The Symbiotic Nation: a Civ 'goal'

[MrBaggins]

Consider the following proposition:

2 players

2 identical islands

30 cities each with 'averaged' terrain

no more viable room to expand

equal 'skill' and knowledge of game mechanics

equal techs

both players have evenly improved their cities, so that their output is increased by 1/4 in all aspects. Additional technical innovations later in technology increase this to 1/3, and later 1/2.

With the current improvments tiles are worth, say, 10 food units, 10 production units and 10 commercial units. Additional technical innovations later in technology can increase all by 2, and later 3.

player 1 has his cities size 10, having just reached that size at all his cities
player 2 has his cities size 11, having just reached that size at all his cities


player 2 has a production and science bonus, that equates to 12 units at each city (2 more than the extra tile alone)

Player 1 will 'catch' up to player 2, but only temporarily, and soon player 2 will be in the same situation having the same 12 unit advantage... again, and again.

A proportion of the total advantage will also be put into science. player 2 will get to advantageous technical innovations quicker than player 1... the most vital of which being growth and science improvements. They give player 2 more of an edge, first... and the science and growth bonus advantage will continue to lengthen, ad infinitum.

MrBaggins

[DrSpike]

Not really, techs can be stolen.......and there is a military in the game as well as science. Also some games penalise the leader in research.

Failing all that.......just don't fall behind in the first place.

[MrBaggins]

Spys (and invasion troops) can't get there... there is a naval meat grinder between the two islands. Which player 2 has an advantage in, due to his production and science advantage.

And your quote, "Failing all that.......just don't fall behind in the first place. " actually can be paraphrased, "Yes, Bigger is better IS a problem."

MrBaggins

[DrSpike]

So in your world of constrained growth for the leader there is no way to capitalise on a lead. What is the point of being the best civ then?

[MrBaggins]

Being bigger still should be better, just never as good as the last time... the advantage should regress, at each growth, so that accelerating away becomes more difficult. It flattens the game state, to bring opponents closer together.

It doesn't stop you finding edges... just that those edges in terms of resources, as less profound, which allows you to concentrate on the game more as an exercise in achieving different victory conditions, than getting to an advantageous resource situation... which is *THE* victory condition now.

MrBaggins

[DrSpike]

Originally posted by MrBaggins
Being bigger still should be better, just never as good as the last time... the advantage should regress, at each growth, so that accelerating away becomes more difficult. It flattens the game state, to bring opponents closer together.

Well if its any consolation the latest game (Civ3) tried in part to do this by penalising the research costs of the leader. I have previously stated in this forum that this works well for SP, so perhaps we are not so far apart in goals........though some of your arguments need to be rethought IMO.

[MrBaggins]

Originally posted by DrSpike


Well if its any consolation the latest game (Civ3) tried in part to do this by penalising the research costs of the leader. I have previously stated in this forum that this works well for SP, so perhaps we are not so far apart in goals........though some of your arguments need to be rethought IMO.

Research is one aspect of the game which is a 'positive-feedback-loop'. Research spread is a good way of dealing with this.

I've numerically proven that city size... given multiplied rewards is yet another 'positive-feedback-loop'.

If you disagree with the arguments, i've made, why don't you state which ones exactly, and why they are not correct (using numerical situations)?

MrBaggins

[DrSpike]

You can numerically prove whatever you want......try playing some civ.

Even in your example where optimal empire size is essentially fixed you still face trade-offs. You could get the maximum number quickly, at the expense of technological development, or you could develop a few quickly and forego some expansion to capture some of the rewards you allude to.

Perhaps the guy that expands can steal tech, or capture cities with a bigger production base, or perhaps the guy who techs will get better units and buildings and pull ahead.

All these things are possible..........everyone starts with equal resources, all civ is just trade-offs. If someone sets up a situation where they hold all the cards then the other players did not play well - end of story.

[MrBaggins]

I do play civ. That is irrevelant to the discussion.

To analyse the exact effect of a component of a complex system, you must isolate the component. The situation given is exact and equivalent in all regards except one... the city size... the component in question. Its actually irrevelant which path has been taken to this point, if the situations are exactly identical NOW.

All strategies in civ are predicated on the resources gleaned from cities. Units, science,improvements and gold.

The simple fact is that you have more resources... given the 'positive-feedback-loop' from your cities. The other 'paths' are merely additional layers of advantage or penalty. These layers are actually useful in terms of giving choices. True, they need to be balanced... but they should allow for differences, and in some cases, advantages. Its just the underlying resources AKA city production that should be compressed.

