Say you're designing a Civ game. How do you make the combat system?

[luddite]

I came up with a civ 4 mod that changes the combat system a bit. All the siege units are greatly weakened (the catapult has only 1 strength, the cannon 3, artillery 5, and mobile artillery 7, with 50% collateral damage) however they can damage up to 200 units on a tile and will always withdraw.

I've also changed the terrain and fortification bonuses. Open terrain has a 20% defense boost, and hills give 35% (hill forests give 65%). Fortifying gives a 20% boost/turn, up to a maximum of 60%. If a unit is on open terrain with no other bonus, it takes 5 artillery to damage it enough for an equivalent unit to attack at even odds.

Hopefully these changes will make it so that being the defender is a bit easier, and allow a slow methodical advance where you take territory. It should also make large stacks more vulnerable to seige, while small stacks are less vulnerable.

I didn't do anything about flanking damage, no idea what to do about that one. I'm tempted to just remove it completely- just having 2 moves is already a big advantage. Anyone have any thoughts about it?

[SevenSpirits]

One thing I've been seriously pondering is letting units use enemy roads by default (i.e. free commando for everyone).

- This makes it important to guard the ends of your roads (at the edge of your territory, where enemies can potentially enter your road with movement left).

- It makes it useful to have (defended) forts on your roads, so that you can use the roads more effectively than any invaders.

- It makes it critical to plan your road network carefully and not just spam them everywhere.

- It makes it more important to engage the invaders early when they enter your territory, before they can claim and abuse mobility.

- It allows a greater variety of movement plans by the invader.

Certainly this would constrain the other parts of any combat system that included it. I don't think it would be that hard to design around it, though.

[Square Leg]

One thing I've been seriously pondering is letting units use enemy roads by default (i.e. free commando for everyone).

- This makes it important to guard the ends of your roads (at the edge of your territory, where enemies can potentially enter your road with movement left).

- It makes it useful to have (defended) forts on your roads, so that you can use the roads more effectively than any invaders.

- It makes it critical to plan your road network carefully and not just spam them everywhere.

- It makes it more important to engage the invaders early when they enter your territory, before they can claim and abuse mobility.

- It allows a greater variety of movement plans by the invader.

Certainly this would constrain the other parts of any combat system that included it. I don't think it would be that hard to design around it, though.

This is actually very interesting. I don't think that using railroads would be viable though.

Only usable in MP since the AI (unless modded) would still spam roads whereas a human will set up "checkpoints"

[Square Leg]

however they can damage up to 200 units on a tile and will always withdraw.

This is what I meant with my post earlier in the thread. Basically a densely populated tile is actually more likely to receive collateral then a sparsely populated one. By populated I mean with units.

[luddite]

One thing I've been seriously pondering is letting units use enemy roads by default (i.e. free commando for everyone).

- This makes it important to guard the ends of your roads (at the edge of your territory, where enemies can potentially enter your road with movement left).

- It makes it useful to have (defended) forts on your roads, so that you can use the roads more effectively than any invaders.

- It makes it critical to plan your road network carefully and not just spam them everywhere.

- It makes it more important to engage the invaders early when they enter your territory, before they can claim and abuse mobility.

- It allows a greater variety of movement plans by the invader.

Certainly this would constrain the other parts of any combat system that included it. I don't think it would be that hard to design around it, though.

I don't really want to let the attacker use the defenders roads. I feel it would just encourage blitz attacks on the capital even more than what we already have. This would be especially problematic for simultaneous turn multiplayer- someone could destroy your capital before you even realize that they've declared war.

Maybe you could change it so that the attacker can build roads inside enemy territory, so that you can have your own road network if your units are there long enough. Or a system where tiles with a unit fortified on them become part of your culture.

[SevenSpirits]

I don't really want to let the attacker use the defenders roads. I feel it would just encourage blitz attacks on the capital even more than what we already have.

That would require all of the following:

- The defender has decided to road from their capital directly to their border.
- They have decided to not defend that road and/or their capital.
- There is no mechanism by which siege is fairly necessary to bombard cities before capturing them.
- There is no way (e.g. zones of control) for players to stop their opponents' units from running by them unconcernedly.
- Attacking only cities with a single large pile of units is at all a good idea in this combat system.

I think you're making too many assumptions. Though city sniping is a useful problem to have concerns about, there are some pretty easy ways to completely invalidate that tactic.

This would be especially problematic for simultaneous turn multiplayer- someone could destroy your capital before you even realize that they've declared war.

That is a good thing to keep in mind, though I think I care much less how well simultaneous multiplayer ends up working. Trying to design a single combat system that works well both turn-based and real-time seems like an unnecessarily onerous constraint. Plus I have no evidence that it's possible at all (as I don't consider civ IV's system to work "well" in real-time, though it isn't terrible).

[Mist]

It sounds like the idea of suicide units really bothers you. It never really bothered me much, because I just thought of it as the equivalent of using up ammunition. But it might be fun to try out a different combat system.

What bothers me is idea of using suicide unit as a tactic staple and the solution for all problems. Having it disguised as siege only adds insult to injury

I'm an old school strategy gamer. My first games used folded paper maps, cardboard square tokens, six sided dice and plenty of combat tables. Combat was slow, methodical, rarely ending with decisive victory ( destruction of opposing unit ). You had to work for your breakthrougs and protect your units from destruction. Suicidal maneuvers were akin to chesslike gambits rare and with an idea behind them, never as an afterthought.

Not that I want to make civ into that type of strategy, or compare it to one, apples and oranges. But it does leave a mark on my thinking. And I think Civ combat idea is wanting. It suffers from unnecessary the idea that combat has to be decisive, from forced dash to the city combat style, and from godawful implementation of siege and colateral.

