How would you improve combat in Civ7?

[historix69]

I always thought there should be two modes for moving your troops : march and fight.

About 20 years ago there was an Axis & Allies RTS featuring different move modes for units of Division size :
- attack : normal combat efficiency, -25% movement speed, (AI aggressiv)
- normal : -25% combat efficiency, normal movement speed, (AI aggressiv)
- column : -50% combat efficiency, +25% movement speed, (AI passiv)
Units which do not move for some time will automatically fortify and get combat boni. (Probably increased combat efficiency and less damage.)

In a game like Civ were most units move only 1-2 tiles per turn, small boni like 25% movement speed do not really matter. Road and Railroad have stronger effects.
So the movement boni probably should better be in tiles per turn or in combination with R/RR boni, eg only a unit in "column" mode could use a R/RR bonus.

 

[aieeegrunt]

These kinds of tactical details are way way below the scope of abstraction for the scale of this game

Redesigning the game to allow for limited stacking would be a far bigger ROI for making combat not terrible

Experimenting with Civ4, and modded Civ6 and Civ5 seems to indicate that three unit stacking is the sweet spot for avoiding all the issues with 1UPT like having to solve a sliding tile puzzle when moving your units, and not turning into micromanagement Stack Of Doom hell.

This replicates the trend in Consim Boardgames where Civ takes it’s roots, and they also rapidly iterated towards three units per hex

 

[Boris Gudenuf]

These kinds of tactical details are way way below the scope of abstraction for the scale of this game

Redesigning the game to allow for limited stacking would be a far bigger ROI for making combat not terrible

Experimenting with Civ4, and modded Civ6 and Civ5 seems to indicate that three unit stacking is the sweet spot for avoiding all the issues with 1UPT like having to solve a sliding tile puzzle when moving your units, and not turning into micromanagement Stack Of Doom hell.

This replicates the trend in Consim Boardgames where Civ takes it’s roots, and they also rapidly iterated towards three units per hex

Had a lot of experience with 'stacking' board games Back In The Day, including the Europa games which allowed up to 8 or more counters per hex stacking - specially modified tweezers were almost mandatory to play those games!

But most of the problems with the stacking in board games were physical: the act of moving tottering stacks of cardboard, especially when playing a game with 1000s of unit pieces (like Europa's Scorched Earth covering the War in the Soviet Union 1942 - 1945) was a monumental pain every single turn.

Computer games, if properly designed, avoid all that. Any stack can be, with a click, opened up for inspection without disturbing its placement on the game map. Any attribute of any individual or group of Units can be examined or modified without worrying about how steady your hand is on the tweezers. A limit of 3 or any number of units per tile in a computer game is strictly artificial from a physical standpoint, which was always the basic problem in cardboard and paper board games.

Now, the Cognitive Problem is something else. Even well-trained and experienced people (like, real Generals in real armies) have trouble keeping track of more than 5 - 6 distinctly different units, which is why real military forces are Layered: a company commander has 3 - 4 platoons to worry about, his battalion commander gives orders to 3 - 4 companies, and the division commander orders about 3 - 4 brigades or regiments, not their component 256 platoons!

So, if the worry is how to keep track of multiple units, invoke the Order of Battle and group the units to make control easier. This is actually easier with stacks than with the wretched 1UPT we have now: instead of moving 30 individual units, you move 2 - 3 stacks/armies. How the individual units act and interact within those stacks IS NONE OF YOUR BUSINESS as the Grand Poobah of the Civilization. "If a sparrow falls" doesn't matter to you, as long as you personally aren't directly underneath the air-to-ground avian. You move the armies. When they get into battle, you tell the little digital General generally what you want him to do (attack carefully, attack everywhere, defend, slowly retreat, etc) and let him and the computer tell you what happened afterwards.

As Grand Poobah, you build the units, group them into armies, and get them at the enemy. After that, to repeat the often-quoted von Moltke: "No plan survives first contact (with the enemy)" - and the game should reflect that, rather than you as God The Commander moving every individual sparrow and unit on the map. If you want to play at the sub-tactical level, go play Advanced Squad Leader or its electronic ilk, the first person shooters.

Civilization the Grand Strategic Game should run battles at an entirely different level.

