So, after having played after around 10 games of Civ 6, my conclusion is . . .

[ghost88]

I find that an interesting idea, especially given my thoughts on the military side of the Civ franchise. I’ve always sort of looked at the warfare side of it as a bit metaphorical. All the other aspects of a Civ game (the modeling of governance, scientific development, economic development, religion, etc.) have always been the parts of the game that have held my deepest interest-even though I have also always appreciated the need for a robust representation of military struggle.

It might sound counter-intuitive to many, but I feel that this reflects how real life works. Despite the fact that there have always been wars-big ones, small ones, devastating ones, etc-I don’t think human society and achievements revolve around warfare. It might look like that, but I actually think Warfare is a reeeeaally big and prevalent side-show to the main event: all the other peaceful pursuits that I mentioned. I look at it this way: we don’t work and live in order to war, we war in order to work and live. In order to work and live in a certain way-the way we want to work and live.

Having said all that, I will also say I can hear the screams of protest from so many, in my ears right now, were the military layer of Civilization abstracted down to such a level.

Even I think that the answer lies somewhere between the way things are now and the way things would be with your proposal.

But I like your general idea though.

Just off the top of my head things that came about because of the military necessity, Malaria and Yellow Fever Vaccines, Anesthesia and Anti-Biotics.
The Worlds militaries have been at the forefront of tech discovery or innovation since Humans first left their Caves. To say that War has not influenced the growth of Civilization is to show a outstanding lack of understanding history. Every great civilization/ empire in history got there on the shoulders of their military.
Without the Military aspect of Civ you might as well be playing Sim City, but at least they would have little trouble programing this games AI to hand you your Fourth point of contact on Settler difficulty.

 

[Orup The Great]

I was never attempting to argue that warfare has not deeply influenced history, achievements, discoveries, and itself has not created new eddys of technology, currents of thought or human pursuits.

What I was trying to say was-at least in my opinion-for all those ways in which Warfare has contributed to all these things-Human Beings, by and large, do not make war for the sake of itself. Warfare is far too bloody, wasteful, disruptive and risk ladden for that to be true.

We have made war for reasons, and those reasons tend more often than not to amount to protecting-or being able to project-a way of life. Very, very few of human kind are so sick that they simply wish to engage in warfare just to kill. We have killed for something. Those “somethings” have been things like: our religion, our desire to gain resources (economy), our desire to be stronger than our enemies (technology), our desire to spread (or protect, or consolidate), our government. Our . . our . . . our etc. (again: my opinion).

And, it’s those things that we have tended to fight over that are more interesting to me, anyway. Which is why, in principle, I don’t mind a warfare system being abstracted, seeing as I am neither interested in warfare per se. I’m interested in general Struggle, and the reasons why we’re struggling.

Therefore I fully embrace a military layer to this wonderful game because Warfare is an enormous part of history, and has-as a tool-driven so much of history. That it has driven history as a reason unto itself (i.e.: Tribe A masses on the steppes, and storms into the land of the rich Urban Neighbors B-just for the hell of it), well-though it might seem to some as though that's the way history has worked-as a life-long historian, I personally would have to politely reject that notion.

Hope that clarifies things.

 

[Arithmion]

For what gain? A tactical layer that is totally off topic in a Civilization game, and the faculty to see a carpet of beautifull units.. so great! ^^

Off topic in a Civ game, perhaps, but relatively standard for Sid Meier games. (Pirates comes to mind. The whole "fortress storming" chess game felt out of place, but I did enjoy it.)