Unfortunately, in most of the games I've seen with seige units that can attack without being killed, the seige units end up being far too powerful. That's how it is now, in civ 5- artillery can crush everything else without ever dying. Which is maybe appropriate for a world war 1 mod, but not very fun for a game.

Yeah, but tht's also poor implementation. Siege does no have to be able to attack on it's own. Neither does it have to gain in power dramaticaly when leveling.

Archers and longbows could be a "light seige" unit that's good at damaging enemy units but unable to destroy fortifications.

Yes. You could also make skirmisher/guerilla type unit that causes light colateral and has a chence to inflict loss of movement.

I guess I misunderstood what you guys were saying. It sounds like you're looking for a different combat system which would reward a more slow and methodical form of warfare, taking ground, instead of the civ 4 style where you dash forward to raze a city. That's sort of how it is in civ 5, or in early wars in civ 4, where you have to fight over specific tiles that can be defended.

Maybe there's some way you can tweak it so that both styles of warfare are rewarded?

And that would be the best outcome as far as I'm concerned. But it's SevenSpirit's mod. I'm just throwing ideas around

[SevenSpirits]

Another aspect I've been thinking about is healing. (I was wrong earlier; healing is actually what makes big stacks so strong in Civ IV. Or would if there was no collateral damage, which I will ignore for this post as I wish to imagine a world without it.)

Specifically, the fact that much of the damage you take in combat is partial and temporary unless locked in by having your unit fight again and die. For example, say two stacks of 10 warriors fight. The average outcome is 5 wounded warriors left for each side. If your stack had 15 warriors instead, you could probably get (almost) 5 free kills just by outnumbering your opponent. And then there's the fact that while your opponent's wounded guys die, yours promote!

Just for fun, I tested out how many stacks of 10 warriors I could kill with a stack of 30 warriors, if I was allowed to heal up between engagements but had to just stack attack every turn while there were enemies to kill. Anyway, despite some very bad luck in the first combat (I lost 8), my guys killed 77 enemy warriors before finally dying to the eighth stack.

Civ V took this even farther, by having combat not necessarily result in any deaths at all, and allowing instant heals. Oops. The result there is that the human can micro to almost never lose anything against an AI. It's terrible!

The rationale for healing of course is a) to cut down on the number of troops you build in the game, b) to allow the RPG aspect of building up experienced veterans, and c) through veterancy to make war not only be a terrible military loss for both sides. But I wonder if the price (basically turning many battles into lopsided routs) is worth it.

One possible system that would remove healing is trivial: every combat results in one unit dying (at certain odds) and the other unit living on with no wounds. Of course, that makes the outcome of each fight even more swingy. But the result is that it doesn't matter as much if you engage with 10 units now and 10 later, vs with 20 all at once. Nevertheless this system seems pretty dumb.

Another idea would be to allow actual stack attacks that work like this:
It starts the same, with one attacker stepping forward and the best defender facing off against it. But then, whichever side wins the combat has that (wounded) unit fight off against the remaining enemy that has the best odds (or perhaps, the best ratio of odds to base strength). This way, pretty much every wounded unit dies. However, damage still exists and is relevant in how much damage the next combatant will take. It's much less swingy. In fact the algorithm for an individual fight could be made significantly less variable as well, and it would still work fine. With this system, I think you'd want to lessen the unit counter bonuses (50-100% being a bit too high). I haven't figured out yet if this second system would be any good, but I haven't disqualified it yet at least. It also has the advantage of the best move tending to be really easy to execute (stack attack, rather than choosing the best attacker yourself at each step).

[T-hawk]

Another aspect I've been thinking about is healing. (I was wrong earlier; healing is actually what makes big stacks so strong in Civ IV. Or would if there was no collateral damage, which I will ignore for this post as I wish to imagine a world without it.)

Specifically, the fact that much of the damage you take in combat is partial and temporary unless locked in by having your unit fight again and die.

This is absolutely true. How to fix it? Make healing have a cost. Starcraft got this right 12 years ago, where repairing damaged units costs money, and also time from a utility unit (SCV) and logistics in making them meet.

The implementation in Civ could be done any number of ways. Maybe you simply click a repair button that costs gold, like the upgrade button. Maybe a worker or a new medic unit or a unit with the existing medic promotion gets a button and action to heal damaged units. Maybe a damaged unit in a city consumes up to 10 hammers per turn of the city's production in order to heal.

[Shoot the Moon]

This is absolutely true. How to fix it? Make healing have a cost. Starcraft got this right 12 years ago, where repairing damaged units costs money, and also time from a utility unit (SCV) and logistics in making them meet.

The implementation in Civ could be done any number of ways. Maybe you simply click a repair button that costs gold, like the upgrade button. Maybe a worker or a new medic unit or a unit with the existing medic promotion gets a button and action to heal damaged units. Maybe a damaged unit in a city consumes up to 10 hammers per turn of the city's production in order to heal.

Perhaps the "cost" to heal could also be implemented by switching around the medic promotion. What if you were to make healing incredibly slow without a unit in the stack with medic, while also making the medic promotion only useful for a certain number of units at a time (so you would need more than just one medic unit in a stack). Furthermore, if you were to make the medic promotion say -50% strength, you would essentially have to invest hammers into units that allow your troops to heal. The exact balance of the cost would have to be worked out, but I think it is worth considering that the cost to heal doesn't necessarily have to come from an entirely new system of gold for healing but could just change promotions already there.

In regard to the roads, if you do make roads useable by the attacker, it would be important to allow someone to raze roads within their own culture I think. It might also be smart to have some other way of hooking up resources other than roads.