The Social Policy Situation

[Zachriel]

Secondly, what do you mean "average"? Average of the number of cities I've ever had given every turn of the game? That's extremely hard to compute.

The average is the weighted average over time. And the easiest way to determine the weighted average is exactly your proposed mechanic. If you take on a bunch of new cities temporarily, it will only affect culture accumulation temporarily. If you gift away a bunch of cities, it will only affect newly accumulated culture, so it resolves the exploit.

It is the 'correct' mechanic. Notably, no one has yet mentioned a significant flaw or exploit.

 

[Louis XXIV]

It only fixes one side of the coin. I read this after I wrote what I did above. You're taking away the retroactive benefit a player sees when a city is sold/razed, and that's cool and kills a lot of exploiting. But you're still not even touching the retroactive harm a player sees when a city is built/bought. I also think this method might be a bit opaque, because who likes to see a "-200 culture lost!" bubble pop up?

What is the retroactive harm a player gets? It is intended that you have higher costs from more cities, since those additional cities can produce additional culture. It's not retroactive, it's proactive

They're trying to fix the cheesy exploit of selling all your cities so your number of social policies can skyrocket.

 

[Celevin]

The average is the weighted average over time. And the easiest way to determine the weighted average is exactly your proposed mechanic. If you take on a bunch of new cities temporarily, it will only affect culture accumulation temporarily. If you gift away a bunch of cities, it will only affect newly accumulated culture, so it resolves the exploit.

It is the 'correct' mechanic. Notably, no one has yet mentioned a significant flaw or exploit.

Thanks for that boost of confidence. I'm fairly certain that there's no holes in my reasoning. What needs to be done now is show just how large of a problem my proposal actually covers. I think there's a lot of people who are still unconvinced.

What is the retroactive harm a player gets? It is intended that you have higher costs from more cities, since those additional cities can produce additional culture. It's not retroactive, it's proactive

They're trying to fix the cheesy exploit of selling all your cities so your number of social policies can skyrocket.

I described it here:

Take a 5 city empire on a standard map. You just bought a policy, and are 20 turns from your next one. When you build a city on the xth turn towards your next policy (where 1 is now, and 20 is the next policy), how much did it set you back? Assume that this new city produces 0 culture for at least 20 turns.

Before and after building the new city, the policy costs are multiplied by 2.2 and 2.5 respectively.

Say I built the new city on turn 1. Then through every turn the average contribution from each of my 6 cities towards my policy gain rate is divided by 2.5 from turns 1 to 20. OK, that's fair, I built that new city right away and am being penalized for it on every turn of the 20 turns.

Now say I built the new city on turn 10. The average contribution from each of my 6 cities towards my policy gain rate is once again divided by 2.5 from turns 1 to 20. So I'm being hurt retroactively for turns 1 to 9. Doesn't this feel like micromanaging overflow at this point, just in a more opaque way? We lose valuable turns towards our policy not based on a strategic decision, but micromanaging city build time. What about a more extreme situation, like a 3 city war trade? If that happens right before you gain a new policy, it could easily set you back 10+ turns.

I would argue that in our second scenario, the average policy gain rate should be divided by 2.2 for turns 1-9, and divided by 2.5 for turns 10+. My proposal accomplishes this.

 

[Slowpoke]

Really, it all depends on how its implemented. We got a week or two to find out and who knows? Maybe they already took razing into account and do not coutn it now. Maybe while your policy costs can never go down if you build another city it won't go up until you've surpassed any previous city level. That would be just fine. Remember we're only hearing an abbreviated version from 2k greg, and likely the coding for the bolded statement is actually easier than it sounds, perhaps even being the default.

 

[rikkia]

Celevin I agree with Slowpoke, depending on how it feels in game I may or may not like the new mechanics however...

That aside you should definitely look at options to mod this proposal post patch(or if not possible post patch then post .dll for sure).

It'll give you an actual version people can play with and see exactly how it'll work in-game.
You'll be able to identify what, if any, exploits remain and use actual player feedback to support your argument for the mechanic.

I for one would actually like to play with an organic culture mod, as I doubt the devs would implement your idea as is.

 

[neptune2000]

Really, it all depends on how its implemented. We got a week or two to find out and who knows? Maybe they already took razing into account and do not coutn it now. Maybe while your policy costs can never go down if you build another city it won't go up until you've surpassed any previous city level. That would be just fine. Remember we're only hearing an abbreviated version from 2k greg, and likely the coding for the bolded statement is actually easier than it sounds, perhaps even being the default.

I thought that was the assumption, that policy costs were calculated by the max # of cites, so that if you had 15 then lost some, they stay at 15 and would go up iff you got > 15.

Otherwise I agree, they may have at least taken razing into account, or done something that is just summarized as "policy costs never go down". But it's fun speculating, and there've been some good suggestions in the meantime.

 

[Louis XXIV]

I described it here:

Like I said, that's proactively, not retroactively

Still, their goal here was to fix an exploit people were using for cultural victory. What you're describing is the known way the system works. I think, unless the system is demonstrably unfair/unfun, they have greater priorities.