On an unrelated point, I like to think I can speak relatively well for the community Firaxis is trying to attract: I'm a younger guy, an adult that can afford to buy a game here or there, but none of that crazy "family" stuff I've heard so much about to get in the way. I played Civ IV, but was way too young for it and it left a bad taste in my mouth as a result. (How do I make it not "Too crowded?!") Only just recently learned that was just the population counter unhappiness that I was supposed to be countering. I came back with civ 6, bought the base pack, liked it, bought the expansions, and have become a civ fan, and a King/Emperor level player. I never liked the doomstacks, but that's probably just because I stank at the game. I do enjoy the AI in this, trading aside. As someone that can win about 1 in 5 games on Emperor, I get why people can kind of manipulate the AI, but I am also a frequent failure at warmongering. (I frequently declare war and find the war ending with my capital being taken, or win the war after a 90 turn struggle to the bitter end.) Point being, I don't consider myself a casual player, though I am a novice. I have about 300 hours in civ 6, and am not anywhere near deity level play. I think 1UPT is more "realistic" if we're going to throw that word around, and I would also like to offer a reminder, similar to many in this thread, that very few of those who complain about the AI have under 250 hours in the game (though that does get into the nature of a message board. Nobody posts on the civ fanatic message boards with 4 hours in the game, after all).

Sorry for the long-winded response. Long story short, it seems to me that the average (not casual, just not a longtime diehard) player is at a pretty good place playing civ 6. Just my completely biased and unprofessional opinion. Occam's razor says I just stink at civ, so what do I know?

 

[acluewithout]

My thoughts are in my signature under “Difficulty”.

In short, the game is too easy and can be quite passive in mid to late game. The AI is part of the problem - it’s cheese if you know how to play well (or well enough) - although the AI has got a lot better and can be really fight if you get your timing wrong (agree with @Victoria above). But the AI is only one aspect of what makes the game challenging. Fixing the overall difficulty would require improving other aspects of the game too, not just AI.

Key areas for improvement beyond just better AI are filling unit gaps and balancing units better (that would really help the AI, because I think it’s often building the right units in principle, but then getting screwed because the balance of certain units is off - eg AI building siege to take cities (good idea in principle) but siege are underpowered and lack a needed medieval upgrade); making it harder to get science (campus spam is too easy); just making empire management harder overall (so much gold and amenities in the game, so little maintenance); and perhaps tweaking what the various difficulty levels do (eg perhaps higher difficulties should increase required era score or increase maintenance).

Hopefully these things will get addressed in an expansion. The AI will also probably keep getting better in patches and indirectly from tweaks to game rules (eg unit balance), but the big jump in AI will have to wait for the .dll and modders.

In the meantime, you can make games more challenging and more interesting by avoiding snowballing (limit the number of campuses, limit number of cities), actively playing with certain mechanics more (eg diplomacy works much better than people give it credit for, it’s just that many people ignore it and so lose that whole dimension from the game; likewise, Religion is fun to play with either founding one or just proactively cultivating one from another Civ), using mods (8 ages of pace, mods that add power requirements or tweak loyalty) and setting your own goals and achievements and / or trying to finish games faster.

 

[mitsho]

I never argued by the way for a Civ without a military side. I just want fewer units on the map to move around with, thus making the moving less complex since it seems moving around takes a lot of time in the late game. These fewer units can be meatier, more complex and more diverse, but they are fewer and due to this ease the calculating needs of the AI and preserve my needs when trying an intercontinental invasion.

 

[Gedemon]

I never argued by the way for a Civ without a military side. I just want fewer units on the map to move around with, thus making the moving less complex since it seems moving around takes a lot of time in the late game. These fewer units can be meatier, more complex and more diverse, but they are fewer and due to this ease the calculating needs of the AI and preserve my needs when trying an intercontinental invasion.

Less units on the map make them more important, including how they're used. Not good for the AI, unless you give it a big combat bonus, killing immersion faster than your units.

Yes, low processing requirement would mean faster turns, but not better AI, unless you code it to be better, and we're back to the development cost/benefit ratio.

But I can see a 1UPT model with supply (ie "healing") and upgrades linked to cities production, allowing the AI to benefits from production bonuses, at the condition that whatever limitation of numbers you use, it could still field enough units to cover maybe not the map but at least multiple fronts.

Linked to that, and also about movement, you could have a model where you deploy/spawn troops directly around a position (city, encampment, fort), speed/numbers depending of your infrastructure (sea, river, land routes), techs, buildings and distance from capital, with the combat units not being physically on the map most of the time.

