Combat in Humankind vs. Civ V/VI

[Amrunril]

I'm curious how other players feel about Humankind's system. While I feel like this was put forward as a major selling point of the game, I don't personally see much to recommend it over Civ V and VI's system. Both revolve around a mixture of melee and ranged units engaging in one unit per tile tactical combat, spread across the game's overall strategic map. While there are plenty of more subtle differences, Humankind's biggest adjustment is, effectively, to allow stacking outside of combat. While this is extremely helpful in cutting down unit micromanagement out of combat unit micromanagement, it also introduces several significant issues:

1. Thematically, armies consolidating for movement and spreading out for battle is pretty much completely backwards. A smaller military unit can generally move faster and keep itself fed more easily, so the standard practice for most of history was for armies to spread out while moving but consolidate for battle.

2. On a mechanical level, the need for battles to have discrete beginnings and ends creates a number of problematic artifacts

• Deployment zones for battle are difficult (or at least unintuitive) to predict. This can result in one army being stuck in too small an area to fully deploy and makes stationing an army in a defensive position very difficult.
• The attacker can usually strike on the first turn of combat with most or all of their units. This can be a massive advantage when using large numbers of ranged units.
• Most importantly, a battle usually ends in the complete destruction of one army. This is obviously ahistorical, and I think it contributes significantly to the game's snowballing tendencies- if winning a battle entails obliterating the opposing army, then the risk of going on to capture cities is dramatically reduced.

3. While stacking units out of combat decreases micromanagement per unit, this seems to have led the developers to balance around increased numbers of units, meaning that, in the context of actual battles, "carpet of doom" dynamics can actually increase relative to Civ V and VI.

 

[8housesofelixir]

The "everyone is ranged" is probably the best feature of HMK's combat; and the "consolidated Army" might satisfy many 1UPT-lovers as well (although I'm not one of them).

On the other hand, the Tactical Map, albeit interesting to deal with and work with, contributed to most of the issues of the combat system, namely
1. Attacker's advantage. This one is not easy to deal with though; the recently added minimum damage changes might be a nice direction to tweak it. Many also argued for a "retaliation" ability for Gunners and Naval units as well.
2. Land and naval combat sometimes disconnect with each other, due to the limited coverage of Tac Map and there is only one ship that can do Bombardment. Air combat can also feel disconnected, as Air units have some strange behaviors when included in Tac Map.
3. Battle usually ends in the complete destruction of one army, as one cannot escape from the Tac Map unless a battle is won or lost. I understand that a "battle" basically means "an engagement you cannot rescue yourself from" in HMK, as the opportunities to retreat is before the battle; but the retreat is costly for War Support, and most of the battles will become a total annihilation which is at least not very realistic (before WWI, it was very common for half of the losing side of a battle fleeing from the battlefield, instead of becoming Romans in Cannae or Zhao Army in Changping) and favors the "winner" too much.
4. Tactical Map, once expanded, will block everyone else from coming in, and cities inside it. Blocking everyone else from joining in the battle is unrealistic to begin with, albeit I can understand the coding limitations. Blocking cities can result in the issue of, Empire A is fighting near the border of Empire B, and B's city becomes inactive due to get caught in the Tac Map; even if Empire A is just hunting a Bear or a Deer, B's whole city will be shut down for at least 1 turn.

As one can probably tell, Civ doesn't have issue 2-4 because it doesn't have a Tac Map to begin with. For HMK, the issue being Tac Map is sometimes not interacting properly with other game mechanics. I do think these issues can be tweaked via some mechanic changes, such as figuring out the air support within a Tac Map and allowing Gallipoli-style mid-battle retreats, to have the Tac Map better included within other game mechanics.

 

[9gkj548hkm20vk79dj689bnj4]

I think the intention of Humankind's combat was indeed to limit the number of stalemates and/or battles that continue for years (in-game). No more Kheshigs shooting and running away again; you must go "all in".
So that's why there's no mid-battle retreat option, and why attackers always seem to have an advantage because they can just pummel the weakest units immediately. Of course, as you say, the game gets around this by simply making units relatively easy to build after they are lost.

I don't know if that's really "better" than what Civilization did or not. It's just a different way of handling combat.

Other than that, I think a lot of your problems mirror my own.

 

[Aristos]

HK battle system is eons ahead of any version of civ. Granted, it has some issues, but they are not game breaking and will be balanced and solved sooner rather than later. The best part is that the AI can actually use it, maybe not perfectly yet, but can use all subsystems within the system and may even surprise you once in a while; it will only become better with time. No comparison to the lame, non-existent Civ 6 AI.

