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

[novice]

To make it possible to defend against stacks of doom, how about giving the attacker the option of which unit in a stack to attack? To make this work, let defensive bonuses such as +50% against mounted apply to an entire stack. Thus, the attacker can go after the stack defenders providing the bonuses first, or he can go after the dangerous offensive units in the stack.

I figure defensive bonuses should scale with the unit's health, and also there should be a limit to how many units can benefit from one defender. E.g, a spearman provides the full +50% against mounted for up to four units. If there are more units in the stack, the bonus scales down. Thus, If two full-health spearmen defend a stack of 10 units, the effective defensive bonus is 50% * (4 * 2 / 10) = 40%.

Obviously this requires a complete rebalancing of all units. It also opens up the possibility of some cool abilities, such as
- Assassin: When a unit with this ability attacks, bonuses from co-defenders do not apply.
- Sentinel: Must be targetted at least once before other units in the stack can be targetted.
- Guardian: Must be killed before other units in the stack can be targetted.

[Cyneheard]

To make it possible to defend against stacks of doom, how about giving the attacker the option of which unit in a stack to attack? To make this work, let defensive bonuses such as +50% against mounted apply to an entire stack. Thus, the attacker can go after the stack defenders providing the bonuses first, or he can go after the dangerous offensive units in the stack.

I figure defensive bonuses should scale with the unit's health, and also there should be a limit to how many units can benefit from one defender. E.g, a spearman provides the full +50% against mounted for up to four units. If there are more units in the stack, the bonus scales down. Thus, If two full-health spearmen defend a stack of 10 units, the effective defensive bonus is 50% * (4 * 2 / 10) = 40%.

This sounds like a really fascinating idea, although I'm not sure the exact implementation is effective: I don't think you want players to have to do any math to know how large of a stack they should build to manage their bonus. (How much would Spear #8 give me for a stack that's currently 37 units? Uh...)

Two possibilities that I can think of:
Make the co-defender bonuses of the same type not stack. So a spear gives, say, +40% vs. Mounted, for the entire stack. 2 spears give +40% vs. Mounted for the entire stack. Pikes could give +50% vs. Mounted, but it doesn't stack with the Spear bonus. However, a Spear +40% vs. Mounted and an Archer's +10% vs. Everything bonus would stack for +50% vs. Mounted.

What about the "defensive strikes" method from FFH, coupled with this idea? This means that if I want to target Combat Unit X, then I get shot at by the other units in the stack first. You could even have targeted defensive strikes, like each spear gets a 50% chance to hit an incoming mounted unit for 10HP. FFH has spells as one solution to getting around DS (Various Marksmen who assassinate targets work as well in some cases), so I think you'd want some type of damage-everything-in-the-stack trick as well.

Defensive strikes could have a limit on how effective they can be (say, once a unit's taken 50HP of damage from defensive strikes, no more DS can take place).

The suggested promotions would still work (Assassins are immune to DS, or ignore Guardians/Sentries, for example).

[SevenSpirits]

I whimsically considered something similar, namely the game engine picks the defender in the attacker's favor instead of in the defender's favor. The effect of this is that instead of the defensibility of the stack growing as it gains more units, it shrinks as it gains more (different) units. Thinking about it again, that is sort of desirable; it makes splitting stacks good when logistics allow. Obviously though the other effects of my idea are pretty awful. (It encourages homogeneity and completely kills the concept of a stack defender.)

Your idea has the side effects of requiring quite a bit more combat micro, which I see as a big negative even if the decisions are interesting. (Also, that math bit is bad as Cyneheard said.) The stack-wide defensive bonuses are a good idea though. That makes it better than mine for sure.

This discussion is very useful: it really emphasizes that the reason large stacks are so good is the "best defender" rule. Maybe there is a good idea out there for stack vs stack combat that would make having a smaller defending stack somewhat good. (E.g. the attacking stack can't have more units than the defending stack, combined with the right resolution system.)

[Mist]

This discussion is very useful: it really emphasizes that the reason large stacks are so good is the "best defender" rule. Maybe there is a good idea out there for stack vs stack combat that would make having a smaller defending stack somewhat good. (E.g. the attacking stack can't have more units than the defending stack, combined with the right resolution system.)

There is actually, and all Paradox games make extensive use of it. It's called combat width. The idea is that only combat fronts of certain size are manageable in battle, and with each unit having set optimal combat width only fixed amount of them will fit at the front line. Everything else is set to reserves and gets cycled in and out of combat depending on leader skill or known tech.