Have you read the Brian Reynolds article, btw?

MrBaggins

[DrSpike]

I read the article, and for SP many of the points are valid.......it is easy for the game to be over when the player gets an advantage. Hell, I have been civving since Civ1 so I know the problems. I have stated above I support the doctored research costs for SP.

I think some of your posts are a misinterpretation though. Let's stipulate 20 cities with maximum workers is your goal. You still have to choose between an expansion path getting 20 as quick as you can, then growing them, or trying to build up a few to lock in a research lead, and adding the rest later. If you can reach the goal quickest avoiding all the possible setbacks then you should win.

Now for SP sometimes things do need to be tweaked, but you cannot go too far else you risk all strategies just having the same result (regardless of their intrinsic merit) because of restrictions the game imposes to keep things equal.

[MrBaggins]

You fall into a trap if you try to analyse a complex situation in its entirety.

You are forced to use reductionist analysis:

* Isolate the layer

* Isolate the component.

* Test the component in an isolated case.

* Iterate with different layers and components.

The city 'case' is the most important because cities are the primal construction component of empires. A case can be made that techs are at the same level, but ultimately techs are researched from resources generated by cities. Units are a property of cities. The construction of anything else, is a property of cities.

Any strategy, including military, infrastructure or technological 'options' are predicated on cities... specifically that the more productive a city, the better qualitively or quantitively the strategy can be pursued.

These strategies can be fed back into the 'system'. I don't believe in less importance to these subsequent strategies. They form the basis for option that a fun game is predicated on (*assuming that all opponents are free to choose and employ similar options.)

A strategy can give an advantage without the 'city-production-positive-feedback-loop'... for instance... say you develop a military technology which allows you to develop an improved infantry unit. You would still have an advantage on the battlefield... Say you develop a growth improvement instead. If you develop a growth component instead, your cities *will* grow faster, and produce increasing amounts, just that the rewards will diminish. No strategy is crippled.

Being ahead, with less of a positive-feedback-loop, means a declining advantage, as you get further ahead= more challenge, yet one that you can still win... but won't necessarily, which is the whole point of continued play= funner game, even in MP.

MrBaggins

[DrSpike]

Originally posted by MrBaggins
You fall into a trap if you try to analyse a complex situation in its entirety.

You are forced to use reductionist analysis:

* Isolate the layer

* Isolate the component.

* Test the component in an isolated case.

* Iterate with different layers and components.

Or you could just try playing some civ and see.

Seriously though, the bigger is better problem as you outlined it in your first few posts is just not an issue. Of course holding everything else constant bigger is better.....what else would you expect?

The fact remains that whatever the game there are various ways to reach the optimal 'big' state, and if you can get there first you will probably win. So what?

If you aim to get there as quick as possible and I come along and take all your goodies because I prioritised unit production then tough cookies. If you can hold off my attack and lock in a scientific advantage tough cookies to me. That's the game.

[MrBaggins]

Originally posted by DrSpike


Or you could just try playing some civ and see.

Seriously though, the bigger is better problem as you outlined it in your first few posts is just not an issue. Of course holding everything else constant bigger is better.....what else would you expect?

The fact remains that whatever the game there are various ways to reach the optimal 'big' state, and if you can get there first you will probably win. So what?

If you aim to get there as quick as possible and I come along and take all your goodies because I prioritised unit production then tough cookies. If you can hold off my attack and lock in a scientific advantage tough cookies to me. That's the game.

Arbitrary experimentation through "play" is the least accurate and efficient way to discover how a change in a component of a complex system affects the whole. Isolate first, recurse... then test.

You are entitled to your opinion, that bigger=better is fine and well... Brian Reynolds disagrees, and I concur. QED... The mathematics of the isolated situation show it to be so. It happens equally in SP as well as MP. Brians comments are written not only for SP but for MP, and I agree with that also.

Insisting that something isn't so when faced with absolute evidence to the contrary strikes me as "flat-earthian" in the extreme.

Sure you can play the game 'as is' and sure it will produce a viable game each time. That doesn't mean that the game isn't unbalanced- in mp as well as sp getting a city advantage= a massive advantage each and every time. I've statistically showed that this is a 'positive-feedback-loop'. The designer of the game that you are playing thinks that those are counter productive in good game design. So, common sense says eliminate the obvious underlying 'positive-feedback-loop'.

