out, the reason I threw "portable" in is that it is a single word. The other possiblity is "chariot", but that does violence to the rights of the 2nd Bavarian Bicycle Battalion.

However, so far the choice appears to be between "lightwheel" (one word) or "chariot".

Another possiblity is to introduce another level of category: "light", "medium", "heavy" (and possibly "extraheavy"). As a generic grading this could well have other uses. Just a thought.

Cheers

[Mark_Everson]

Yeah I knew why you had no space. I just thought it was clearer to write it "light wheel". I'm with you that the designations heavy and light could have merit on their own. I'll be happy (for the moment ) whichever way you would like to do it.

[Gary Thomas]

Damn, another decision.

Laurent, if you happen to see this, what do you think?

Cheers

[LDiCesare]

I don't have an opinion on this, thus I said nothing...
Since we must choose, however, here is what I propose:
portable is hard to understand. It applies only to wheeled elements as far as I can think (chariots, jeeps).

The medium/light/heavy categories I can see the use of only for wheels, unless we have mounted instead of horses, and light would be horses, heavy elephants... but even elephants can cross mountains (Hannibal).

So in conclusion, I'd tend to use lightwheel.

[Gary Thomas]

Agreed. "lightwheel" it is.

Cheers

[LDiCesare]

Warfare and economy

Rodrigo made a few comments in the following thread:

http://apolyton.net/forums/showthread.php?s=&threadid=45946

I answered a bit there and more here:
The idea is: Warfare should cost a lot. This is to prevent people (and AI) from being at war at all times.

Rodrigo's proposal:
Rodrigo suggests we could change the timescale in order to adjust to small wars. This means tweaking the non-military models by a scale factor. Thus, we should have a parameter used a bit everywhere which is turn-length. However, I see a glitch with this proposal: If AI1 and AI2 want to wage a war, can they switch the turn length down to 1 month for the player? If the player changes the turn length, how does it affect far-away peaceful AIs? Worse, how does it affect other players in multi-player?
It is not that I don't like the idea, I believe it is hardly applicable in multi player, and may be difficult for the AI to handle.

Another proposal:
Ways to handle that include not having a fully operational army all the time (historically, there were very few standing armies). Looking at the MOO3 specs, I saw they have an active fleet and a reserve fleet. The active army costs more than the reserve army to maintain, the reserve army being in a "void" and not usable or seeable on the map. CtP2 had something a bit similar, in which the army supply cost could be halved, which halved the hit points of units.
I think such a system could be used, but I know that CtP2 scheme doesn't work, because a 1/2 cut is not enough to make a difference. I also think that military units could be mobilised very fast in order to defend their territory.

We could give orders to a unit/TF like "reserve". With this order, the army would have to be in friendly territory, would cost maybe 1/10th of its normal cost to maintain, and be unable to move or receive other orders. They would probably have a defense penalty, but not as severe as 50%. Getting the army out of reserve could take one or two turns.

[Mark_Everson]

Laurent, thanks for moving this discussion here!

On the turn-length thing...

I think it could work with the following prohibitions: It works for Single-player only, and only the player is in charge of the change. I *think* it can be made to work under those circumstances. We will need rules for how to handle varying time scales as we go from antiquity (long turns for everything but military) to modern (short turns). So this is the matter of putting in a switch under player control. If it doesn't work, then we can get rid of it, but it seems to me it should work fine. Since some players like Rodrigo may really value the option, and it should be very little work, it seems to me we should provisionally support it, and only ban it if it turns into a quagmire.

For mobilization/demobilization we already have plans to definitely have it but details are TBD. Searching on "mobilization" or "demobilization" would I hope point to the relevant discussions which could be in either military or econ threads.

[roquijad]

I'm putting here this as Mark suggested (from the "From CIV3 to Clash" thread):

Warfare: defense vs. attack points and terrain
Why in Civ my bronze-age soldiers have the same attack rating a stone-age warrior has? Why my bronze-age soldiers defend greatly but are useless to attack? I don't understand why the Civ series makes such a sharp distinction between units, where some are meant to attack and others to defend. Maybe it's a requirement of the AI they used. I don't know. But I sure know I don't like it. Once I develop bronze working, I want to use it and attack those stone-age neighbors. I don't understand why I only can defend myself better.