 

[aieeegrunt]

Had a lot of experience with 'stacking' board games Back In The Day, including the Europa games which allowed up to 8 or more counters per hex stacking - specially modified tweezers were almost mandatory to play those games!

But most of the problems with the stacking in board games were physical: the act of moving tottering stacks of cardboard, especially when playing a game with 1000s of unit pieces (like Europa's Scorched Earth covering the War in the Soviet Union 1942 - 1945) was a monumental pain every single turn.

Computer games, if properly designed, avoid all that. Any stack can be, with a click, opened up for inspection without disturbing its placement on the game map. Any attribute of any individual or group of Units can be examined or modified without worrying about how steady your hand is on the tweezers. A limit of 3 or any number of units per tile in a computer game is strictly artificial from a physical standpoint, which was always the basic problem in cardboard and paper board games.

Now, the Cognitive Problem is something else. Even well-trained and experienced people (like, real Generals in real armies) have trouble keeping track of more than 5 - 6 distinctly different units, which is why real military forces are Layered: a company commander has 3 - 4 platoons to worry about, his battalion commander gives orders to 3 - 4 companies, and the division commander orders about 3 - 4 brigades or regiments, not their component 256 platoons!

So, if the worry is how to keep track of multiple units, invoke the Order of Battle and group the units to make control easier. This is actually easier with stacks than with the wretched 1UPT we have now: instead of moving 30 individual units, you move 2 - 3 stacks/armies. How the individual units act and interact within those stacks IS NONE OF YOUR BUSINESS as the Grand Poobah of the Civilization. "If a sparrow falls" doesn't matter to you, as long as you personally aren't directly underneath the air-to-ground avian. You move the armies. When they get into battle, you tell the little digital General generally what you want him to do (attack carefully, attack everywhere, defend, slowly retreat, etc) and let him and the computer tell you what happened afterwards.

As Grand Poobah, you build the units, group them into armies, and get them at the enemy. After that, to repeat the often-quoted von Moltke: "No plan survives first contact (with the enemy)" - and the game should reflect that, rather than you as God The Commander moving every individual sparrow and unit on the map. If you want to play at the sub-tactical level, go play Advanced Squad Leader or its electronic ilk, the first person shooters.

Civilization the Grand Strategic Game should run battles at an entirely different level.

Oh Mer Gawd I remember those tweezers thank you for that

The “Cognative Problem” as you aptly described it covers some of why I think 3 UPT is the sweet spot but I think the biggest reason is still a UI one

Three units per tile is about as big as you can go without an extra step to “expand” the stack to see what’s in it

You can comfortably put three easily distinguishable unit icons in a hex at most reasonable zoom levels and see at a glance which units you have on the map and where without any extra steps. This is a mechanical limitation just as unavoidable as “stack skyscrapers” in a cardboard medium.

During my experimentation I had vox Populi running for Civ5. I ended up purging it (and boy was it annoying getting it removed) because there was no combo of mods that allowed unit stacking AND seeing what was in the stack at a glance. There were other reasons, but that was the biggest one. The ease of use hit was that bad.

The other big reason is micromanagement that comes with having the number of units stacks of doom requires. Going back to Civ4 the nostalgia glasses came off fast because that game has a ton of micro, and building swarms of units, combining them into stacks, figuring out optimum stack composition etc was a TON of micro that in the end did not change the actual end result much over a 3 UPT setup

Think of of all those individual units in a Stack of Doom as essentially being equivelant to pieces of the health bar for a larger scale unit

Building 4 swordsmen, moving them all into a stack, then moving that around is functionally basically the same as having one swordsman in a 3 UPT hex and moving it around.

Having that stack be beat up in combat and reduced to one swordsman is basically the same as that 3 UPT swordsman being reduced to 25% health

The difference is that I didn’t have to make all of those extra clicks to build and combine that 4 swordsman stack. I also don’t have to deal with expanding stacks to see what is in them and the other UI issues I addressed

Best Of Both Worlds I say

 

[historix69]

The charm of 1upt is that game rules are easy. Combat is always one single unit against another single unit, you can apply unit type boni, etc.

Once you have more than one unit in a tile, you have to define new game rules how combat results are calculated, how many units in a stack can attack, how many can defend, how bombardment damage is spread in a stack, how many attackers move to the conquered tile, etc..

