The core of 4X games "boring endgame problem" - a short essay about Rapidly Increasing Complexity

the game should have concription system that allows creating lower level military units from peasant by forcing them to join army. These should sometimes turn into counter and super aggressive barbarians that can pillage all tiles of your empire in one turn and also capture half of your cities instantly
You mean like Germany in civilization 5? You conquer an encampment and you take their units?
You mean a more challenging type of barbarian AI.
Or barbarians suddenly appearing in your civilization from all the sadness? That did happen in civilization 5.
 
You mean like Germany in civilization 5? You conquer an encampment and you take their units?
You mean a more challenging type of barbarian AI.
Or barbarians suddenly appearing in your civilization from all the sadness? That did happen in civilization 5.

No, I mean you use citizens to instantly create a military unit. For example you have 10 city and you force your citizen to join army and city loses one pop resulting pop 9 city. This will in turn create some minor temporary local unhappiness
 
No, I mean you use citizens to instantly create a military unit. For example you have 10 city and you force your citizen to join army and city loses one pop resulting pop 9 city. This will in turn create some minor temporary local unhappiness
Ohh.. that's like drafting from civilization 4. You need to adopt nationhood in civilization 4 to do that but there would be unhappiness in exchange but you instantly get a unit.
 
Ohh.. that's like drafting from civilization 4. You need to adopt nationhood in civilization 4 to do that but there would be unhappiness in exchange but you instantly get a unit.
Thanks. I had forgotten that particular English word though I think term conscription was use din civ 3? Never played 4

Edit: It seems term draft was used also in civ 3 but resulting unit was called conscript
 
Perhaps the game needs less "this will last the entire game" of a mentality. Like, playing civ without losing anything ever, probably shouldn't be so anywhere near as easy as it is. There should be more situations where we lose stuff but we choose to because of the greater good. The ability to play without losing anything ever can only turn the game into who has the biggest snowball.

Obviously an "aging up" mechanic could do this. Your cities might revolt or whatever BUT you gain access to better technologies so you at some point (Chosen by the player) will choose to age up.
Colonies should be immediately profitable, created with the understanding that they WILL revolt at some point.
 
I believe I wrote this already somewhere, but I like how in SPORE, whenever you evolve you play a different game. It does have the side effect that 5 little games are more underwhelming than 1 good, but makes proper difference between managing one animal hunting in the wild vs one tribe of ten such animals vs civilization of cities of buildings.

Likewise, maybe not in Civ but another game, actual change of game mechanics could make you feel that there is difference between managing tile improvements for each tile early and thinking of single city as as important as one such improvement in lategame. Turning Civ 6 into Civ 5 by Renaissance into "Europa Universalis" in Atomic etc.

Whole game you have mentality of what single unit yields. And how to improve it. It just shifts what that unit is. Starting as 1 Province of 4 Cities of 10 Tiles each, tile improvable with farm or mine being a unit, ending with 1 Empire of 10 Provinces of 10 Cities each, each City being the unit improvable with buildings now. Etc.
 
@Jeppetto
Yeah, while I wouldn't completely change the game, it makes sense to shift what you're focusing on as the game advances. Detailed control of worker assignments, district and and building placement and so on, is meaningful and enjoyable early on, but extremely tedious I'm the late game. I think it might make sense to have some detailed control being lost as a tradeoff for more advanced government types. Ruling as a god king who makes every little decision makes sense when you have one or two cities, but it would be a highly ineffective way to run a nation later on. As you phase out these parts of the game, you could instead introduce higher level mechanics, such as more advanced diplomacy, ideologies, international agreements, cooperative projects, and so on.

Once you've reached the modern era, you should be launching satellites rather than building granaries.
 
Perhaps the game needs less "this will last the entire game" of a mentality. Like, playing civ without losing anything ever, probably shouldn't be so anywhere near as easy as it is. There should be more situations where we lose stuff but we choose to because of the greater good. The ability to play without losing anything ever can only turn the game into who has the biggest snowball.
I agree. And as told earlier following the discussion on Sid Meier's conference, the fact people refuse any negative effect, even if known as certain, may be manageable through difficulty levels. We could imagine that on easier difficulties, the player can snowball pretty much everything without much problem, but the harder would be the difficulty, the more severe are downturns to be expected. This way, it may be better accepted by the player as a harder difficulty extra challenge whereas those who wouldn't enjoy it could stick to easier difficulties.

As a matter of fact, I personally never won Civ 4 on Deity because the challenge sounded ridiculous to me and I always better enjoyed playing as Monarch or Emperor. This has never bothered me at the time, I was quite happy this way.
 
