Fortifications and Sieges

[Mark_Everson]

Thanks for the example Laurent! It looks basically reasonable to me, although I'm not as with it tonight as I might be

I'm embarassed to ask, but is a round a tick, a turn? If a round is a tick then these times may be too fast, if a turn, maybe too slow... I'll await the definition, and more brain-power, before I comment further.

[LDiCesare]

A round is somewhat abstract here, as are the figures. In the code, there are several rounds of fight in a tick, so the figures are really bad here. I don't recall the max number of rounds in a tick, must be around 30. The round stuff is there mainly to allow fight to be staged at different distances (long range, then the enemies move to medium, short range and melee, each move taking a number of rounds depending on the mobility of the units and their desire to move closer).

If the model sounds OK, I will code the stuff that is still missing, and then change my batch test to tune the figures. Hopefully I will have an e-mail working to send the code when I am done.

[Mark_Everson]

Hey Laurent, I agree its good enough for you to take a shot at it. If something doesn't work I'm sure we can get around it with a few rules tweaks. Looking forward to checking it out!

[LDiCesare]

Balancing values

I have a first funtional code, but I have a few issues.
First, I had to add a variable to tell the economic model to shut up because I cannot model a siege if the square isn't owned, and the economy doesn't allow me to own a square with no population, or I need to create an economy in order to run a batch military test, which I don't want to (fixing it everytime the econ changes - arglll, no, htank you). I had to touch a few classes here and there, also about the City class which holds some Wall information. I am not yet clear as to where the wall info will be stored, but I think the xml will add forts to a location the way a city is added to a location.

Now my balancing problems:
I have made a test where a phalanx is in a wooden fort, and is attacked by legions. To make things even, I can test with legion inside instead of phalanx. I assume legions have engineers who can sap the walls. Now, what should be the ratio number of legions needed in order to win the fight? In the open 2 vs 1 is enough, and 1 vs 1 provides roughly 50% (slightly less because of terrain advantage for defender). What if it was warrior vs. warrior (i.e. no wall-breaching)?
These questions are very important as it will decide the value of defense (warriors) and health (legion with engineers) of the wall.

Right now, I decided that it should be at least 3 vs. 1 before the attacker would decide an assault, but maybe 2 is enough? What ratios should we have? If you have no idea about engineers/sappers. opinions about rams/trebuchets/whatever are welcome too.

[Mark_Everson]

Re: Balancing values

Hi Laurent!

Originally posted by LDiCesare
I have a first funtional code, but I have a few issues.
First, I had to add a variable to tell the economic model to shut up because I cannot model a siege if the square isn't owned, (snip)

Ok with me, we can figure how to do it right when the city stuff is done properly.

Now my balancing problems:
I have made a test where a phalanx is in a wooden fort, and is attacked by legions. To make things even, I can test with legion inside instead of phalanx. I assume legions have engineers who can sap the walls. Now, what should be the ratio number of legions needed in order to win the fight? In the open 2 vs 1 is enough, and 1 vs 1 provides roughly 50% (slightly less because of terrain advantage for defender). What if it was warrior vs. warrior (i.e. no wall-breaching)?
These questions are very important as it will decide the value of defense (warriors) and health (legion with engineers) of the wall.

Right now, I decided that it should be at least 3 vs. 1 before the attacker would decide an assault, but maybe 2 is enough? What ratios should we have? If you have no idea about engineers/sappers. opinions about rams/trebuchets/whatever are welcome too.

I wouldn't fine-tune too much up front. For your case with equivalent units with one side behind wooden walls something like 3:1 for success sounds ok provisionally. I would expect the attackers to lose between 1/3 and 2/3 of the attacking force at 3:1. Of course wooden walls would be relatively easy to breach, and that would make the attack less expensive.

On the issue of rams etc, in my mind we should just have a "siege weapons" element. It would be costly, and move slowly, and have primarily a 'breaching' attack. I don't see the point of modelling in detail lots of obscure siege weapons. What do you think? If you agree, I'd just set the cost and breaching capability together so that the unit is cost-effective compared with 'normal' military units.