HK battles combine the best of both the strategy layer (you have to somewhat preplan your battle plan, reinforcements, positioning, terrain matters, etc) and the tactical layer. It solves the relative problem of the SoD and the horrible problem of the CoD in one big shot. A similar system should have been in Civ since the late 90's after Call to Power showed the way, but Firaxis was too busy not listening.

 

[Taefin]

I think the intention of Humankind's combat was indeed to limit the number of stalemates and/or battles that continue for years (in-game). No more Kheshigs shooting and running away again; you must go "all in".
So that's why there's no mid-battle retreat option, and why attackers always seem to have an advantage because they can just pummel the weakest units immediately. Of course, as you say, the game gets around this by simply making units relatively easy to build after they are lost.

I too suspect that the no retreat was intended to destabilize wars so that one side captured some cities before war score hit zero. Unlike Civ6 and Old World, the AI is very reliable at capturing one another’s cities.

But the cost might be too steep if Amplitude cannot find a way to better counter the player.

In ancient to early classical the AI does pretty well at fielding a large army of often superior units, so that as they throw stacks against you you win several even fights, but another comes right after it, creating some action and peril. But as soon as your army gets bigger than the 3-4 stacks they are willing to move together, then the game becomes a short string of lopsided battles and much less enjoyable. All the more because by medieval/early modern, the AI refuses to build back up past a stack or two.

I hope Amplitude will change something. It seems like it should be simple enough (build more units).

As to the combat mechanic itself, I do actually find it overall more enjoyable than Civ 6. Certainly changing infantry to ranged gunners with LoS adds a lot to the immersion of transitioning to gunpowder warfare. I find the tac map makes gameplay flow better, getting to focus on one battle at a time battle rather than managing multiple, and cities, between each moves. It also seems to allow larger movement ranges to feel natural (even though it’s the same terrain, my impression is the tac map is meant to represents a much smaller scale than the strategic map).

I will forever insist that midbattle retreat is the only way the system can reach its fullest, even if this would require rebalancing AI to conquer each other again. One idea is to fix what is stopping sieges from working. Currently every siege is either an immediate assault or a sortie (that becomes an assault). With 6 catapults I bet the AI could conquer a city.

 

[Gedemon]

A lot of good points, but IMO none of them make civ4-5-6 combat looking good in comparaison (civ4 is used as a reference not because it's good, but because civ5-6 are worst), I do think HK do it far better, and that's coming from someone who was not convinced by the deployment from stacked strategic to 1upt tactical map for combat when it was discussed in Idea & Suggestion before HK was announced, and even after that announcement, I still wasn't convinced until I actually tried it.

Now, I agree with most issues, and I also think most can be solved, there is just one point I disagree with:

1. Thematically, armies consolidating for movement and spreading out for battle is pretty much completely backwards. A smaller military unit can generally move faster and keep itself fed more easily, so the standard practice for most of history was for armies to spread out while moving but consolidate for battle.

There is a major change of scale between the strategic and the tactical representation, when they are stacked out of combat the armies are moving on tiles that are for the smallest (meaning at map max sizes) representing a ~100x100km area, they can be as spread out as you can imagine in such an area. They actually consolidate for combat, at a much smaller scale in which individual formations (archer, melee, ...) can be represented.

 

[Pfeffersack]

I overall like HKs approach here - and especially the AI being capable of using the mechanics of tactical battles.

My criticism is more directed to balance, visual things and interface:
- overall, the battles feel a tad too lethal (units do a too much damage each round)
- the formula for how combat strength affects damage feels a bit rough (and someone has already done an analysis what is wrong with it)
- the coloured overlay marking the battlefield looks ugly and I often have troubles to find out, how terrain elements affect a fight between two units
- maybe I just haven't learned the game here, but I have troubles to predict where the flag will appear when engaging a battle or to determine which armies are in range for serving as reinforcements

 

[Don Andre]

I've had a couple of battles that ended in sort of a draw. When you're defending and you don't use your units to attack, because they're too weak, sometimes the time is up before the other play has captured your flag and both sides retain some units. However, there is just too much time with larger armies and this happens only in smaller skirmishes, in earlier eras.

In any case, I feel that the instant resolution and the manual battle are miles apart. The total strength is not at all an indication of who's winning. It also depends a lot on who got the first move (especially with ranged units). I often use manual battle to minimize my unit losses. My main gripe is that some battle turns take extremely long - for instance the hunnic hordes. There are often more than 5 units and it takes time until they approach, fire the shot and move back to safety again. But manual battle becomes very tedious if you're having to do a lot of them.