- Civ 1 had the rule that the attacking unit and the best defender unit of a stack simply made a duel and if the attacker won the attacked stack commited suicide and vanished. There was an exception for city defense and fort defense where a defeated defender died alone and the defending stack did not die as a group.
- (I don't remember much of Civ 2 besides the fact that combat was broken up into more single steps of RNG combat to reduce the number of tanks and battleships being destroyed by fortified Phalanx units on mountain tiles.)
- Civ 3 separated combat so that the attacking unit and the best defender unit against the attacker (unit type boni) made a duel and if a unit died other units were not affected as long as the defending stack still had a defender. I remember that ranged units with movement points could support the defender by bombarding the attacker before 1st round of combat. If the last defender of a stack was defeated, remaining artillery units were captured, leading to insane stacks of about 100 artillery units in late game. (Later patches added shortcuts to move whole stacks of a unit type with a single command.) There also was the feature that units fortified near a road automatically sniped on enemy units passing by (probably instead of Zone of Control). Ranged units could damage defenders down to 1 HP but not kill them.
- Civ 4 was similar to Civ 3 in combat except that artillery type units now sacrificed themselves when attacking and could damage several units in the defending stack.

This fall back to single unit combat steps in Civ 1-4 kept rules for combat with stacks simple, but it ignored the fact that a side with more units usually has kind of a superiority in combat, eg having several attackers shooting on a single defender increases the chance that the defender is hit (or suppressed).

If you consider that a hex tile represents a much bigger area than a single unit can occupy, then both attacker and defender should be able to send several units into a battle depending on the hex' front length (which may vary depending on terrain) and the (individual) front size of involved units. Further the attacker should be able to attack a hex from more than one adjacent tiles at the same time so that the overall battle might involve several attacking and defending stacks along the borders of a single tile. (Technically a 2nd defender stack on a tile neighbouring an attacker should be able to also support the defender and counterattack the attacking stack.) While infantry, tanks, etc would fight in the 1st line of the front, ranged (artillery) units would support attack and defense from the 2nd line. Since both attacker and defender might have a mix of unit types in their stacks, several unit type boni might or might not apply.

Defining new consistant game rules to calculate the outcome of such stack-against-stack battles is not easy and probably the biggest challenge ... and once you have the algorithm defined it does not matter if you limit the stack size to 2 or 3 or 10 units. However since a hex has 6 sides, for consistancy reasons the overall stack limit should be much bigger than six times the number of units allowed to defend on a single front line.

 

[Boris Gudenuf]

The 'charm' of 1UPT is that game (combat) rules are easy. It could be said that's the charm of Checkers, too. But I don't need a several thousand dollar computer to play checkers, and if a computer game can't do better than checker-level rules why should I waste my time playing it when a $5 checkers set will do as well?

This is a silly argument. 1UPT is a tactical mechanic, and the battles in Civ have no business being tactical battles. The largest military units we have a record of in ancient Mesopotamia and Egypt were the Sanga of about 700 men (Sumer/Akkadian) and the Pedjet of 1000 - 1200 men (New Kingdom Egypt). Your ancient 1UPT battle, then, would have about 1000 men in a tile covering hundreds of square kilometers facing off against another 1000 men in another tile in a 'battle' lasting At A Minimum 40 years/1 turn, and, frankly, usually lasting 80 - 160 years. In other words, 1UPT in a Grand Strategy Game like Civ is by it's nature ridiculously out of Time and Distance Scale compared to the rest of the game.

To have any hope of being in line with the rest of the game, ALL battles in Civ must meet two criteria:
1. The battle is resolved in One Turn
2. Most battles (until the modern Mass Army Era) take place in One Tile.

That No. 2 means, also, that most Armies take up 1 Tile. 'Stacking' is inherent unless you make all Units mass armies, which results in a bland and uninteresting unit mix, to say the least.

IF the game is going to keep something resembling Unit mixes, then they have to be stacked for battle and the tactical interactions among them in battle have to be largely Computer Resolved. It's no business of the Gamer, the 'Spirit of the Civ', to maneuver a bunch of archers onto a hill for best effect, or charge into the flank of those archers with a bunch of chariots: there are Captains, Colonels, and Generals to make the units do that,(and tactical battle games if you want to do it yourself) and the gamer's (Grand Panjamdum's) job should be to form the army from Units and get it onto the battlefield (into the tile with an enemy Army) - and, I propose, give basic orders for the type of battle he wants the digital General on the Field to fight just so you don't feel like you are throwing all those units you spent turns building into a Black Box to be ground up.