In general, IMO, the whole concept of attack and defense is incorrectly handled in the Civ series. Did you see the movie Braveheart? Sure you did. Two armies met at some point in the countryside and fought each other. You can't really tell who's attacking and who's defending. Both armies are attacking and defending at the same time. They are just "fighting".

IMO, there's a problem of scale. We know some type of military, like cavalry, were used to "attack", in the sense they were more effective when launched against the enemy than waiting for the enemy to come to them. But that's true only at the battlefield level. When the player moves a unit to other square to attack an enemy, his troops are traveling hundreds of km's to meet the enemy and start a battle. The map isn't a battlefield. It is at a higher scale. If someone attacks my cavalry unit on the map, that doesn't mean my men have to stupidly see the enemy come from afar and decide not to charge. The cavalry should attack/charge them once the enemy has come in range. On the game map, my cavalry should have the same strengths (in the general sense of the word) no matter if it's attacking or repelling an attack.

IMO all (or at least most) units must have the same attack and defense rates. I bet some of you would refuse this arguing it'd be boring because you wouldn't have to make "tactical" decisions (attack with the cavalry first, then send the legions, etc), but in Clash, really, that type of tactics is already dead once we chose to use task forces. (at least that's what I expect if the task force sums up all strengths from all units)

The important tactical decision is, like IRL, where the battle takes place, due to the relation between different types of units and the terrain. In the Civ series you can use your cavalry to attack a unit in a jungle..... give me a break! The cavalry shouldn't be of any use attacking a walled city either. Nor defending it! What's my cavalry inside the walled city supposed to do if the city is being attacked? Open the door to attack them outside? No way, leave that door closed! Shouldn't also cavalry be more effective in plains than in a forest?

The Civ series doesn't treat terrain correctly IMO. The terrain defensive bonus only covers some situations. I think Clash must do better than Civ in this regard.

Finally, note that fortifying a unit is a special relation between a unit and terrain. When you press "f" what you're saying is "take a defensive stand, taking advantage of the terrain". IMO that should be the only case where a terrain defensive bonus can be applied. If the attacked unit isn't fortified, then both attacker and defender are in the same type of terrain (it's the place where the battle takes place) with no advantage for neither of them, so terrain shouldn't give a bonus to the defender (remember the issue about scale).

LDiCesare answered this way:

There is no such distinction in Clash. There is a bonus to defense in terms of:
1)Scouting phase bonus,
2)We will have fortify orders to increase defense.
Mixed warfare is achieved through an advantage of having high mobility units (cavalry), long range weapons for fire support (archers), and units that have strong attack values (infantry).
Terrain defensive effect depends on the type of units. In the code, currently infantry is advantaged in rough terrain. Ultimately, each terrain/unit type will yield different bonus/malus.

That is a big relief for me. Thanks LDiCesare!

[LDiCesare]

I still have a problem with odds to attack. I coded it so that the attacker will attack at 1.5 vs 1 or better, and the defender will defend up to 1 vs 2. Even if a party doesn't always manage to disengage, the result is boring as you attack at 3 vs 1, deall little damage, then retreat a bit, attack with 2 thirds of your strengths, deal good damage, but then the defender is so weak they always flee and you never manage to get rid of them.
Is it enough if attackers attack only if they consider they have a good strength and defenders always defend? It may not be good form an AI or historic point of view, but gameplay-wise, it would help a lot: You attack a square only if you are numerous enough, and you don't have to run around smallish armies that escaped total destruction but managed to survive due to their small numbers.

On another topic, it may be soon needed to open yet another thread for the military model.

[Mark_Everson]

Originally posted by LDiCesare
Is it enough if attackers attack only if they consider they have a good strength and defenders always defend? It may not be good form an AI or historic point of view, but gameplay-wise, it would help a lot: You attack a square only if you are numerous enough, and you don't have to run around smallish armies that escaped total destruction but managed to survive due to their small numbers.

Hi Laurent, your solution sounds fine to me for the time being. I'd like to have good ways to handle defender retreats at some point, but there are more pressing things to get done for now!