 

[Orup The Great]

and also about movement, you could have a model where you deploy/spawn troops directly around a position (city, encampment, fort), speed/numbers depending of your infrastructure (sea, river, land routes), techs, buildings and distance from capital, with the combat units not being physically on the map most of the time.

That’s an interesting suggestion. Could you illustrate your vision a little more? How would what you’re talking about look, in a gameplay situation?

 

[Orup The Great]

Like are you suggesting essentially a separate place for combat to take place?

 

[mitsho]

Less units on the map make them more important, including how they're used. Not good for the AI, unless you give it a big combat bonus, killing immersion faster than your units.

Yes, low processing requirement would mean faster turns, but not better AI, unless you code it to be better, and we're back to the development cost/benefit ratio.

But I can see a 1UPT model with supply (ie "healing") and upgrades linked to cities production, allowing the AI to benefits from production bonuses, at the condition that whatever limitation of numbers you use, it could still field enough units to cover maybe not the map but at least multiple fronts.

Linked to that, and also about movement, you could have a model where you deploy/spawn troops directly around a position (city, encampment, fort), speed/numbers depending of your infrastructure (sea, river, land routes), techs, buildings and distance from capital, with the combat units not being physically on the map most of the time.

Sure, whatever works. Is it really easier to code a realistic (and satisfying) "teleport troops from their barracks to the battlefield when necessary"-function though? I can see many player complaints as with every Automation function. That would have to work flawlessly to not bring about more complaints about killing immersion.

As what that discussion tells us for the current state of civ6 unit system? Ironically, I would make them cheaper to build for now.

 

[acluewithout]

Less units on the map make them more important, including how they're used. Not good for the AI, unless you give it a big combat bonus, killing immersion faster than your units.

Yes, low processing requirement would mean faster turns, but not better AI, unless you code it to be better, and we're back to the development cost/benefit ratio.

But I can see a 1UPT model with supply (ie "healing") and upgrades linked to cities production, allowing the AI to benefits from production bonuses, at the condition that whatever limitation of numbers you use, it could still field enough units to cover maybe not the map but at least multiple fronts.

Linked to that, and also about movement, you could have a model where you deploy/spawn troops directly around a position (city, encampment, fort), speed/numbers depending of your infrastructure (sea, river, land routes), techs, buildings and distance from capital, with the combat units not being physically on the map most of the time.

There is probably a limit to how much troop management Civ VI can bear. For me, I think supply lines would be too much. That sort of Mechanic would do better in something like EUIV (which has much more complex troop management).

I think a happy medium would be something like a Force Limit or Manpower Limit. You can only maintain x number of units (with some unit types costing or using more of you manpower limit, and maybe Levy units having no manpower cost), which if you exceed requires exponentially more maintenance and impacts loyalty and or happiness. You could then maybe have manpower costs increase when you can units on continents where you don’t have some sort of military base. The AI might struggle with that Mechanic, but you could maybe help it out at higher levels by giving it additional manpower or letting it ignore additional manpower costs on foreign continents. You could also tie the mechanics to Dedications, Alliances and or Policy Cards / Governments they let you increase manpower or ignore additional manpower costs in certain situations.

 

[Gedemon]

That’s an interesting suggestion. Could you illustrate your vision a little more? How would what you’re talking about look, in a gameplay situation?

Sure, whatever works. Is it really easier to code a realistic (and satisfying) "teleport troops from their barracks to the battlefield when necessary"-function though? I can see many player complaints as with every Automation function. That would have to work flawlessly to not bring about more complaints about killing immersion.

As what that discussion tells us for the current state of civ6 unit system? Ironically, I would make them cheaper to build for now.

Small "standing armies", few units, always on the map, and a "mobilization" mechanism, using stockpile of units elements (pool of personnel, based on empire size, pool of equipment, based on cities production)

Mobilized units are temporary, with a turn counter, you draft them in cities/encampment/fort using a UI similar to the production menu when you buy something with gold or faith, what's available in a position depend of the stockpile, distance from capital, route, techs, etc...

