Clash Combat system, a quick sketch.

[jacobo]

Would a battle last until one army group died? Or would it end after a specified number of casualities, or until one side surrendered/retreated?

[Mark_Everson]

Hi jacobo:

When a fight ends depends on the missions of the armies involved. Armies will have missions determined by the player or AI. For instance, the mission on a given front might be to "destroy enemy army" or take X strategic city/territory, or "make a pain of yourselves till they buy you off". Defensive orders might be defend at all costs, roll-back in order with a scorched earth policy, or 'get the H out of there'

So it depends on the missions of the two armies and their relitive forces. An army bent on the complete destruction of the other will pursue if fled from to the extent supply allows. Armies that are clearly outgunned will tend to retreat to fortifications or sue for peace at whatever cost the civ thinks is worth it, or 'roll the dice' and try a battle.

The tactical combat system will include modifiers for things like 'defender withdrawing in order, and a bunch of other things.

-Mark

[Mark_Everson]

Clash Combat system, a quick sketch.


By Mark Everson & Andrew Warwick

My thoughts so far on modifications from the military / combat system in civ are that Clash will have:

Simultaneous Movement
Army Groups
More Strategic combat system / combined arms…
Good AI

Good AI is really a different subject from the game design, which is what I want to get at here. I'll talk about that in a different
thread…

Simultaneous movement is good both for practical reasons (in multiplayer everyone can do turns at once) and "stylistic"
ones (you get effective "fog of war" without special rules). I think simultaneous movement gives a lot of good feel to the game with a relatively small amount of resulting problems. One thing it does require is a "support" order system so that one
army / column can support another "automatically" when it is attacked (a unit/army could also have orders to support/protect a square its not in). IMHO we don't want the player always questioned about "mid-move" issues (I would not allow the player to modify their movement slightly before it is implemented; that seems to me more of a field commander call rather than
overall strategy). However, some way to include "new" information gotten from scouts during actual movement would need to be included to make the automatic last-minute corrections that a field commander would.

An Army Group (name and writeup by Andrew W.) is a collection of units, built individually and combined into one force. They have the speed of the slowest unit, and an overall strength rating, plus special abilities gained from various unit types and technology. (Ie cavalry gives you a 'shock attack' bonus, artillery gives bonuses against defenses, bombers [if included]
might ignore ground defenses and so on.) Individual units can be removed from the group and added at will. For example a
civ might have Army Group 1 with the following units.
2 x Cavlary
4 x Phalanx
3 X Skirmishers
etc
An Army Group with only 1 or 2 types of units would be somewhat limited from a "combined arms" standpoint as some units are more effective against some, and weaker against others. Diversity is generally needed to make a successful Army Group. This would lead to more realistic handling of armies, and speed up game play, since you don't have to move dozens of units individually.

Enemy Army Groups can only be spotted if they are adjacent to one of your AGs or squares, or be scouting (spies, planes, satellites etc). The contents of the AGs can only be found out during battles or by spies (though sometimes they may get it wrong...) If AGs meet while they are moving, they battle, and depending on the results, may be able to continue their projected movement.
During and after combat, individual units within the AG will take damage and/or give ground based on their individual roles in the combat. (And of course with a heavy random element). For instance in Napoleonic warfare troops that give a "shock attack" advantage to the army would tend to be heavily damaged in "even" battles. In even battles that are lost, those shock troops would sustain Very heavy losses.

BTW I (Mark) found a useful reference on a potential strategic combat system in "Understanding War" and "Numbers, Prediction, and War", both by Dupuy. You
might be able to find copies at a university library.

I guess my proposal at this stage would be to try the idea of a simple system that has army groups with simultaneous movement, good AI to move them for you, and support orders. We could then see if we need the military vs strategic turn setup (see Telescoping Time Scales thread).

[Mark_Everson]

An Aside...

There's an interesting article on game balance in combat systems, in regards to Civ:CTP at http://apolyton.net/ctp/devcor/1_joyofcombat.shtml . Its definitely worth a read.

They found that giving Armor its real striking power imbalanced CTP. I think our more reaslistic technology model (includes diffusion of technology, making it harder to keep a tech to yourself for a long time) and better AI will mean our system can use realistic combat powers without screwing up the game balance.