[LDiCesare]

I agree we can have few siege weapons. I thought about having one engineer to encompass all mining/sapping/light catapult(scorpion) stuff, and other elements more detailed.
I agree that having a single element could be a good thing, but it has to be very tech-dependent, as there were real evolutions in siege warfare, from ballistas to trebuchets. Calling the thing "catapult" would be nicer than "siege weapon". One important difference between various siege weapons is range: sap, mine, rams, are all close-range, whereas catapults are long range. So engineer would be the short-range, and catapult the long-range breaching element. Later in the game we would have engineer and cannon.

Note that the tuning I do is not really "fine". I have to check the model so that it doesn't behave in too chaotic a way. Adding a single unit can make lots of changes right now.

A last point on engineer elements. I think they should be able to repair their walls a bit during the siege, not much, but a bit (it was typical to counter-sap by adding a second stone wall behind the first, or to build a wooden wall behind a weakened stone wall in order to delay the breach more. I won't code it right now, though, as I have to balance things out before.

[Mark_Everson]

Hey Laurent, everything you say sounds fine to me. I had indeed expected there to be changes of siege weapon effectiveness with tech. Your two-element approach (engineers and catapult or whatever) seems quite good to me.

Can you check out the discussion in the D7 thread? Pierre is finding Clash to run really slow for him, and it would be good to figure out if its a garbage collection issue with the mac java implementation, or just plain the game takes up that much memory and his system can't handle it. If you're up it, it would be good for you to work together on it since you can work directly without the web!

[LDiCesare]

Building forts/walls: User interface and coding issues

I have wondered how to do in order to be able to build fortresses. Walls look like units, but you want to place them manually on the map, a bit like roads.
So I think we need a button to place forts as we have one to place roads. There would be a button for selection of the square, a special section in the econ orders for the various fort types (wood, stone...) and when funds are allocated and a square is clicked, the fort starts being built. Note that ideally, the button should spawn a dialogue asking for the kind of wall asked for, and how much income is to be spent on it, at the province and at the civ levels. Otherwise there is no telling whether fort1 is to be a stone fort and stone2 a wood fort or vice versa.

Also, how does that translate into econ orders? I checked the road-building stuff, and it is quite hard-coded. I don't want to do the same for forts. I need a way to be able to give a build infrastructure order to a square, or to precise a target square to a buildable, which may not be the capital of the province (contrary to what happens with units). Ideas are welcome.

[Mark_Everson]

Re: Building forts/walls: User interface and coding issues

Bonjour Laurent, sorry I missed this a while ago. . .

Originally posted by LDiCesare
I have wondered how to do in order to be able to build fortresses. Walls look like units, but you want to place them manually on the map, a bit like roads.
So I think we need a button to place forts as we have one to place roads. (snip)

Sounds good to me. Priority for building forts will be arranged the same way they were ordered in time.

Also, how does that translate into econ orders? I checked the road-building stuff, and it is quite hard-coded. I don't want to do the same for forts. I need a way to be able to give a build infrastructure order to a square, or to precise a target square to a buildable, which may not be the capital of the province (contrary to what happens with units). Ideas are welcome.

I would do this the same as roads for the near future. Specifically there would be just a "fort-building" class, analagous to the road-building one. The class would build up points, and when you get to the point cost for the first things on the list, it'd be done. So as to make it not be completely unrealistic, we could determine the max rate of building locally so you can't flash-build forts. In terms of programming, a fort could not be built until a number of turns passes such that it could have been built locally. But to keep it simple there would just be one civ-wide pool of fort and wall-building points. Sound ok to you?

[LDiCesare]

The difference between road and fort-building is that there are different kinds of forts. This means I need a class which can build different kinds of infrastructure. I'd rather build a new one which would also specify what kind of UI it wants, so I could fold the roads into its framework. There are differences since one needs path, the other doesn't need it, but I think it can be done, and it would allow for roads, railroads, etc.