MrBaggins

[DrSpike]

Yours and Reynold's points are very different. Reynolds is talking solely about technology costs, and I have already repeated stated I concur that something needs to be done to stop a small early tech lead snowballing.

I was restrained in comments before, but the rest of your stuff on positive feedback loops through getting a city advantage is rubbish, and I begin to doubt you have ever played much civ of any flavour. If I have more or bigger cities than you, with no other disadvantage, then I am better placed and have the best chance to win. What else would you suggest? As soon as one player get a city lead of 2 cities he is not allowed to build any more? If that sounds silly, there's a good reason, because its damned silly, along with your precious feedback loops.

You start with the same resources, and if I have more cities you must have bigger ones, and probably a slight tech lead, or you are not a good player.

Anyway this is boring me now........I think I might just play some civ.

[MrBaggins]

Originally posted by DrSpike
Yours and Reynold's points are very different. Reynolds is talking solely about technology costs, and I have already repeated stated I concur that something needs to be done to stop a small early tech lead snowballing.

Complete and utter rubbish...

Brians comments use an example of tech, but goes on to say that this applies to all general game mechanics

In The Poor Get Richer: The Ancient Art of Game Balance, Brian Reynolds said
The rules of the game should work to keep the game competitive for as long as possible. Players have the most fun when the game is a tight contest, so our game systems should help keep players who fall behind "in the race" and try to prevent players who get ahead from simply "running away with it."

Ideally, a game should end at the exact instant a player has effectively been guaranteed victory, a player should be eliminated at the moment victory for him becomes essentially impossible, and both of these situations should occur approximately when the players have experienced the expected game arc--that is, in a game about a 10,000 year timeframe we don't want to have 80% of the games end after the first 1,000 years. These ideals cannot be practically achieved in every game session, but our games and game mechanics should be tuned to push toward these goals rather than away from them.

Originally posted by DrSpike
I was restrained in comments before, but the rest of your stuff on positive feedback loops through getting a city advantage is rubbish
*SNIP*

So point out EXACTLY how it doesn't. Try using proofs... numbers are the only way to ultimately prove what comes down to a formulaic situation. You should be able to simply disprove this in the isolated example, if its true. Disprove the math that somehow a culmulative bonus is not a culmulative advantage.

Originally posted by DrSpike
*SNIP*(what) would you suggest? *SNIP*

Diminishing returns, not caps or 0 returns, except in extreme cases.

MrBaggins

[DrSpike]

For the last time its rubbish because you cannot possibly get an advantage in city count or size without your opponent getting another advantage. You cannot analyse all the possibilities numerically........civ is a fluid game with tech, growth, expansion, and production. YOU HAVE TO HAVE SOMETHING EVEN IF YOU GIVE UP THE OTHERS. Maybe I have more cities, and this could potentially snowball. But will it in reality? Maybe, but not if you have an army to take my cities away, or lesser well developed cities that give you a tech advantage that you can use against me.

You can go on about your stats all you like (and I do understand the need for scientific analysis......I have a ph.d, I build and evaluate economic models for a living) but ultimately the question of whether a game is balanced or suffers too much from positive feedback loops can only be seen from playtesting as the disparate components cannot be analysed properly in isolation.

From my extensive playtesting I conclude that there is a possiblility for technological progress to snowball (and also this is clearly what BR had in mind in that interview), and I agree with how that has been addressed civ3, but otherwise there is no need for complaint.

All you concrete example shows is that if I have bigger and better cities I will probably win. That's bloody obvious......but in reality one advantage can only be obtained by giving up another, so the case you analyse is of no use whatsoever.

[MrBaggins]

The ONLY culmulative advantages are those that add back to the city production model>

By capturing cities you gain culmulative advantage, not only because you are size+1 and 'they' are size-1... but that ultimately you can improve and grow this city... and gain culmulative values. Strategies creating units (or more advanced units) which do not conquer cities do not create culmulative advantages. They may prevent catastrophic loss, but thats not the same thing.

Techs which improve the city... also beget culmulative advantage, but only by adding back in to the city production model.

Cities are the central pillar of resource, and it still remains that a larger city gains and continues to gain, faster than a smaller city.

Your argument is that the entire game is a discrete 'blackbox' chaotic system, and that that the components (specifically the proposed positive-feedback component of city size) are communally.. and not hierachically, related. I believe that the city production component is discrete enough to use reductionist analysis, as the component is the primal member in that system, and that other components only give culmulative advantages, because they add back to the city production model; the gain is recursive.