[Hrafnkell]

I don´t know how combat will be computed (has it been decided yet), but I stumbled upon a neat idea in a book by Frederick William Lanchaster from 1916 titled "Aircraft in Warfare: The Dawn of the Fourth Arm." In it he puts forth an idea of his he calls 'The N-Square Law'. He writes:
"The fighting strength of two forces are equal when the square of the numerical strength multiplied by the fighting value of the individual units are equal. Therefore the figthing strength of a force may be broadly defined as proportional to the square multiplied by the fighting value of its individual units.
As an example of the above let us assume an army of 50,000 giving battle in turn to two armies of 40,000 and 30,000 respectively, equally well armed; since (50,000)2 [this is a small 2]=(40,000)2+(30,000)2. If, on the other hand, the two smaller armies are given time to effect a junction, then the army of 50,000 will be overwhelmed, for the fighting strength of the opposing force, 70,000, is no longer equal, but is, in fact, nearly twice as great - namely, in the relation of 39 to 25."
I don´t know if this is something we can use, but here it is anyway.

[Mark_Everson]

Yes, Lanchester's Law is a good start. I stumbled upon it a few years ago. The books by Dupuy I'm always going on about have a modified version of the Lanchester "square laws" in them. I am formulating a system based on the Dupuy books, and hope to have something written up by maybe Saturday.

The concept of the Lanchester law also applies to alliances in power politics. Consider countries A,B,C with A having power 10, B 8 and C 6. Clearly if war is the only option B and C must ally against A or they are doomed. But C must leave the alliance with B at a certain point during the destruction of A or it will be doomed to fight a more-powerful B when A is gone.

[F_Smith]

Personally, I think the rule is not a good one. The fighting power of an army group of 2 seperate groups on the attack is often far greater than that of 1 large group of the same size -- esp as you get into the modern era of mechanized armies.

Mobility is a much bigger factor in combat strength than most traditional analysts ever gave credit for. That is one of the biggest things you can learn from a 'real-time' tactical game like AoE or Starcraft. If the one big group is defending, then it can more easily be outflanked by the two smaller groups, thus is at a disadvantage. If the one big group is on the offensive, then the two smaller groups can work in a coordinated fashion, one providing a screening force while the other chooses the ground for best combat advantage (or retreats!).

That was Napolean's big breakthru, I think, and the beginning of the understanding of modern warfare by Gustavus Adolphus. Break big armies up into smaller operational units, then focus on coordinated use of those units. After a certain point, army groups become too big to effectively deploy in anything besides a static defensive position.

[Mark_Everson]

Lanchester's Law really does make sense in an abstract way.

An Example:

This is using the Lanchester Square Law, which assumes aimed fire.

There are guys with guns standing on two sides of an open field, fighting. Everything is equal between the sides. (Never is in real combat) The only difference is that side A has 30 men, side B 20. Lets assume there is a 10% chance that any guy from side A will hit a guy from side B and vice versa. The combat looks like this:
A___B
30 ,20 At the start.
28 ,17 After one round of shots 30A guys kill 3 (30x0.1) of B, 20B guys kill 2 of A
26 ,14 Rounding nearest, so 17x10% -> 2
25 ,12 round .5 ones down
24 ,10
23 ,8
22 ,6
21 ,4
21 ,2
21 ,0 So after nine rounds its all done.

What does Lanchester's Law Predict?
(30^2 - 20^2)^0.5 = 500^0.5 = 22.4

I would've gotten closer to 22.4 if I hadn't rounded.

Class dismissed


[This message has been edited by Mark_Everson (edited May 06, 1999).]

[Mark_Everson]

You are correct If the two small separate groups can be effectively coordinated. But then they are in the same battle and would count together. And in fact one of them would probably get a "flank bonus" in a field battle. What Hrafnkell and I are talking about is if the two forces are engaged in separate battles. If the bigger force can break the coordination between the two smaller forces, and engage them separately. Napoleon was great at this.

Mobility counts, both strategic and tactical mobility. That'll be in the system.

Spreading units out a la Napoleon is great for operational reasons, but only practical if the disparate forces can be pulled together for a battle. Otherwise all the small forces will be beaten in turn by the larger.

I think everything I've said is consistent with current military doctrine. But of course I'm too lazy to give all the caveats

[Hrafnkell]

Yes, mobility is one thing, communications is quite another. As we´re talking about Napoleon it was one of his favorite tactics to place his army between the armies of his enemies and beat them in turn. Even if the enemy armies where quite close they didn´t know what Nappy was doing until too late. There are more historical examples, one that springs to mind right now is in the Crusades (the first one I think) where the Christian armies where in two sperate columns with only few miles apart, the Muslims ambushed one group and the other didn´t realize what was happening until many hours later.

[F_Smith]

I see. I think.

Altho, if the two forces are seperate, and the two battles are fully seperate, then there's no reason to add the square of their numbers for combat purposes, is there? It would just be figured as 2 seperate battles, and the larger force has an advantage in open terrain.

I thought you were discussing a situation like 'Waterloo', in which the three armies came together in the same place to do battle.

And in these games, we assume a level of communication/coordination between units that was impossible until just recently, anyway . . .

I still don't think the rule works. Just a personal opinion.