Perhaps the game needs less "this will last the entire game" of a mentality. Like, playing civ without losing anything ever, probably shouldn't be so anywhere near as easy as it is. There should be more situations where we lose stuff but we choose to because of the greater good. The ability to play without losing anything ever can only turn the game into who has the biggest snowball.

Obviously an "aging up" mechanic could do this. Your cities might revolt or whatever BUT you gain access to better technologies so you at some point (Chosen by the player) will choose to age up.
Colonies should be immediately profitable, created with the understanding that they WILL revolt at some point.

It would be unusual for video game design, and maybe "psychologically risky" (like many revolutionary ideas Humankind had which turned out to be very counterintuitive, alienating and ultimately not fun) but I'd love to see a strategy game at least trying to work like this, where you know that at some stages it is preferable (or at least very serious strategic alternative) to lose important part of what you've built in order to transition to something bigger.

Imagine strategy game which is designed so there are vast swathes of land to colonize or conquer across earlier ages (medieval to industrial) and it is yet inevitable huge part of this is going to rebel sooner or later (we're depicting real life two waves of decolonization and rise of nationalism), and yet you still do that because it is either simply a very powerful temporary boost for certain empire building strategies, or because it is still somehow calculated towards victory that 'sure fellas your empire had to fall like all others but it did change the world to unparalelled degree, so you win history anyway'. The strategic dilemmas you have is 1) Whether you engage in colonialism or territorial expansionism or not (as not all 'builds' and 'meta' revolves around it and there are alternatives) and 2) At what point and how do you let them go (ofc it shouldn't be hard forced but rather increasingly pointless to prolong, as colonies are more and more rebellious and less profitable and diplomatically acepted, and it may be much more profitable to invent the controversial practice of neocolonialism :D )

We' re talking about colonial (or expansionist) empires abandoning their lands here, but you can imagine the same idea with political reforms (the player loses some total micromanagement control of the State here and there in exchange for very powerful indirect mechanics (as long as they actually work well)), or diplomatic concessions and compromise which are actually viable strategy (instead of typical civ "take everything you can from everybody, it's rarely worth to play nice"), or cultural and institutional changes which make you lose something big and gain something big...

I believe I wrote this already somewhere, but I like how in SPORE, whenever you evolve you play a different game. It does have the side effect that 5 little games are more underwhelming than 1 good, but makes proper difference between managing one animal hunting in the wild vs one tribe of ten such animals vs civilization of cities of buildings.

What you wrote actually surprised me, because I was so disappointed in Spore as a 14 - year old back then that I held that grudge all this time forgetting this game actually implemented some great, unorthodox concepts you can be inspired by... Just to disappointing extent.
 
Last edited:
What you wrote actually surprised me, because I was so disappointed in Spore as a 14 - year old back then that I held that grudge all this time forgetting this game actually implemented some great, unorthodox concepts you can be inspired by... Just to disappointing extent.

Yeah, Spore is the game where Alpha footage seemed seomwhat more creative than final product which is weird. It disappointed me in many aspects too. But I still find a lot of nice things there too. I do mentoin that making a game consisting of 5 different games has exactly that downside that each phase cannot reach its full potential, which IMHO was the primary issue Spore faced. So a little less radical, more tame stance to the same idea would be prefered.
 
Perhaps the game needs less "this will last the entire game" of a mentality. Like, playing civ without losing anything ever, probably shouldn't be so anywhere near as easy as it is. There should be more situations where we lose stuff but we choose to because of the greater good. The ability to play without losing anything ever can only turn the game into who has the biggest snowball.

Obviously an "aging up" mechanic could do this. Your cities might revolt or whatever BUT you gain access to better technologies so you at some point (Chosen by the player) will choose to age up.
Colonies should be immediately profitable, created with the understanding that they WILL revolt at some point.
I agree. And as told earlier following the discussion on Sid Meier's conference, the fact people refuse any negative effect, even if known as certain, may be manageable through difficulty levels. We could imagine that on easier difficulties, the player can snowball pretty much everything without much problem, but the harder would be the difficulty, the more severe are downturns to be expected. This way, it may be better accepted by the player as a harder difficulty extra challenge whereas those who wouldn't enjoy it could stick to easier difficulties.

As a matter of fact, I personally never won Civ 4 on Deity because the challenge sounded ridiculous to me and I always better enjoyed playing as Monarch or Emperor. This has never bothered me at the time, I was quite happy this way.
Another way to mitigate this type of snowballing, would be to simply get rid of all the permanent effects. As it stands in all current titles, the buildings the player builds work by providing the empire with a certain amount of specific yield each turn. How about instead, the buildings only provide immediate boosts? One example would be to have monument, instead of simply giving +1:culture:/turn, it'd instead just give a flat sum of culture all at once, and nothing else. Need more culture? Build another monument! And so on and so forth...