Regarding the other point, Brian Reynolds quite clearly using the plural and not the singular case when discussing these matters.

MrBaggins

[MrBaggins]

Furthermore... discovery in a system (Civ2, Civ3) where there can be no alternate case (diminishing returns vs. non-diminishing returns) is useless... how will a system which cannot simulate a diminishing return show the validity of that case?

MrBaggins

[DrSpike]

Well let's test your theories by playing a game of civ........any civ except CTP........though I get the impression you are one of those that would rather analyse the game than play it and gain some true insight.

[MrBaggins]

Originally posted by DrSpike
Well let's test your theories by playing a game of civ........any civ except CTP........though I get the impression you are one of those that would rather analyse the game than play it and gain some true insight.

Well... because diminishing returns can only be simulated in CtP, perhaps??

[DrSpike]

No because that is the game I present no opinions upon since I do not know it. Since you conclusion applies to all civ games......you take your choice of civ2, civ3, SMAC and we will play. You can do whatever it is you think gives you positive vibrations or karmic energy or whatnot and we will see.

[MrBaggins]

Yes, the conclusion applies to all games, but you can't measure the effect in a game which does not give you the option to do so (by giving an option for diminishing returns)

What would you seek to find out, by playing without any ability to measure the quantitive effect of cities being positive-feedback-loops anyway?

but... if you'd like I'll put together the mod for CtP2, and give you a demo...

MrBaggins

[DrSpike]

But I don't think substantial postive feedbacks (other than the tech issue) exist in civ games even if they do not have capacity for diminishing returns.....you seem to think they do.

But never mind......this is now.

[LDiCesare]

Mr Baggins

Back to a few asumptions:
On city sizes: CtP2 has city size limits. Unless you build improvements, yor population reaches a cap. Same in civ, except it is more brutal and teh growth decrease isn't as important. That means the initial lead in you size 10 vs. size 11 cities example, the size 10 will always trail behind, but the size 11 will not have such an exponential advantage because it will have stunted growth. If, f.e., it requires a tech to go from size 12 to 13, it will reach the tech before the other civ, but by then the other civ may have made up for the population disadvantage and be behind only in tech. Staging various city size limits (2 in civ2, more in CtP2) allows to limit the big city size effects somehow.
Note pollution also limits to some extent the exponential gains (although it is usually not very efficient, and, from a gameplay point of view, often found annoying).

About ICS: There is one way not to have it: No cities. In civ, it is ridiculous that in your territory about two thirds of the squares are not worked at all. The cities thing is artificial. If you have population in every square, and the ability to group squares together into regions, then there is no ICS but there is still land-grabbing.
To avoid land-grabbing, look at the number of civs in 4000BC on Earth, consider that Earth was populated almost entirely by then, although by small tribes, and you have a solution: Make the whole map populated at start. The civs will start with 1 , 2, maybe 3 squares, and have to expand into other civs peacefully or by war, but you have lots of civs, and only a few of them should grow. That also makes the big gets bigger a good thing to a certain point, because it allows some civs to start grabbing territory at the expanse of others for the whole length of the game.
This doesn't require any arbitrary limit that is supposed to prevent a strategy but hurts gameplay IMO.

I'd like to see your stance about OCC. This doesn't work in CtP2, or if it does, it doesn't seem to have attracted as many enthusiasts as in civ2. With a single city, you have huge advantages over other players: You are considered weak and puny, never fight offensive wars, so the ai likes you for a long time, which alows you to get techs and gold and trade from the ai for all the length of the game. This means you can in fact be better than a wide empire with a single city, thanks to trading. The gain obtained by city improvements isn't very big in comparison with what you'd get from building a second city, but building a second city would probably make it very hard for you from a diplomatic perspective, and could probably ruin your whole game (unless built very late in the game to rushbuy spaceship). Still, the best strategy at that point is not the one dictated by the mathematical formulas but the one dictated by diplomacy.
I chose that example to show that solutions may exist outside the problem, by adding game features (diplomacy here) that aren't taken into account inside the problem itself.
This is important because limiting factors (number of cities cap, pollution) are not popular things. You should empower people rather than limit them.