Defining any 'tactical' (or operational) rules for resolving battles is complex, but it has been done (see Dupuy's book History, Numbers and War, full of massive formulae for calculating 'combat power' with various weapons, terrain, weather, posture and numbers), and it only has to be done Once. After that, resolving battles under those rules is done by the computer and doesn't require any direct input from any gamer.

 

[historix69]

This is a silly argument.

You probably misunderstood my post.

In the first part I more or less tried to describe how Civ 1-6 stuck to combat rules limiting the game mostly to atomic battles with one unit against another unit. And imho the reason for the devs always was that combat rules for one unit against another unit are simple, easy to understand, easy to implement and probably widely accepted as a compromise by a majority of players.
In the second part I started to describe the complexity of a battle of (several) stacks attacking a single tile involving multiple units on both sides and that the challenge is to define the algorithm which calculates the result of such a battle. It gets even more complicated if you allow several lines of units, eg skirmishers, archers, infantry, cavalry, reserves, long range artillery as well as terrain modifiers, rivers, fortifications and other obstacles for each front ... Having a 300 pages manual to understand how battles are calculated will be a challenge, especially for players who actually are used to command such battles in detail in Total War games ...

You mentioned Dupuy in your own thread before, but his prints from the 1980s (?) are not easily available to people outside university or army in Europe. (A single copy of a similar book "Numbers, Prediction, and War" (1985, 256 p.) is listed at amazon for about 500 Euro. Dupuy died in 1995 and the book is out of print.) Maybe you can link a free copy of the book or the relevant chapters ... or post the battle algorithm.

 

[Boris Gudenuf]

The complexity of tactical battles has no place on the Grand Strategy scale of Civilization games. The smallest time interval is a 1 Year turn. If that year was 1944 you would have to fight 5 major land campaigns in the Soviet Union, a major amphibious assault and 3 major land battles or campaigns in France and fight off another major amphibious assault and land campaign in Italy if you were playing Germany. You would have to be able to rebuild about half your army on both fronts at least once during the turn and raise 50 new infantry divisions in a single turn. Civilization just doesn't play well at that scale. In fact, it can't if it is also going to cover 6000 + years and be playable in the average lifetime.

It's not alone in this. The old Europa boardgames were billed as Operational Level, yet they included units as small as battalions (which are tactical units) and had 2-week turns, which means the entire Battle of Kursk, involving 2.5 million men and great assaults, counterattacks, and maneuvers, had to all play out in a single turn!

But, and this is my main point, we can at least try to keep the game at a reasonable level that is consistent with the game's expressed content: 6000 years of political, cultural, technological, economic, social and military events. That means the gamer cannot play at the level of an individual General or Colonel, no matter how much he may want to.

I happen to think that the game cannot do that completely: without Some input to the outcome of the battle, gamers will be, I suspect, extremely dissatisfied with the 'military' portion of the game. That's why I posited using Dupuy's Postures as gamer input to how a battle would be fought, giving said gamer some say in what his Units and stacks will be attempting to achieve. I think this can be argued as appropriate given the amount of influence politicians and 'Higher' civilian leaders have had, or attempted to have, on military operations, from Louis XIV and his successors and predecessors to Churchill, Stalin and Hitler's baleful influence on operations in WWII.

Sorry about the Dupuy books. This is a problem when you are older than the proverbial Dirt, as I am: I have copies of Dupuy's works and numerous other military history works that I bought while they were still new or at least In Print 'way back when. AbeBooks | Shop for Books, Art & Collectibles, my 'go to' on-line bookstore, lists copies of Numbers, Prediction and War used for around $25 American, but I don't know of any free copies on line and the equations and information charts run for pages. I'll see what I can do.

 

[aieeegrunt]

Here is how 3 UPT works for the amusingly sarcastically named “Stacks of Doom” mod for Civ5, and basically the same for the ARS mod for Civ6

Combat is you pick the attacking unit and either assault (melee) or bombard (ranged) it. Strongest unit is automatically selected to defend

That’s basically it. This does an excellent job of simulating hex combat at operational and above levels of scale, which is why just about all of the old consim games that didn’t use combat odds tables did it that way.