What I really liked in my last game was the battles between the Hunnic horse archers and my Persian immortals. The immortals could insta-kill a horse archer and thus were quite apt at fending of attack waves of the Huns. In the end I even managed to win a war that was totally against my odds and that was a fun moment. Basically, I had to deal with -3 war support per turn and no per turn war support loss on the Huns. So I kept my head over water by winning battles until I finally managed to take one of his cities. After that I could manage to get my war support a little above and won as he lost -4 per turn and I was losing just -3. Had a lot of 3-star units after that long war.

The line of sight restriction is pretty annoying for some heavy weapons. It's not entirely clear which tile is visible / attackable from which other tile. Sometimes you have to prioritize movements so that there is a window of attack for the artillery, before you close it again with another unit and block the LoS. Aren't cannons like the Howitzer ballistic? I know there is this red attack cross which is brought up when you try to move your units to different positions, but nothing like that is displayed during deployment phase and nothing is put for tiles not containing units.

Sometimes the own deployment zone is very tiny.

 

[KayAU]

I like the idea of the system, and I think it is an improvement over Endless Legend, but I really wish they had done something different with the "tactical map". Other games have had army systems before, the semi-recent examples I keep bringing up are Fallen Enchantress and Age of Wonders 3. These both used separate maps for combat, which ensured that you had enough space and allowed the designers to create something which worked well for tactical combat. Spreading the units out on the strategic overview map and designating part of that as the tactical map doesn't work nearly as well, in my opinion.

If the point is to make the tactical map not seem like a completely different place to the strategic map, this could be solved visually by just zooming in on the tile where combat is taking place, and subdividing it into an appropriate number of sub-tiles to create the tactical map.

 

[Siptah]

The line of sight restriction is pretty annoying for some heavy weapons. It's not entirely clear which tile is visible / attackable from which other tile. Sometimes you have to prioritize movements so that there is a window of attack for the artillery, before you close it again with another unit and block the LoS. Aren't cannons like the Howitzer ballistic? I know there is this red attack cross which is brought up when you try to move your units to different positions, but nothing like that is displayed during deployment phase and nothing is put for tiles not containing units.

For all the attacks that LoS got at release, my gut feeling is somehow pretty accurate. Not perfect, but maybe 95% of the time. Not that it is optimal that a player needs to fight dozens of battles until the intuition is there. But it was an interesting learning process how and when it works, and how to make the best use of tactics and unit movement. Still, I would love to have the red attack cross when deploying units, that's a great idea!

Sometimes the own deployment zone is very tiny.

Yeah. And I wish I would know why. I mean, sometimes, it's obvious that you are cornered or in a valley or something. But I also had battles in "standard" hilly terrain in which I couldn't deploy 4 units...

- maybe I just haven't learned the game here, but I have troubles to predict where the flag will appear when engaging a battle

This also puzzles me. It's especially annoying when you are sieging a city and the defender makes a sortie. I often end up with a flag in a depression next to the enemy walls, which is just such a bad place to station units. Also, at least once, I had a flag that couldn't be reached from within the battlefield, as the route around the cliffs was too far away. You needed reinforcements from the back to capture that flag...

 

[Taefin]

For all the attacks that LoS got at release, my gut feeling is somehow pretty accurate. Not perfect, but maybe 95% of the time. Not that it is optimal that a player needs to fight dozens of battles until the intuition is there. But it was an interesting learning process how and when it works, and how to make the best use of tactics and unit movement. Still, I would love to have the red attack cross when deploying units, that's a great idea!

Haha explain LoS around rivers to me

My main gripe is that some battle turns take extremely long - for instance the hunnic hordes. There are often more than 5 units and it takes time until they approach, fire the shot and move back to safety again.

Switching to 3x animation speed really helped me tolerate this. Brought back the simultaneous feel where stuff moves as you click around the battlefield.

 

[NombreyApellido]

combat zone size and deployment areas are another component of the blatantly "cheater" AI... for instance, you'll notice that when in areas of ridges the cpu players will get access to upper levels on the first round, or that they get one forward deployment tile on an upper level, or simply how fast they can trigger a combat seemingly moving faster than the locked unit animation speed

but the most egregious one to me is how after you deploy on a siege that of course is now a sortie for them, they are able to grow the combat zone AFTER the pawns are set so they can now flank units you had at the limits or they can go somewhere elevated previously unaccessible

--

hordes being able to move tile by tile during combat instead of all at once fells out of place. it should be available for all, to avoid pathfinding issues

 

[Don Andre]

Switching to 3x animation speed really helped me tolerate this. Brought back the simultaneous feel where stuff moves as you click around the battlefield.