Last point, about techs. You have the possibility in civ2 to do, if not diminishing returns, something that works quite similarly: Prevent tech trading/stealing (possible at least in ToT), and make a tech tree with branches that are independant (the tech "tree" is actually a net, not a tree at all). If you choose to advance along all branches at once, you are penalized in terms of cost of research. If you choose only one branch, this branch has ever increasing benefits, but you are more and more limited in terms of other, often basic, advances you can get in a reasonable time.

[MrBaggins]

Yes, CtP2 has a 4 city size caps, effectively only 3 during the civ2 time period. They do retard growth (and everyone is subject to them), yes, but don’t do anything to deal with the production value problems (I.E. city size*increasing modifier). Overcrowding growth reduction works nicely as a parachute. Cities getting to their caps generally have the buildings to deal with these issues, due to their heavy science production aiding movement through the tech tree. I’ll agree that pollution is an annoyance, I don’t like how its balanced now, either… however at least it can be fixed, exactly the way you want.

ICS: I’m not sure I like the no cities approach. Firstly, I’ve always viewed pretty much everything in a civ game as an abstract: cruise missiles aren’t just 1 cruise missile, and so on and so forth. Cities… especially big cities, can represent a region to my mind, and CtP2 is interesting in that you get city spread, insofar as your city spreads out to encompass more land as it grows… it starts at 1 square and grows up to 3 times. The growth points are adjustable.

An approach that Dale, and now I’ve been looking at, is Colonization, that is settling without settlers (at least for the initial game) You start by founding a city by a settler, but your cities can’t produce more settlers. At particular size points, you colonize, as the script searches for an appropriate site, considering distance, terrain, your or foreign units and so on. You can set how fast and where (inland or coastal) that you want to settle and the script also handles the AI’s settling. Settling stops when you reach an empire cap size. Dale has settlers reappear as building options in the renaissance, but I’m not sure whether it’s not a viable option for the entire game.

What I’ve found is that I most certainly can pick better sites than the AI does, given yet more sophisticated weighting.

I, I’ll reiterate, find nothing unhistorical or unrealistic about empire size caps. I find them good and believable facsimiles of the limits of bureaucracy of differing government types. I’ve never been hurt in a game play sense by the Cap… I do plan advancing around it, keeping room for conquest, and maybe sneaking over the cap, if I’m really close to the next government.

I shall definitely check out Clash when it comes out, however.

OCC came about as a reaction to the no-more-worlds-to-conquer situation for the best players, in Civ2. Victory had become formulaic. The game was popular and victory decision trees had been analyzed to death. The original game was released before the trend for releasing patches and so on and so forth…

CtP2 was never as popular as the Civ series... little brand awareness, and generally poor reviews. CtP2 was an odd release... commercially, in that the greatest advantages were effectively invisible; its level of versatility and flexibility. Thus there was no great crowd of players, playing the (badly balanced) original game.

There were and are fans of the series, who, as they had for CtP1, set to work on complete game balance overhauls. The process is no small matter, but is completely viable, and some very finely crafted and tested mods have been produced. Playing these on Deity level is still a challenging exercise and games really do differ, based on situation. If the game is still providing a challenge into AD, why make up an artificial challenge?

The other reason why OCC isn’t viable is that trade has always been different in CtP. Trade affects *just* money. Cities raise commerce from outlying regions, and that is split between science and gold. Gold is multiplied by gold improvements and trade adds to the base gold, not the base commerce. Activision screwed up by drastically altering trade between CtP1 and 2. The number of resources you controlled at a city, mattered and you could build trade monopolies. The figures were out of whack, and I’d prefer that system over CtP2 trade, merely balanced. I guess they must have changed (simplified) it due to the AI’s inadequate usage of it.

Now as to whether you *could* OCC in CtP2… sure… build in a separately balanced government path which allow you to have bigger and bigger production (particularly science) coefficients yet keep the empire size at 1. Do a diplomacy mod… ala diplomod, that encourages dealing with weaker civs fairly, rather than beating on the player just because... The end game is more challenging to ‘solve’, since you’re not building a space ship, but a Gaia Controller; that assumes you can build 10 cities with Controllers and Satellites, not least cover 60% of the world with Obelisks. Tough to do if you’ve only got 1 city…

I agree that Tech ‘control’ is a very important aspect of game balance, however, I want to see how the (scripted and hence highly flexible) knowledge dissemination plays out in final play balance before I start messing with the tech tree... especially since the tech tree is going to have split paths... for different religious ethos' already.

MrBaggins