Simple, clean and effective

 

[historix69]

Here is how 3 UPT works for the amusingly sarcastically named “Stacks of Doom” mod for Civ5, and basically the same for the ARS mod for Civ6

Combat is you pick the attacking unit and either assault (melee) or bombard (ranged) it. Strongest unit is automatically selected to defend

That’s basically it. This does an excellent job of simulating hex combat at operational and above levels of scale, which is why just about all of the old consim games that didn’t use combat odds tables did it that way.

Simple, clean and effective

There is the rule of thumb that the attacker should have 3:1 superiority in combat strength to have a good chance to break through and push the defenders out of their position.
How do you have 3 units attack simultaniously a defending unit?

What you describe seems to be mostly a quality-of-life mod to remove the (annoying) traffic jams, but (melee) combat seems to be still limited to single one-unit-against-another-unit battles as in Civ 1-6 before. The stacking however allows to move more ranged units into range to soften the target before a single unit attacks to break through, so ranged units like artillery become even more important and OP.

 

[Evie]

In a turn based system where you can freely attack with all your units before opponents have any chance to react - you should have to work hard (ie, attack from multiple tiles at once) to achieve that three on one superiority that mostly guarantees a breakthrough.

Otherwise it's too easy to overwhelm an oppnent before they even have a chance to fight, and that's not a desirable outcome.

 

[aieeegrunt]

There is the rule of thumb that the attacker should have 3:1 superiority in combat strength to have a good chance to break through and push the defenders out of their position.
How do you have 3 units attack simultaniously a defending unit?

What you describe seems to be mostly a quality-of-life mod to remove the (annoying) traffic jams, but (melee) combat seems to be still limited to single one-unit-against-another-unit battles as in Civ 1-6 before. The stacking however allows to move more ranged units into range to soften the target before a single unit attacks to break through, so ranged units like artillery become even more important and OP.

You have three units attack that hex in succession.

Just like you wpuld in 1UPT

Except you don’t have to solve a sliding tile puzzle to move your units problem

Of course if you focus all your attention on one enemy stack you are leaving other ones unengaged

In a turn based system where you can freely attack with all your units before opponents have any chance to react - you should have to work hard (ie, attack from multiple tiles at once) to achieve that three on one superiority that mostly guarantees a breakthrough.

Otherwise it's too easy to overwhelm an oppnent before they even have a chance to fight, and that's not a desirable outcome.

Ranged units being OP is one million percent because they have multi hex range. Which you have to have in 1 UPT for obvious reasons.

Having 3 UPT means you don’t have that design constraint.

The turn based system thing you mention also applies to 1 UPT, which also has the added benefit of solving a sliding tile puzzle when moving units AND having to have multi hex ranged units

Of course if you are focusing all your attention on one enemy stack, that leaves the rest of your enemy stacks unengaged and free to return the favor

 

[Boris Gudenuf]

The old '3-1' rule turned out to be almost completely obsolete during World War Two and since. Mainly because too many armies (the US Army for a prime example!) used numbers instead of actual combat power when calculating '3 - 1'. Superior firepower, maneuverabilty, reaction time, protection (as in field fortifications or camouflage), air support, any or all could skew the equation to one side or the other.

At one extreme, in front of Moscow in October 1941 the Germans frequently attacked successfully with inferiority in numbers but a massive superiority in tactical doctrine, automatic weapons' firepower and heavy infantry weapons. Most of the Soviet infantry forces were militia or raw recruits and replacements, so there was also a massive dissimilarity in training and combat experience. Where there wasn't, as at Ilynskoye when the Germans hit a pair of officer schools or Borodino where they hit the 32nd Rifle Division, a veteran full-strength unit from the Far East, the result was a bloodbath on both sides and a breakthrough only after the attacker bludgeoned his way through over several days.
At the other extreme, in February 1945 in the fighting at Königsberg, 2 Soviet rifle platoons attacked with the support of 2 combat engineer platoons, a battery of 122mm self-propelled assault guns, a battery of 45mm antitank guns, a battery of 280mm siege howitzers, and several battalions of 76mm - 122mm field artillery: the numbers of men actually going forward was the least important component of the combat power massed for the attack!