Though admittedly, this is just an idea, and I have no idea how it'd work practically. Heck, this approach could arguably be considered more appropriate for an RTS context, for all I know. But at the very least, I do wonder if this could possibly work
 
I love the early game but late game does have two big issues. 1) I already snowballed into victory and I have to spend multiple hours of mindlessly clicking meaningless things in order to get the official Victory. 2) My empire is so large that I am forced to spend my time micromanaging inconsequential small details every turn in a large empire.

I love some of the ideas in here. I'll throw in my own which I think may make late game easier to manage as well as throw in some opportunity for internal strife mechanics and politics.

There are now two types of governors, Official Governors and Local Governors. Official Governors are the governors we have now in Civ6 with their own promotion tree that you can unlock through policies and assign to cities. Local Governors are for your cities that you did not bother assigning an Official Governor too.

There is also the Civ Leaders (Teddy Roosevolt, Cleopatra etc.)

If a city is controlled by an Official Governor it acts the way cities in Civ 6 behave, meaning you control what the city builds.

The Capital City would be governed by the Civ leader and they could each have their own unique promotion tree too.

If you didnt assign an Official Governor, a city gets a Local Governor and behaves more like puppet cities from Civ 5 with stunted growth. Each local governor has its own personality and flavors. Some may favor building military units, some entertainment, some may be corrupt and funnel GPT from the city. There could also be new spy missions for example you could attempt to become elected Local Governor of an enemy city. Meaning that you would now control what the city produces and also open up new spy missions, like attempting to cause a total city rebellion turning it into a free city. There should be some way to oust your local governors if you do not like them or reelect new ones, including extreme actions like military action against your own city if you really think they are sabotaging you.


I would also say that micromanaging and moving builders around in the late game is also an annoying problem. I often dont develop or harvest land in the late game because I have conquered so much territory and controlling builders becomes annoying. I think it would be nice and someone already suggested to be able to just place icons of what you want improved and when you get a builder it automatically goes and improves the nearest icon. I would also suggest turning icons into a lens that you can toggle on and off.
 
Another way to mitigate this type of snowballing, would be to simply get rid of all the permanent effects. As it stands in all current titles, the buildings the player builds work by providing the empire with a certain amount of specific yield each turn. How about instead, the buildings only provide immediate boosts? One example would be to have monument, instead of simply giving +1:culture:/turn, it'd instead just give a flat sum of culture all at once, and nothing else. Need more culture? Build another monument! And so on and so forth...

Though admittedly, this is just an idea, and I have no idea how it'd work practically. Heck, this approach could arguably be considered more appropriate for an RTS context, for all I know. But at the very least, I do wonder if this could possibly work

I mean I thought about it, what would happen if buildings were more like projects basically. I think you would need some mechanic where each building costs progressively more production as you build them. Kinda makes sense, like newer monuments need to be "more monumental" to attract attention. Otherwise we are back at most biggerest civ wins territory.
Try find a way to make buildings more about timing than just wanting all of them.
 
I think the late game slog, (as early as after turn 120 in civ6) is mainly a UI problem. There's no auto production, no system to tell where built troops should go, there's also no way to select multiple troops and send them somewhere all at once. I also think there's just too many policy cards that you have to constantly switch out every 3-5 turns. The civics system from civ4 was far superior.
 
maybe make some changes like this

1.the power of new tech always throw away the old tech
although Player1 have a leading position and have a snowball in the game
we can push back Player1 by active the new snowball(tech) first, because the new snowball will be much more powerful than the Player1's old snowball

2.applying the benefit of a new tech must renew(by workers or by money) the old improvement and buildings

3.everyone can take loans to buy an army or buildings or something like that
 
I think the late game slog, (as early as after turn 120 in civ6) is mainly a UI problem. There's no auto production, no system to tell where built troops should go, there's also no way to select multiple troops and send them somewhere all at once. I also think there's just too many policy cards that you have to constantly switch out every 3-5 turns. The civics system from civ4 was far superior.

It's also the fact that decisions matter less. Like in the early game, if you have to move Pingala, that's a big decision. If you have 2 titles and are getting 6-7 science or culture from them alone when your entire empire maybe is only producing 20 science or culture, you are giving up a serious chunk to make it work.

