A way to end the snowball effect of huge empires

Loppan Torkel said:
Not that corruption will solve everything, but it's an easy way to hinder expansion. If you raise the military upkeep at the same time you will soon reach a point where you have to leave your outer cities unprotected to not go bankrupt. It shouldn't be done like this though.
That's a pretty extreme predicament. ;) As long as cities provide free unit support then you will rarely, if ever, see this situation.

I agree that random events shouldn't steer the game too much, but on the other hand they are necessary to some degree. It's about calculating risks and there will always be complaints about them, like when spearmen beat tanks etc.

Aussie_Lurker said:
Yep, I have to agree with Lopan on this one, Trip! Although I believe that the random element should be kept to a minimum, there should still be a random element in determining if civil wars actually occur! Randomness is a part of many other elements of civ (like combat, for instance) and is also a big part of most wargames (which you claim to be such a huge fan of) so why should this be any different. You will see, though, that my model seeks to keep the worse elements of this randomness at bay, which is why I speak of getting a pretty decent warning first!
It depends on the effects of that randomness. Do you think my rebellion model would properly cover civil war situation? I think its the best way to impliment it simply because A) it's realistic, and B) it's not a HUGE impact. If you have some cities flip all of the sudden it isn't good for gameplay. Just as a player can choose a Scientific civ and aim for techs he doesn't think other players will choose, if he DOES get that SGL and builds the Pyramids in 2000 BC the whole game is screwed.

MAJOR random events, good or bad, are not good for gameplay. I have no problem with smaller events, but something to the scale of many cities flipping and dark ages aren't acceptable...
 
I feel the main issue we are having trouble with, here, is random versus semi-random events! As I said on page 3 of this thread, history sometimes turns on a dime and I, for one, one to play a game where I can recreate this element of history if I want to! At the end of the day, though, I also want the power to avert disaster-and my model does allow room for both options! Also, we are not talking 'culture flips' a la civ3! What I am talking about is when things get so bad that your domestic advisor is coming to you to say: 'Sire, I fear that some of your citizens in the city of '___________' are becoming restless and productivity is way down-perhaps you should try.....(insert helpful suggestion here)' . Then, later, he comes back saying 'Sire, the people of '__________' are becoming riotous, and are commiting wanton acts of vandalism-you MUST do something to quell their anger'. Later still, you get ANOTHER warning of 'Sire, the people of '___________' have become rebellious, some of whom have broken away from the city and headed for the hills! You must try and put down this rebellion'. At any time between these warnings, though, you could get a message from your advisor saying 'Sire, recent acts of sedition have caused many of the citizens of '________' to whisper against your rule, and I fear they might try and seccede. Their main complaint seems to be......(insert trigger event here)' If the city (or cities) seccede, they will form a NEW NATION, one which you have the power to reincorporate into your nation at any later date (by force or diplomacy!)

Do you see how I want it to work now, Trip. There is SOME small degree of randomness there, and your rebellion model is definitely a part of my overall idea but, at the end of the day it can be averted by player actions-assuming they don't take too long to do something about it! The same kind of system, though, also applies to my plague and Dark Ages model!

Yours,
Aussie_Lurker.
 
Actually, I feel that I have been distracted from the main issue here which is: How do we 'decouple' past success from future success-if that makes sense. My point is that, historically, those nations which became large and 'successful' did not continue to get bigger and more successful ad-infinitum. In fact, the largest empires in history, be it the Assyrians, the Egyptians, the Romans, the Spanish, the Dutch or the British Empires are either no longer around, or are now just fairly minor players in the grand scheme of things, and there MUST be a way to properly model that within the game WITHOUT removing the fun element of the game.
Oh, one last point though, Trip, is that in my model you should also have the power to peacefully allow cities to secede. Doing so would improve your future relations with whatever nation forms from that seccession!

Yours,
Aussie_Lurker.
 
Aussie Lurker, you might never find a fun way to let the players civ wither and die. I think most people would rather detest something like that eer happening. Success should really be determined by the players policies, and the AIs policies. If the AI outwiles you, then you can lose.
 
But MC, that IS what I am talking about. Having players set up to 'wither and die', as you say, would be nothing but predestination (not very fun at all). I agree that ultimately player choices MUST be the determinant of success. The problem I have, though, is that EARLY success feeds in on itself, creating the 'snowball effect' we have been debating here (and which is mostly based on sheer size of the civ)-that is to say that the player who gets the early lead is all but predestined to remain in the lead for the remainder of the game-this is not only ahistorical but also not much fun. Instead I want a system where the following states exist:

1) Size is no LONGER the sole determinant of success, but where good management of what you have, and strategic acquisition of what you need, is far more important.