The problem is, combat power, the actual factor the 3 - 1 equation, is a combination of physical firepower and maneuverability, and 'soft' factors like training and morale and the command's ability to react and get inside the other side's decision loop. And all the combat takes place, in Civ terms, within a fraction of a single turn, so somehow, I would hope, has to be modeled as a component of the campaign/overall battle rather than dropping the gamer down to the tactical level and slowing the rest of the game to a halt (the Humankind tactical battle problem).

Basically, most of the Combat Power should be 'baked in' to the Unit, a function of the weapons available, the training undergone (Professional or Amateur Units), and the doctrine of the unit or army. The Command Influence during the battle is the other component between the Unit and the Gamer, and really should be separate from the gamer's decisions, represented by a combination of Great Generals and Posture in which guided by a general command from the Gamer the invisible digital Army Commander runs the battle.

 

[Evie]

My criticism wasn't so much directed at the 3 UPT (it's a workable approach), as to the idea that the game need to enable players to achieve the kind of concentrated attack that historix69 wants.

I'd suggest that in Civ 6, *armies* are supposed to represent that whole "bringing a 3-1 advantage over the opponent".

 

[historix69]

Ranged units being OP is one million percent because they have multi hex range. Which you have to have in 1 UPT for obvious reasons.
Having 3 UPT means you don’t have that design constraint.

If ranged units in 1upt are OP, then they should be even more OP in 3upt since you can have 3 ranged units (artillery) per tile ...

I'd suggest that in Civ 6, *armies* are supposed to represent that whole "bringing a 3-1 advantage over the opponent".

Yes, I think that's why they introduced the Corps of 3 basic units.


In general I think that the devs should experiment with a division/corps/army model similar to HOI series where the player commands units which consist of dozens of smaller battalions or batteries adding specific combat values and boni, allowing the player to build different kind of formations to solve specific problems on the battlefield, eg engineers to cross rivers, massive artillery to break through, mechanized units to have higher movement rate, etc.

 

[Boris Gudenuf]

If ranged units in 1upt are OP, then they should be even more OP in 3upt since you can have 3 ranged units (artillery) per tile ...



Yes, I think that's why they introduced the Corps of 3 basic units.


In general I think that the devs should experiment with a division/corps/army model similar to HOI series where the player commands units which consist of dozens of smaller battalions or batteries adding specific combat values and boni, allowing the player to build different kind of formations to solve specific problems on the battlefield, eg engineers to cross rivers, massive artillery to break through, mechanized units to have higher movement rate, etc.

Let me throw out a couple of suggestions for discussion here.
First, I would really like to see 'combat' occur within the confines of a single tile, thus avoiding the grotesque out of scale unit attributes and battles we have in Civ VI (archers shooting across a major city, a single unit of a single weapon-type taking up 100s of square kilometers, etc).
So, in order to make that work without the entire combat system devolving into some form of SoD, we need to emphasize In The Combat System all the potential variations, not just in the weapons types, but also in the relationship between the weapons-types.
Examples:
Spartan Hoplites were virtually unbeatable on an open plain, the typical Greek battlefield. But in one famous incident, they were caught in rough, uneven country by a force of lighter, javelin and sling-armed troops and cut to pieces - they couldn't catch their lighter enemy, and had no distance weapons to oppose the enemy fire.
Alexander the Great's army had heavy infantry armed with pikes (the Pezhetairoi), faster infatry with spears and swords to provide a link between the heavy infantry the the cavalry (Hypaspists) and heavy shock cavalry that charged with a heav lance (Hetairoi - the 'Companion Cavalry'). While individually the troops were good, collectively they were simply unbeatable in every major battle against everyone: Greek hoplites, Persian cavalry, Scythian horse-archers, Indian elephants and archers.
Tanks are notoriously useless by themselves. The whole point of the Panzer Division of WWII was that it combined tanks with mobile infantry, combat engineers, antitank, antiaircraft, and artillery all able to keep up with the tanks. The combination was many times more effective than any single type of weapon or unit that comprised it.