But in the late game, even if Pingala is giving you 20-30 science and culture from promotions, if your empire is at 300 science per turn, then that's not nearly as important. So to me, I find it a slog because every now and then, you are forced to perform actions which simply don't matter as much. Policy cards as you mention end up in a similar boat. Early game I will make sure to swap in God King for 5 turns if it makes sense, but later game, you either need to keep micromanaging, or you just pick the ones you want and only change them when you change governments or one of the old ones becomes obsolete.

And yeah, UI issues obviously don't help. If a tornado takes out your campus, you have to repair the district and buildings one per turn without the ability to queue them up. If you get a bad disaster, I will have cities that have to spend 15 turns repairing their infrastructure, but cannot simply queue them or even just run like a "repair everything" project that just allows me to come back and check out the city when it's back and ready to work.
 
We all know the eternal endgame curse of 4X games (and some other strategy games but not not all of them). When a typical 4X game begins, it is so interesting; but we abandon so many games we begin, because they often (not always) become significantly less exciting the longer game session goes on. What's the problem? War system, victory conditions, diplomacy, yield balance, pacing, eras, tech tree? [well it's visible part of the problem is "I know I have won already, game is set halfway through, but I have devoted separate post below just to that aspect of the same problem]

The problem is deep: growth of complexity across every 4X game session is exponential and those games design fails to keep up with that inherent growth in terms of mechanics and interface, leaving the player more and more overwhelmed and burned out as the game goes on. Player has to use interface and mechanics built around early game very low complexity, to deal with the scope orders of magnitude greater by the endgame. And every game session, in every 4X game, trying to somewhat model human history over time, it simply has to contain rapid, exponential growth of complexity. In the early game you deal with two tile improvements and two barbarian units in one city, over the course of several turns. By the late game you can juggle dozens of cities and units with hundreds of tile improvements, and the more game goes on, the more happens in shorter amounts of time. This is unavoidable.

Not only the more game goes on the more and more happens in shorter and shorter amounts of time, here systems sciencr gives us the idea of emergence - at some threshold critical points, rising complexity creates entirely new qualities which cannot be reduced to their precedessors. In the early game you think of individual tiles and your little city neighborhood. Over the course of game you fluently start thinking in terms of local diplomacy, larger diplomacy, finally "globalisation" and victory conditions. In 4X games thus in every game session you dont just do more and more per turn; your ingsme activities and their menral framework and scope and scale of complexity fundamentally and dramatically changes as the game goes on.

This is not typical for most other video games. In most of racing, sport, action, rpg games you progress, unlock new toys and obstacles, but you think fundamentally about the same things from the beginning to end of your campaign. In Doom Eternal you begin with few toys and enemy types and end with dozens of toys and enemy types but you essentially have the same mindset as you begin and end the campaign: kill monsters by shooting gun at them. Even in complex tactical rpg games, even though you unlock tons of more items and skills and enemies with time, you do fundamentally the same things with your 4 - 6 team members: use skills and items to kill enemies. Relatively few game genres have this rapidly expanding and deepening cognitive load which 4X games have.

This is the problem, because interface and mechanics of 4X games being built and tested and designed around early game, collapse under weight of huge end game numbers. 1UPT district system feels great in civ6 early game, and it is a tedious terror of the late game - by the very nature of things, 100th tile improvement (+1% imperial yield) matters orders of magnitude less than 2nd tile improvement (+100% imperial yield), but you still have to spend the same amount of time to manage it. It becomes more and ml ore meaningless and therefore more and more boring!

It doesnt matter what shallow another mechanics, or number balance, or fancy fireworks you throw at that problem; you cannot solve the above issue of rapidly increasing cognitive load and rapidly decreasing meaning of individual player's actions by throwing even mire garbage on top of that, or tweaking some numbers.
The only solution I can imagine which could work would be to rise to the level of challenge and make game's mechanics and interface evolve as the game goes on. So for example player moves from having to manage every damn mine in an empire of 100 mines (tedious, boring, meaningless) to now having to manage imperial mining policy on a strategic level, or mineral trade in the global context. Less tedious, more impactful.

"Do you have any examples of 4X games dealing well with this problem?"
I think civ5 had one mechanic which was good in this regard, and the fact it was removed is miserable. Ideology system emerging in the industrial era. It was building upon your previous achievements and imperial backbone (so it doesnt "come out of thin air", it emerges from previous complexity), didn't require tedious micro on previous levels of interface, and introduced an entire new layer of complex interactions between players just in time to shakeup lategame.

However, significant offensive against this peoblem woukd need more than one such mechanic dealing well with the endgame/rising complexity problem, the core game design should recognize this issue and attempt to figure out how to keep "not feeling overwhelmed and bored" ratio to "making meaningful decisions" less asymmetric over a game session. Over the entire game design and pacing and balance and all mechanics.