2) Expansion in the very beginning of the game is somewhat curtailed, and where expansion occurs more in fits and starts, as it was in history-rather than in an exponential fashion, as it is in Civ!

3) A player who fails to get in the lead in the early game has ample opportunity, via good management and strategy, to get in the lead at various points in history.

4) A player who gets off to an early lead may, through bad managment or poor strategy, end up falling back towards the middle or even back of the pack!

5) Success can as easily be achieved through building and co-operation as it can through conquest and war.

No matter HOW they are achieved, if civ4 can meet all-or even the majority-of these desires, then I will be a very happy man indeed :)!

Yours,
Aussie_Lurker.
 
Aah, get you now ;) Yeah, that would be good, and fun. Semi-random events are fun; events based solely on chance are not. I would love to see minor civs implemented and they would certainly curtail many of the aforementioned problems, but exactly how do you reckon they should be implemented?
 
Those ideas are nice, but how to actually implement them? Well, i am sure the civ4 development team can come up with a good idea.
If they can do this right, i will keep the game challenging, (not discouraging) for longer period of time, and hence more fun.
 
I definitely agree with Aussie_Lurker and his list of points, but I am at a loss to explain how that could be accomplished without a major overhaul of Civ.

One fundamental aspect of Civ that might be preventing this is the relative simplicity of Civ. Although a simpler game is easier to learn and start enjoying, a simpler game is also easier to master and perhaps become boring. Unfortunately, this is the case for the snowballing effect observed in Civ--it is too easy to figure out all of the AI's tricks, too easy (and tedious) to figure out how to extract every resource from our own civs, and too easy to make plans that few forces or events in the game can derail. There is no effective way to combat the snowball effect unless the game structure is heavily altered or the AI is imbued with human intelligence. I assume that the latter would be difficult with our present technology, and therefore I advocate some sweeping changes to the game.

My specific ideas are contained in the comprehensive UET II thread, but I think it more important to establish my more general take on this matter. I would like to see Civ be more complex, such as in introducing more forces and variables to consider for every course of action, and thus grant the AI a subtle advantage--the ability to compute and factor in more variables than a human might bother with. Since the AI is mechanical in its "thinking," its inherent advantage would be to have more calculations that it might do more precisely than a human can. For example, a human could easily estimate, using experience, logic, and even their "gut feelings," pretty accurately how long it might take to construct an army or develop a region. However, if there are more variables involved in this process, the human estimate might become less certain, while the AI can continue to crank out the calculations. In this way, the AI can have an advantage without really improving its "thinking" capabilities very much.

In any case, having more complexity would reduce the player's ability to figure out all the tricks and build foolproof plans, which is critical to reducing the snowball effect. In addition, the AI would get a boost as its prescience becomes increasingly more accurate than the human player's, giving the AIs more leeway to continue doing unstrategic things that are difficult to correct in AIs. So, yes, I do advocate a major revision of Civ to solve the snowballing problem, and we already have three versions of Civ to see that past revisions of Civ have not sufficiently curtailed snowballing.
 
I don't think that making Civ more complex and random is a good idea to combat this though. There is already too much micromanagement involved, and people won't play a game with hundreds of variables and rules and so forth (and at the end of the day, Firaxis needs to produce an appealing game to make some money). A significant review of various things are needed desperately.
 
I do agree with T-P that it should be incredibly difficult to account for ALL factors when making decisions about the future of your empire-thus making it harder to 'outwit' the AI. At the same time, I feel that the people of YOUR civ should act more independantly of you, the ruler. If you do something which is against THEIR percieved interests, then they should react against it. This might include which improvements/wonders you build, what governments you adopt and which nations you form diplomatic relations with! Angering your people in this way could also tie in directly to the chance of revolts and civil wars.
Speaking of which, I think that a number of factors should play into the chance of revolts and/or civil wars in a city or region. Though it should be VERY hard to control ALL of these factors, it should not be impossible, and warnings from the Domestic Advisor should help greatly!
Random events should be triggered by 'triggering events'. In the case of civil wars, this could be government change, revolts, loss of your capital or changes to your civ characteristics. In the case of a dark age, loss of a certain percentage of your current culture, plague outbreak, loss of capital, loss of a percentage of your cities and/or population in ONE turn. Whether or not the 'triggered event' actually occurs will depend on whether the various prerequisites for the event both exist within each city, and if these PR's combined match or exceed a particular threshold. How much they exceed the threshold by determines how EXTREME the even is. For CW's, the PR's might include unhappiness/corruption/war weariness levels, distance from the central or provincial capital, # of units etc etc. For dark ages, the factors might be current science funding, # of cities/citizens lost that turn, current level of religious influence, current age you're in (and where in that age you are) etc etc.
Anyway, I think this and a more 'independant' citizenry will make for a more intriguing game, AND along with the other things I have discussed, will assist in limiting the snowball effect.