Now, let's take a look at Humankind: every single type of Unit in that game has a separate specialized attribute - something it does a little better or worse or different from all the other Units. This can tie in with multiple units per tile: the interaction of the historical units and armies can be automatically calculated by the computer so that your 'stacks' are never Uniform piles of combat factors: 4 units of Hoplites might be invincible on a flat plain, but they are much less effective on hills, forests, or almost any other kind of terrain, and get shot to pieces by Horse Archers that they cannot catch or shoot back at

Interactions like this among the Unit types will keep the game's combat system interesting, even if all the combat takes place semi-invisibly on a single tile: it's your job as Great Ydvig of the Slobbovians to produce an army of units that can successfully apply their interactive bonuses against the enemy, not to individually order each unit into battle.

Now as to Corps, Armies and such. This, to me the military historian, reflects increasing the Span of Control of the digital General. Even an Alexander could not give orders to 12,000 individuals. He could, however, give orders to the commanders of 6 Taxeis of Pezhetairoi and have all 12,000 pikemen in those units do pretty nearly what he wanted (never entirely: "War is the province of uncertainty" and Clauswitz's "Friction" in combat always have their say in matters).

But it means that, for instance, if the right criteria are met, you could combine separate units of Tanks, Motorized Infantry, Artillery, and Combat Engineers into a single unit - let's call it a Division Cuirassier, or Armored Division, or Panzer Division - and that would now be a single Unit for stacking and combat, with a whole new set of bonuses and attributes compared to the individual units that made it up. You might be allowed, by the Modern Era, to stack, say, 6 - 8 'Units' in a tile, but with better Span of Control and Brigade, Division, Corps units you might actually be stacking up to 40 individual 'Units' in pre-built configurations.

Now, of course this also requires rethinking the requirements to build units, because potentially 'armies' are going to consist of many, many more units than now. Greater concentration of 'units' into single tiles also changes the relationship between Units and the maps - for the better, because genuine maneuver is now possible since in most cases you cannot Blanket the Border with 1UPT on every tile and hold for turn after 1 - 10 year turns: 1 Unit on a tile will get wiped out by a genuine concentrated Army in a fraction of a turn, and that army will keep moving until it meets a 'real' obstacle.

 

[aieeegrunt]

If ranged units in 1upt are OP, then they should be even more OP in 3upt since you can have 3 ranged units (artillery) per tile ...



Yes, I think that's why they introduced the Corps of 3 basic units.


In general I think that the devs should experiment with a division/corps/army model similar to HOI series where the player commands units which consist of dozens of smaller battalions or batteries adding specific combat values and boni, allowing the player to build different kind of formations to solve specific problems on the battlefield, eg engineers to cross rivers, massive artillery to break through, mechanized units to have higher movement rate, etc.

You are looking at isolated elements without considering how a system works as a whole. It would save a whole hell of a lot of time if you loaded up one of these mods and tried it yourself

A properly designed 3UPT system for example, would give ranged units a move of 1, have them not be able to move and shoot on the same turn, and give them virtually no defense.

So a mob of unsupported artillery moving up to a properly balanced enemy force would suffer the same fate as one would in real life; the enemy force would be mildly puzzled at such foolishness, and then simply advance and wipe them out at little cost.

This is yet another thing 3 UPT brings to the table that 1 UPT cannot; force composition and combined arms matters.

A three stack of artillery, sure it can dish out a lot of punishment but it’s basically naked. One melee or cavalry and two artillery is a lot of firepower but pretty limited sustain. A cavalry, melee and artillery is well balanced but a jack of all trades is master of none

This is clear, simple and fits the theme and design style of Civ

Something like a HOI system is way way too complicated and micromanagy for the scope and scale of combat in Civ. It also brings nothing to the table that a 3 UPT does not, and still suffers the drawbacks of a 1 UPT system; this is just 1 UPT with a layer of unit customization on top

You might as well bolt on a logistics system from OCS or Campaign for North Africa while you are at it

 

[Marla_Singer]