When the unit's turn counter reach 0, the unit disband in 2-3 turns, sending back gradually personnel/equipment to the stockpile.

You'll mobilize a lot of units shortly in the core of your empire, not so much on the border.

 

[Kruos]

Sure, whatever works. Is it really easier to code a realistic (and satisfying) "teleport troops from their barracks to the battlefield when necessary"-function though? I can see many player complaints as with every Automation function. That would have to work flawlessly to not bring about more complaints about killing immersion.

As what that discussion tells us for the current state of civ6 unit system? Ironically, I would make them cheaper to build for now.

Honestly, when you have archers able to launch arrows at more than hundreds km, or when you are not allowed to stack two little regiments in a tile modelising an area of hundreds km2, the immersion you know... I think it is the last thing cared about in the last 1upt Civs honestly.

 

[rbricker99]

No reason a powerful AI couldn't be developed from the data of multi-player games. It would require Firaxis making allowances for a custom AI to play.

It would require multi-player consent to record sessions. This would provide the data to train the basic AI.
It would also require the game to allow multiple AI's to play against each other at high speed.

After that, it would just take time.

Be careful what you wish for though. AI's are not intuitive, but they would have learned all the tricks employed by multi-player sessions and then by multi-AI sessions.

 

[Orup The Great]

No reason a powerful AI couldn't be developed from the data of multi-player games. It would require Firaxis making allowances for a custom AI to play.

It would require multi-player consent to record sessions. This would provide the data to train the basic AI.
It would also require the game to allow multiple AI's to play against each other at high speed.

After that, it would just take time.

Be careful what you wish for though. AI's are not intuitive, but they would have learned all the tricks employed by multi-player sessions and then by multi-AI sessions.

Now THAT is interesting!!!

What about it guys? Possible?

 

[Gedemon]

Now THAT is interesting!!!

What about it guys? Possible?

This ?
https://forums.civfanatics.com/threads/game-ai-net-based-machine-learning.627729/

 

[AsH2]

An hybrid of 1UPT/MUPT is what I'd like to see in civ7, but divided into strategical (ordinary) turns and tactical/operational (in-between) turns..
(I've mentioned that before and recently started a thread about 'Micromanaging - for fun, but optional' there I also mention this thread.)

Spoiler divide turns :

Instead of this mishmash of strategic and tactical decisions/moves in same turn, keep strategy matters (setup) in ordinary turns and (execute) then deal with any conflicting decisions/moves in Tactical/Operational in-between turns to streamline gameplay; (Then also time and map scaling could work well together during events.)

They (devs) could make it work by dividing the mishmash of strategical decisions and tactical/operational moves into separate strategical (ordinary) turns and tactical/operational (in-between) turns, where tiles (and movement) would be treated differently - effectively halved..

Strategical (ordinary) turns
- hexagon (pentagon) tiles allowing 3 (2) units per tile
Tactical/operational (in-between) turns
- hexagon (pentagon) tiles split into 6 (5) triangles allowing 1 (1) unit per node​

Edit: Pentagons to allow spherical map. https://forums.civfanatics.com/threads/in-civ-vii-spheres-or-cylinders.649608/

Spoiler earlier posts :

https://forums.civfanatics.com/threads/in-civ-vii-spheres-or-cylinders.649608/page-2#post-15544114
https://forums.civfanatics.com/threads/future-update.649590/page-139#post-15656683

 

[PieceOfMind]

I've taken a pretty long break from Civ games but have recently come back to playing a bit of Civ 4. I actually haven't bought Civ6 yet because I have played Civ5 quite extensively and was mostly unimpressed with it compared to Civ4, mainly down to 1UPT and its consequences for the game, and it seems clear Civ6 is more a continuation of the design philosophy of 5 rather than 1 to 4. (By the way, I still plan to buy Civ6 - I'm just not in a rush). The two main consequences are movement tedium and AI challenge. There are occasional moments in Civ5 that I might enjoy the tactical aspect of having to think about positioning, but they're outweighed by those other times where I just want a simpler way to to move an army from point A to point B. I don't want to have to think about how each unit's move will affect the next ones in terms of getting in the way. And if I find that annoying or a challenge to deal with I can only imagine how much an AI would struggle with it.