[Mark_Everson]

Hey Laurent:

Originally posted by LDiCesare
The difference between road and fort-building is that there are different kinds of forts. This means I need a class which can build different kinds of infrastructure.

IMO fort construction is relatively similar whether it be a wooden palisade or a stone fort. Local materials are used, and the most difficult input is labor. I think we should spec out a fort as to exactly what the player wants, and then calculate the cost. The cost would simply be the number of fort infrastructure points needed to build it. It is a simplification, but I personally don't think its worthwhile to have many different classes of fort infra. Maybe one or two over all of history should do. Perhaps rather than calling the wooden one wooden, we should call it a "light wall" or some such. That could cover thin walls made of mud bricks and such which are roughly equivalent to wooden ones. Then we wouldn't get people complaining about building wooden walls in a desert.

I'd rather build a new one which would also specify what kind of UI it wants, so I could fold the roads into its framework. There are differences since one needs path, the other doesn't need it, but I think it can be done, and it would allow for roads, railroads, etc.

Sounds reasonable to me. Certainly building walls like the Great Wall have something in common with road building. But it might overcomplicate things in the short run to put these two together.

[LDiCesare]

Sorry I used the word infrastructure incorrectly. I ment different results, walls being of different kinds the same way units are of different kinds. Overall I think we are in agreement.

[Martin the Dane]

Hi Mark just asked me to post my comment from the Demo 7.2 Proposal! thread here.

(Please note that I have not yet read the discussion here, so I appologice if some of my comments might seem out of place)

Well, since Mark invites comments, I'd just as well vent one of my main frustrations as to walls in civ-type games.

As far as I have been able to determine there has been two versions of the same model as to implementing them:
1. Give the defender a fixed defencive bonus (CTP)
2. Give the defender a persentage bonus (CIV)

And to tell the truth non of them makes much sence.

In reality a single troop of archers - with sufficient supply arrows that is - could hold of a substantial force of non-ranged attackers indefinately and slowly wear them down, by picking off anyone stupid enough to enter their range.

What I'm saying is that no force, no mater how large, should be able to defeat a garison in combat if the garrison is protected by walls, unless they have some form of wall (or gate) breaching equipment. They can of cause besiege the defenders and wait for them to starve.

So in order to add reality to walls only certain units, or units with certain abilities should be able to attack walls and the units within.

Weaponds for wall breaching could be:
Ancient:
- Balistas
- Catapults
Middle age:
- Explosives
- Canons
Modern age
- Any artilery or heavy guns

Weaponds for attacking units within walls could be:
Ancient:
- Balistas with fire arrows
- Catapults with burning shots (tar)
- Catapults with "buckshots"
Middle age:
- Canons with axploding amo
Modern age
- Any artilery

[LDiCesare]

Good points Martin, except they apply mostly for stone forts. There are examples of wooden forts which have been burnt down by a single man (like a man removing leather hides from a part of the wall, allowing to burn it down), also, there are examples of entering a walled city by outwitting the defenders (think Troy, also Bertrand du Guesclin in the 100 year war).
Note that Age Of Wonder had a siege model like what you suggest, but they changed it in AoW2 so you wouldn't need breachers.

The current model gives a fixed defense bonus but also prevents flanking. That means hand-to-hand units can't attack at more than one on one. Plus it allows for siege weapons.

The use of siege weapons can make a big difference, however, the defense bonus needed to mlake it so is very big. You will see that in the next demo, as I made a scenario specific to test sieges. In it, you can take a walled city without siege engines, but it will take lots of casualties and lots of time. Too much time, in fact .
The question is what defense value should the walls provide? You say infinite against close-weapons, and limited against siege weapons. I reached a very big value in the siege scenario. Playtest will tell.

[Martin the Dane]

Well Laurent, you do have a point, and I agree that most cases can be modeled by giving the defenders a wery high defence bonus.

After having read some of the thread, I see that you plan to include provision, morale, helth etc. in the siege model. That is enocuraging so I look forward to play your scenario.