Yours,
Aussie_Lurker.
 
Hmmmm,

In answering the 'Prehistory' thread, I came up with a new idea related to dark ages, as WELL as a good possible means of dealing with the snowball effect!
OK, you have the different ages in the game but, instead of simply getting prerequisite techs in order to move from age to age, you also need to achieve a certain degree of sociological and economical development as well! However, should you somehow LOSE one of those two development criteria, then you end up backsliding into the previous age-losing any techs from that new age in the process. Given that infrastructure would be a major key to the sociological and economic prerequisites required for advancement, then this same infrastructure could prove to be the achilles heel of an especially large empire. Small nations will be vulnerable too, but their smaller size will make it easier to both defend this infrastructure from attack AND pay the neccessary upkeep to keep it all in order. Also, backsliding would be a KEY TRIGGER for a dark age-where your cities lose productivity, science, culture and wealth.

Yours,
Aussie_Lurker.

EDIT: OK, I confess that this is NOT a new idea, but one which first came up when the 'Dark Age' thread first appeared.
 
Just to reinforce some of the stuff that Aussie is saying, just because you say "semi-random" civil wars, it doesn't mean we're advocating random break-offs that happen no matter what.

Under normal circumstances, there's a 0% chance of civil war.

After unhappiness becomes rampant, and the rules of the game determine that the player had his chance to avert civil war and play intelligently, even still, it wouldn't just say "okay civil war now". It would say "probability(civil war)=0.5". That's the essence of semi random, in my mind. Impossible if you play it safe, and simply a matter of inevitability once you pass that point of no return.

When would civil war happen? In my mind, one reason would be because you failed to really assimilate the people you've conquered, and they eventually came back and overthrew you. And it would be hard to assimilate them if you're at war. (1) Because violence would keep the people resentful -- not necessarily unhappy or resisting, but something less than content. (2) Culture production should have a hand in assimilating people, with the most cultured enemies being hardest to assimilate -- and culture production should be slowed while at war. (3) Because funding a war requires you to neglect reconstruction -- and if you fail to aid in reconstruction, it's only a matter of time before the disenfranchised rise up against you.

Assimilation is one of many factors, of course. But just throwing that single item out there as a talking point.
 
Well, one idea is that if you DO implement civil war, rather than the cities themselves becoming overtly hostile to you, they simply break away and form a separate province which, though technically part of your civilization, is not directly controllable by the player. You may model this as a certain group of people who, because of the frustrations and inefficiences of being in such a large empire, simply peaceably break away. Your own units can not attack and reclaim the cities, because your own people refuse to attack their fellow citizens. In fact, you can move in and garrison your own military units in these cities in time of war, to prevent the AI breaking in and razing them. However, you cannot open the city window and change anything. The cities would produce nothing, and commerce would not go to your civ. The cities would never become unhappy, and they would be garrisoned only by drafting via the AI, and only in time of war, in addition to any units you might park in them. Barracks and such would still have the same effect on your units, and you would still collect culture points from the cities, as it would be unfair to deny this if you were going for culture victory. You also receive benefits from all luxuries/resources in any breakaway territory, and can even perform worker actions as needed, in fact it would still be your turf in every respect except for the fact that you wouldn't control the cities per se.

The chance of a group of cities going self governing could be based partly on empire size, and partly on player action, i.e., unpopular war, or some atrocity, like ROP rape or nuking. Unhappiness should play a significant part, to the degree that the player better have access to plenty of luxuries if they are planning on starting a major war. Care would have to be taken to ensure that the cities that broke off were not the players early core cities (the heart of the empire), but also not a group of pathetic little tundra cities that the player doesn't care about either. Cities lost should be those that are conquered, have the chance to offer some meaningful culture and production, but depart from the empire due to player mismanagement. The number of cities that might break off at a time would have to be determined by a model based on other factors, some experimentation would be needed to determine how much slows the player down without crippling him/her.

Using this model, overall ability to wage war could be controlled to some degree, yet it is not completely random, the player can abuse, and possibly lose cities, but the RNG does not get the final say. This system would be guaranteed to produce a LOT more city razings during wars, with the resultant open territory that could be fought over and controlled, if not always rebuilt upon, which might be interesting, if you allow military forts and such in no mans land.
 