The move to 1UPT crippled the AI. It's simply because programming a tactical AI is much more difficult than programming a strategic one. It's like the difference between mastering Go and mastering Checkers.

What was really needed was adequate disincentive to stack everything on a single tile. I wonder if people remember back to Civ1 what its answer was? If a stack of units was attacked and the defender lost, the entire stack was destroyed! That was indeed a strong disincentive to stack everything on top of each other. The exception to this was for stacks in forts and cities. I think it's amazing to think Civ1 of all games had one of the easiest answers to the Stack of Doom issue.

Now, I'm not necessarily arguing we need to go back to that solution as it was arguably too harsh, but there has to be a middle ground somewhere. Allow units to be stacked so that the AI can be competent and the player is not bored to death trying to move more than 5 units across the map. But make sure there is adequate disincentive for players to put their entire force in a stack. Civ4 tried to do this with the mechanics of collateral damage. Amusingly though, when it became clear people were amassing huge amounts of siege in their stacks because it was the most effective way to wage war between stacks (collateral damage), the solution to that was to basically give the same mechanic to mounted units (flanking siege). Yo dawg, I heard ya like rock-paper-scissors so I put another rock-paper-scissors mechanic on top of your solution to rock-paper-scissors.

Basically in any battle in Civ4 there were two main advantages at play:
1) The attacker's initiative advantage - being able to choose whether or not to attack, and with what unit and from what direction. It was the the attacker's turn after all.
2) The defender's advantage of the best defending unit automatically being chosen within a stack. Seems only fair because it's not the defender's turn and Civ games have never had a separate move phase and battle phase.

Because of the rock-paper-scissors style of unit counters in Civ4, the defender advantage for a stack with at least minimal diversity was significant. A mini stack with a spearman (to counter mounted), an axeman (to counter melee) and perhaps an archer to take advantage of defensive positions like hills, forests or cities, was formidable. Even if an attacker had a 2-to-1 advantage of unit numbers he might still lose every battle because of the defender always being able to put up the best counter unit. I'm sure everyone remembers situations like this:

"Ok, I'll choose my City Raider swordsman to attack this city. Damn, they have an axeman. 10% chance of success. Ok, how about my horse archer. Their axeman doesn't counter that. Ffffff, they have a single spearman in there. 5% chance of success. Fine, my axeman with the melee-counter promotion (Shock) should go well. Oh damn, their archer has what feels like a dozen defensive bonuses - 2 or 3 City Garrison promotions, on a hill, archer bonus for being on hill, fully fortified, 40% from city culture, and 1-2 first strikes, 5% chance of success and they have half a dozen more because they're cheap. I guess I'll wait til I can bring half a dozen suicide cats..."

Maybe it needed archers to be able to do collateral damage, or the defensive bonus from forests (50%!) needed to be dialled back. Or maybe defensive bonuses from terrain only applied as the unit gained its fortifcation bonus over time (in Civ4 most non-mounted units gained 5% fortification bonus per turn it didn't move up to 25%).

The point is there are so many things that could have been adjusted to adequately disincentivise huge stacking but they didn't seem to be explored thoroughly. The weird "suicide-siege" mechanic was the only real attempt it seems and it mostly failed. To put it bluntly, if I had an army of 10 units advancing on a city, there didn't seem to be any reason to split it into two stacks of 5, and if anything there seemed to be only disadvantages. The defender had the mobility advantage (because attackers couldn't use roads except with a difficult-to-access promotion) and so usually could occupy the best defensive positions. Maybe Civ4 needed the flanking bonus of Civ5 - that is that when attacking a defender there was an attack bonus applied if you also had another unit next to the same defender. That would have provided some incentive for the attacker to split their stack, and the size of the bonus could be tuned as needed.