Iirc that was added with a patch and I am still at release version on stadia... Too bad

 

[7daysofwar]

Iirc that was added with a patch and I am still at release version on stadia... Too bad

It was never added in a patch, I have my animation on 3x since I bought the game. It is just somewhere in your settings.

 

[Buktu]

It was never added in a patch, I have my animation on 3x since I bought the game. It is just somewhere in your settings.

Then again the Gamepass version of the game misses the feature to rename your cities (at least thats what I read somewhere) so maybe there is something "missing" in the stadia version too?

 

[ThunderLizard2]

HK battle system is eons ahead of any version of civ. Granted, it has some issues, but they are not game breaking and will be balanced and solved sooner rather than later. The best part is that the AI can actually use it, maybe not perfectly yet, but can use all subsystems within the system and may even surprise you once in a while; it will only become better with time. No comparison to the lame, non-existent Civ 6 AI.

HK battles combine the best of both the strategy layer (you have to somewhat preplan your battle plan, reinforcements, positioning, terrain matters, etc) and the tactical layer. It solves the relative problem of the SoD and the horrible problem of the CoD in one big shot. A similar system should have been in Civ since the late 90's after Call to Power showed the way, but Firaxis was too busy not listening.

Agree - a real step forward from where we've been with Civ V and VI 1 UPT and the crazy unit bottlenecks. It needs refinement for sure but very good foundation.

Then again the Gamepass version of the game misses the feature to rename your cities (at least thats what I read somewhere) so maybe there is something "missing" in the stadia version too?

That makes sense - you're getting cheap access so it's good they remove certain extras.

 

[Don Andre]

It was never added in a patch, I have my animation on 3x since I bought the game. It is just somewhere in your settings.

Found it, thx! Better now.

 

[Jkchart]

combat zone size and deployment areas are another component of the blatantly "cheater" AI... for instance, you'll notice that when in areas of ridges the cpu players will get access to upper levels on the first round, or that they get one forward deployment tile on an upper level, or simply how fast they can trigger a combat seemingly moving faster than the locked unit animation speed

but the most egregious one to me is how after you deploy on a siege that of course is now a sortie for them, they are able to grow the combat zone AFTER the pawns are set so they can now flank units you had at the limits or they can go somewhere elevated previously unaccessible

--

hordes being able to move tile by tile during combat instead of all at once fells out of place. it should be available for all, to avoid pathfinding issues

I'm not sure how the game decides how big a zone is, but I've had situations where an army that misses the chance to end up reinforcing that first turn and is outside the battle zone gets included anyway, so i don't think it's just the AI "cheating". There's some really funky logic that goes on behind the scenes where things end up not making sense with deployment. And when a new army does get added through reinforcements, tiles do get added to the battlefield and it slowly gets bigger (which is something I recall from a marketing video too).

I agree that the instant "I get to turn the assault into a sortie" is BS though, the attacker should at least get to first decide if they want to maintain the siege or assault (or back off) before allowing the defender to choose to surrender, wait, or sortie.

 

[NombreyApellido]

i'm starting to believe all of these "cheats" are in place to help the AI because it's designed with a very small allotted time budget to make decisions to fit to the simultaneous turns design... in combat, i've seen the AI being able to choose to stay behind walls when there are only mounted units on a siege, and a few times move into a defensive position when in clear disadvantage. but mostly it plays exactly the same: it will send its strongest unit after your weakest, including ranged fire and charge in suicidal attacks that have no chance of succeding

on other aspects of the game, it's even more evident that it's clueless

in city building, it will place powerfull cultural wonders in awful locations negating their foorprint exploitation

in trade, it may choose a weak lux instead of a very powerful one... not only it will offer to trade when there's nothing to, but will do it while at the same time attacking you and triggering a crisis

it will choose cultures in easily recognizable paths that keep no relation to the map benefits (i've seen samiramis, an expert type persona choosing phoenicians in my pangea seeded map)

so i'm coming to terms with the idea that even if we end user somehow come up with the perfect balance formula in a mod, the underlying AI systems won't be able to play the game at all because of their design constraints

 

[Taefin]

so i'm coming to terms with the idea that even if we end user somehow come up with the perfect balance formula in a mod, the underlying AI systems won't be able to play the game at all because of their design constraints

If unit training priority is moddable then I expect some very challenging AI will be attainable through mods. The one time in opendev that I ran up against 30-40 muskets and dragoons did not end up well for me. Sure the player can overcome swarms of pre-gunner units (and I think that will always be part of the fun of beating a heavily advantaged AI) but once you start trading shots with gunners, the AI behavior feels far more optimal. As I see it, the only problem now is that starting in medieval, the AI won’t train half the units the player does and they fail to upgrade to partisans which could actually defend their cities.