[RefSteel]

This discussion is very useful: it really emphasizes that the reason large stacks are so good...

Curious. In my experience, very large stacks aren't very good - least of all when defending cities. Siege just shreds them, even at trebs vs. infantry, and just about anything can clean up the mess that gets left behind. I find tactics designed to control the zone of combat far, far more effective in most situations: Coordinated surprise strikes (nav-promoted galleons/galleys/transports help a lot, as do workers playing combat engineer) when possible of course; sacrificial road cutters; hilltop outriders and sentries; pre-war cultural combat.... At least in the games I've been playing lately, if the majority of my attacking units are ever gathered on a single tile, it's for strictly logistical reasons. Or because I'm getting lazy, in which case the outcome likely won't be pretty.

(And after physics, air power adds a whole new dimension to the game....)

[SevenSpirits]

Curious. In my experience, very large stacks aren't very good - least of all when defending cities. Siege just shreds them, even at trebs vs. infantry, and just about anything can clean up the mess that gets left behind. I find tactics designed to control the zone of combat far, far more effective in most situations: Coordinated surprise strikes (nav-promoted galleons/galleys/transports help a lot, as do workers playing combat engineer) when possible of course; sacrificial road cutters; hilltop outriders and sentries; pre-war cultural combat.... At least in the games I've been playing lately, if the majority of my attacking units are ever gathered on a single tile, it's for strictly logistical reasons. Or because I'm getting lazy, in which case the outcome likely won't be pretty.

(And after physics, air power adds a whole new dimension to the game....)

I certainly don't disagree about logistics being critical. But are you saying that you intentionally spread your units out across multiple tiles?

[RefSteel]

I certainly don't disagree about logistics being critical. But are you saying that you intentionally spread your units out across multiple tiles?

Yes, at times - also generally for logistical reasons, but sometimes just because an enemy's defenders are very heavy on low-tech siege.

[Sullla]

The notion that giant stacks of units are some awesome world-beating tactical formation in Civ4 is, and always has been, completely wrong. Don't be fooled by the posters at CivFanatics who have never played against anything other than the AI. Grouping all of your units into one big stack makes you extremely vulnerable to collateral damage (as RefSteel pointed out) and greatly limits your tactical options. When your units are in one big stack, you can only threaten one target at a time, and any competent human (non-AI) player will shift defensive resources accordingly.

You can see this really well if you look at the Pitboss/PBEM games here at Realms Beyond. Nearly all of the really successful attacks (that didn't simply have a huge tech or production edge) relied on some kind of movement or misdirection to work. I'm thinking of how Realms Beyond took down Templars and Imperio in the Apolyton game (see here and also here) as well as how Speaker attacked against Jowy in the Pitboss #2 game. We've also seen innumerable attacks where the player just combines everything into one stack and ends up getting crushed by the defender; the best example I can think of being the Dreylin/mostly_harmless war in PBEM #1 (although in Dreylin's defense, the map forced him to attack through a narrow pass).

I don't think there's any fault in a system where the best defending unit defends. It forces the attacker to have mixed, balanced stacks (and they are punished accordingly if they don't!) Just make sure to have some kind of collateral damage system (doesn't necessarily have to be siege), and something resembling the Civ4 system works really well, IMO. Civ5 demonstrates how bad things can get when designers go overboard and ignore what works.

[SevenSpirits]

Sullla, I'm totally with you on that. In fact I didn't mention stacks in my original post at all. But what I did say is somewhat related, namely that I wanted things like forts to be more relevant. As things stand, defensive positions that aren't valuable objectives themselves (cities) have limited value. Essentially, that's because you want to just send all your guys at the opponent's cities, and avoid as much actual combat in enemy territory as possible.

[Mist]

The notion that giant stacks of units are some awesome world-beating tactical formation in Civ4 is, and always has been, completely wrong.

But nobody here raises this notion ( I think ) :] I just pointed out it would be nice to have a counter to huge stacks other than catapults ( which is ridiculous from purely aesthetic point of view. The whole notion of 'suicidal defensive siege' feels just plain wrong )

All the discusion since then is simply brainstorming on counters other than colateral on siege and stack combat in general. Not because we think SODs are uberpowerfull, but because we think this area in Civ4 is limited and might benefit from improvements.