- Civ 1 had the rule that the attacking unit and the best defender unit of a stack simply made a duel and if the attacker won the attacked stack commited suicide and vanished. There was an exception for city defense and fort defense where a defeated defender died alone and the defending stack did not die as a group.
- (I don't remember much of Civ 2 besides the fact that combat was broken up into more single steps of RNG combat to reduce the number of tanks and battleships being destroyed by fortified Phalanx units on mountain tiles.)
- Civ 3 separated combat so that the attacking unit and the best defender unit against the attacker (unit type boni) made a duel and if a unit died other units were not affected as long as the defending stack still had a defender. I remember that ranged units with movement points could support the defender by bombarding the attacker before 1st round of combat. If the last defender of a stack was defeated, remaining artillery units were captured, leading to insane stacks of about 100 artillery units in late game. (Later patches added shortcuts to move whole stacks of a unit type with a single command.) There also was the feature that units fortified near a road automatically sniped on enemy units passing by (probably instead of Zone of Control). Ranged units could damage defenders down to 1 HP but not kill them.
- Civ 4 was similar to Civ 3 in combat except that artillery type units now sacrificed themselves when attacking and could damage several units in the defending stack.

Good summary however you missed the big change of Civ 4, which was to replace "attacks" and "defence" points by a single "strength" criteria, varying according to unit types involved. That is commonly referred as "rock, paper, scissors": a mounted unit defeats a range unit, a range unit defeats a spear unit, a spear unit defeats a mounted unit.

The same logic was applied in following games, but lost in importance with 1UPT, as the better strategy largely consists since then in having a first row of melee units and a second row of range units and that's about it. However, in Civ4, wars were mainly occuring over a city, with an attacker trying to seize it to a defender. 1UPT solved that in bringing warfare to the open field, making terrains more meaningful.

Both Civ4 and Civ5/6 combats have their pros and cons. The question I'm wondering is whether it's possible to get the best of both worlds.

 

[Civinator]

I'd personally love to see something combining Humankind's approach with your AoW - style approach, where we send small stacks across the map in one click instead of a hell of taking 30 units across the entire world one after another, and then for something interesting happening in the form of a field battle, and it would solve sooo many problems of civ5-6, I just have no idea how to do that without falling into the pitfalls of super messy HK combat design, while still making it interesting.

With Civ 3 (C3C) it is possible to move stacks of different units (and also stacks of the same kind of units) with one click (but may be several turns of the game) over greater distances on the same continent of a map.

 

[historix69]

The question I'm wondering is whether it's possible to get the best of both worlds.

(The following is more or less a summary of ideas already posted before ...)

I don't think that it will be possible to go back to stacks without annoying 1upt and "rock, paper, scissors" fans.
Limitations for number of units per tile would be mostly based on logistics and economics and I doubt that the devs will ever implement realistic logistics and costs for warfare since logistics is no fun.
Having 2 Levels of Detail like a strategy map and a more detailed tactical map probably also won't be implemented because moving much more small units on a huge tactical map will distract players too much from the strategy game.
Having bigger stacks would suggest to implement group combat (n attackers against m defenders) rather than individual combat (one against one) but again I doubt that devs will ever move away from the current approved system of individual combat of attacker against (strongest) defender since defining consistant and acceptable group combat rules for "rock, paper, scissors" is almost impossible.

Best solutions imho :
- Allow players to define their own individual stack limits like 1upt, 2upt, 3upt up to unlimited upt depending on category (melee, ranged, civilian, navy, air,...). Allow to stack units in cities. 1upt traffic jams are no fun.
If the devs won't do it, the players will mod this feature anyway ...
- Use single combat of 1 attacker against (strongest) defender ("rock, paper, scissors")
- Allow players to balance unit production costs. 1upt usually has rather high unit costs to limit number of units on the map. Bigger stack limits allow games with more and cheaper units.
- Allow stronger units like armies or corps units based on 3-4 basic units (with appropriate production costs and upkeep).
- Provide the default UI features to select, sort and move units (of a type)/groups/stacks ...
- Adjust unit movement so that units on easy flat terrain or road can travel maybe at least 4 tiles per turn.
- Allow players to define (individual) flanking boni for combat so that occupying neighbouring tiles is key to success.

Optional : Make unit upkeep (base costs + supply) dynamic based on terrain, improvements, number of units in a tile and other circumstances.
(E.g. lower costs for tiles with city, fort, barracks, harbor, road, railroad. Higher costs for healing units, jungle, ice, desert, mountain, neutral and enemy territory, crowded tiles with too many military units.)
(This is to some degree redundant with general limitation of stack size and in general probably will be no fun for most players. I am also unsure how AI would deal with these rules since economic costs are often ignored by AI.)