Time will show at what point will 4X realize the nature of the problem (if I am right ofc; I was inspired by my recent months intensely enlightened by systems science in my academic fields)

EDIT
_hero_ and Kan Boztepe have rightfully pointed at the huge role played by "I know I have won already" aspect of the endgame problem. I devote a separate post to it, few posts below, but I still think it ties to the general theory of everything; "there are clear champions and losers 2/3 through the game" is because of exponentially rising power of early dominators, and to change that 4X games need some bottlenecks which shakeup things and make new rules of the game emerge by the late game, allowing underdogs to potentially win. There is actually such historical thing which can be adapted into great mechanic like that in civ series, Industrial Revolution; I have written more about this idea in that post below. So I think this caveat is still within my theory of What's Wrong.
I think solution to end game for any 4X is progressive difficulty and close to victory joint wars.

For example, progressive difficulty might mean +20% to AI bonuses for each era, but in non linear fashion.
close to victory joint wars means the closer one is to victory the more the AI's will join and wage war against you, not just DOW but actually capturing cities.

Civ series does neither of these things.
 
Last edited:
It is simple : replace improvements by policy cards late. If you choose "production" policy card, suddenly all your hills are mined. If you choose "agriculture", suddenly all your non-bonus or -luxury flat tiles are improved with farms.

Also, ability to attach several cities in one production queue : cities not connected would cost a lot of [insert new currency] points, while the connected ones, by rivers, canals, roads, railroads, ports, or airports would be cheap.

As to armies... just scrap individual units and provide a strategical layer as to where to deploy what we produce into an abstracted covering of our land, and from that base you can create aggressive actions in a strategic panel with arrows and success probabilities ?

Well, that would be kind of a different game I guess. But it might be interesting to have game A early and game B late. (although I think the switch should be player-driven, with a currency like Influence we see in Humankind, to avoid too arbitrary "oh, when the heck this option will come" problems)
 
It is simple : replace improvements by policy cards late. If you choose "production" policy card, suddenly all your hills are mined. If you choose "agriculture", suddenly all your non-bonus or -luxury flat tiles are improved with farms.

Also, ability to attach several cities in one production queue : cities not connected would cost a lot of [insert new currency] points, while the connected ones, by rivers, canals, roads, railroads, ports, or airports would be cheap.

As to armies... just scrap individual units and provide a strategical layer as to where to deploy what we produce into an abstracted covering of our land, and from that base you can create aggressive actions in a strategic panel with arrows and success probabilities ?

Well, that would be kind of a different game I guess. But it might be interesting to have game A early and game B late. (although I think the switch should be player-driven, with a currency like Influence we see in Humankind, to avoid too arbitrary "oh, when the heck this option will come" problems)

I like the simple practicality of those proposals, even if they'd be potentially more or less quite controversial in the community (especially military one, although personally I'd be on board with some radical solution like that). More abstract "game pacing, dynamics, AI, eras, design philosophy" stuff is one way to think about The Problem, but UI - based practicalities are also alternate way (or necessary part of) trying to figure it out.

I wanted to say I am super happy to see how this thread turned out, with a lot of amazing insights, discussions and ideas of you all, and being surprisingly alive after all these months, when I began it I was pessimistically afraid it will quickly fizzle out. In the best case, the splendor and long term presence of this thread may have attracted attention of devs and inspire them to Do Something (some time ago one of them was talking on this subforum in another thread, so they are here). In the middle case, it may inspire some mods for civ7, and in the worst case, well, maybe the real treasure was the discussion we made along the way :D
 
One common failure of "Automation" systems to reduce micromanagement is that they are often worse than what a skilled player can achieve with clever orders. Bad automation is almost as bad as no automation, since it simply won't get used.

So you need to make sure that any Automated system is actually delivering results with some sort of production bonus. But then the question comes: why not just automate everything from the start? Why even play the game?

My idea here is to let the automation features be unlocked by technology or some other achievement. Early in the game, and when the empire is small, you control everything, but as you gain more cities and as communications technology improves, you gain the ability to put more and more of your empire under more efficient decentralized control. Found a city, declare it to be a "Special Economic Zone", and watch as it builds up mines and factories on its own, cheaper and quicker than you could have. But such abilities should be limited, so that, roughly speaking, you're still in direct control of your 5 or so most important cities.

The important idea here is not the specific nature of how the automation would work, but rather that it should be an ability that is unlocked on the same time scale that empire management becomes a headache.
 
Top Bottom