What I would like to see in a civil war is similiar to barbarian uprising. You have insurgents popping up, sometimes in huge quantities, they are aided sometimes by the citizens of rioting cities. The citizens can destroy buildings, preventing civil service, (no hospital service for your troops so they can't recover) or even attack troops garrisoned in the city. The troops might become injured (each timee hippoint go down by 1)or even chased away. (when hippoint is about to become zero). This way it is random, but the player has a full chance to quell the rebellion. And what I like about it is that the user can have more chance for fighting war. I like wars. yes, it keeps the player involved in the process. This is way better than the 'cultural flip' they have now , where your troops in a newly conquered city just magically disappear and the city just somehow flip sides.

The rioting city will breakaway after a certain time if
1) there is no unit garrisoned in the city
2) the unit is chased away by rioting citizens
3) an insurgent unit takes the city

On going riots and rebellion and your failure to quell them encourage further rebellion. If you ignore your people, you will be dead. Hostile foreign power might aid the insurgents as well.

There should be differnt kinds of rebellions and they have different goals.
1) rebellion by conquered ethnic minority - they will try to rejoin their mother nation (if that nation still exist) or they will try to revive the dead mother nation by independence from you.
2) rebellion by people in your cultural border - most likely a revolution. they want to overthrow you, and have a new government form. (such as communist overthrown of the nationalist government in china)
3) rebellion of the colonies - these people are semi culturally tided to you. They will most likely seek just independenc, and form a new civ. Won't want to overthrow you or join another country.
(American indep. from the British Empire)

You can have all three kinds of rebellions going on at the same time. Gives you lots of headaches, and lots of fighting to do on the battlefield. After kicking some ass on the international areana, I think the user would like a change, and fight a demostic war instead.

We shouldn't have rebellion after the capital is taken. It's already bad enough to have the capital taken, why rebellion? And I don't see any real connection between capital being taken and people rebelling, unless those insurgents are minorities who hates you. People of your own culture should bond together even stronger then before and try to resist foreign invasion. Washington DC was taken and burnt in the 2nd war of independence, and no state rebelled. The Chinese capital of Nanking was also taken and razed by the Japanese during WW2, and no Chinese rebelled.
 
Dida, I think you might have missed the point about my civil war post. Though losing your capital is a 'Trigger event' it doesn't mean that you WILL have a civil war (I really disliked that element of Civil wars in civ2 :(!) If you already have a lot of corrupt, unhappy cities though, then losing your capital COULD precipitate an attempt by said cities to break away-especially if they are on the periphary of your nation. As I see it, Civil War is more likely to effect those who engage in rampant and thoughtless expansion and/or poorly manage their cities!

Yours,
Aussie_Lurker.
 
I generally agree with the many ideas presented thus far concerning reasonable factors, NOT randomness, determining whether civil war occurs. I also agree with Aussie_Lurker's idea of having factors besides technology considered in whether a civ "progresses" to the next era.

Allof these ideas have an underlying theme--they increase the complexity of the game by adding a few more things to consider. This, in my opinion, is the ultimate way to reduce snowballing.

Note, however, that I am NOT asking for Civ to be more random (I don't believe I ever stated such in my previous post, but maybe it could somehow have seemed to imply...?). I would rather have a lot of variables and NO randomness than a few variables and a lot of randomness, as Civ is. The reason is that in the latter situation, the player is arbitrarily limited in his ability to control his game, due to the randomness. With many variables and no randomness, it eventually becomes possible for veteran players to have a better understanding of how everything comes together, and develop more advaned and/or specialized strategies. For newbies, the game will seem "random" whether there is really an RNG or just the sheer number of factors to be considered. When they become veterans, the RNG continues to serve to limit them, arbitrarily.

Even so, veteran players dealing with an un-random but complex game would not be able to initiate the snowball effect due, simply, to the number of factors that can force adjustments. This is much more effective than using the RNG, because the randomness gives success, as well as failure, almost equal shots, so that any of the few mechanisms to curb snowballing would not work as often as they should. On the other hand, with many variables, there are more chances to force plan adjustments, and should factors against snowballing be present in a situation, they WILL have an effect. The overall idea is to achieve a "random" effect without being random, and thus ensure that mechanisms that curtail snowballing WILL work.

By the way, I also hate micromanagement, but I do not think that increased complexity warrants increased micromanagement. For this to be possible, the player should act more as a leader does in real history--tell people what to do, and influence the environment in which people can do their work, but do not actually go do the work personally. For example, if I want a city to trade more with another city, I should order a road built between the cities, and thus encourage trade through providing easier transportation. I should NOT have to go make one of the cities build a caravan and then manually move it to the other city.

Hmm, I got a little off-topic there! Anyway, I agree with Mongoloid Cow that Civ needs a "significant review of various things" to eliminate the snowball effect.
 
Top Bottom