 

[Celevin]

Really, it all depends on how its implemented. We got a week or two to find out and who knows? Maybe they already took razing into account and do not coutn it now. Maybe while your policy costs can never go down if you build another city it won't go up until you've surpassed any previous city level. That would be just fine. Remember we're only hearing an abbreviated version from 2k greg, and likely the coding for the bolded statement is actually easier than it sounds, perhaps even being the default.

I assumed that's exactly what it is.

Here's how I think the post-patch works: If you have 10 cities, you lose 2 and go down to 8, you have a policy multiplier of 10. If you then gain 2 more cities and go back to 10, you have a policy multiplier of 10.

I see the problem with how this is set up. You can't lose any cities below your maximum without being hurt big time. No_Such_Reality described a perfectly normal situation that, post-patch, becomes really painful.

I also described a situation, that pre or post patch, is also painful. If you build/capture new cities, the next policy you buy has a price tag equal to if you had all that number of cities all along. This is what I mean by retroactive damages. If you build that new city 5 turns into your current policy, you're priced as if you had that new city from the 1st turn into your current policy. How is that fair?

 

[Rohili]

Celevin - Have you tried posting your suggestion over at the 2K forums yet? It would probably be more likely to be read by Greg over there, and your feedback transmitted to the game designers themselves....

 

[PieceOfMind]

TLDR version:
Instead of altering the global culture rate as you gain or lose cities, the same gameplay effect can be achieved by altering the amount of culture needed for the next policy and making that change depend on how far you have progressed towards that next policy.




Note in this post I use and to distinguish two different formulas. (EDIT... Hmm it probably wasn't necessary in the end, but it makes the post more fun so what the heck. )

f you build that new city 5 turns into your current policy, you're priced as if you had that new city from the 1st turn into your current policy. How is that fair?

Good question, and it has helped make your idea 'click' to me. No matter the implementation, I see now that what you're essentially trying to do is smooth out the discretization of this game mechanic.

A comparison can be made to how combat worked in civ4. Two units of equal hitpoints and strength had 50% odds when facing off against each other, but if one of them had 99hp instead of 100, the odds change drastically. These 'jump points' had interesting consequences, one being that the effect of promotions (like combat I) were not very intuitive or obvious.

You want adding extra cities to be a rate-altering effect rather than a threshold-altering effect. It reduces the micromanagement in waiting for the various jump points (when cities are founded or lost).

A thing to note is that this effect is possible to achieve in a system where you alter the thresholds rather than the rates with additional or lost cities. (explained later below)

Let me give an example. Suppose (using your sort of system) you want your culture rate to be multiplied by 0.75 for having two cities, where it would have been 1.0 if you had only one city. i.e. 25% penalty. Suppose also that you're currently generating 1 per turn and the next policy is at 10.
Build the 2nd city on turn x (between 0 and 10) and you would have to generate x + (10-x)/0.75 culture to get the policy.

Note that this slightly resembles a 'weighted average' calculation. To be clear, rewrite it as:

1.0 * x + (4/3) * (10-x).


The earlier you build the 2nd city (the smaller x is), the larger the weight you give to the slower rate (3/4 or 75% rate) and the smaller the weight you give to the 1.0 (original) rate.

If this is significantly different to the sort of mechanics you're talking about let me know because this is how I'm understanding it.

Notice I said earlier in this post that it's possible to achieve the same effect with altering the thresholds. Maybe it's obvious, but if not I'll try to do this for the same example presented just a moment ago.

So once again we have our next policy taking 10 culture and currently generating 1 per turn. Build 2nd city once you have accumulated x culture (so it is turn number x again).

To produce the same effect as the rate-change by doing a threshold-change instead, and assuming we don't want to instantaneously change the current amount of culture put towards the policy (i.e. we won't just 'cheat'), then we just need to change the threshold to:

x + (10-x) / 0.75

(i.e. it's essentially the same thing)

Basically with this change it would make the culture threshold bump up by a smaller amount if you are close to the next policy. If you lost a city, the threshold would be reduced in basically the same way (reduced by a smaller amount if you are near to the next policy). Rather than having global culture rate vary up and down as we gain and lose cities (respectively), we'd have culture threshold vary instead but using the fairer formula than the existing game.

Now what is left to argue is which feels more intuitive or looks nicer (changing the rate at turn x or changing the threshold at turn x ). Personally I feel that neither of them is clearly better than the other, but the latter w00t stays closer to resembling the current system and achieves the same positive effect as what you desire (just using a different way to look at the numbers).

 

[Slowpoke]

I assumed that's exactly what it is.

Here's how I think the post-patch works: If you have 10 cities, you lose 2 and go down to 8, you have a policy multiplier of 10. If you then gain 2 more cities and go back to 10, you have a policy multiplier of 10.

I see the problem with how this is set up. You can't lose any cities below your maximum without being hurt big time. No_Such_Reality described a perfectly normal situation that, post-patch, becomes really painful.

I also described a situation, that pre or post patch, is also painful. If you build/capture new cities, the next policy you buy has a price tag equal to if you had all that number of cities all along. This is what I mean by retroactive damages. If you build that new city 5 turns into your current policy, you're priced as if you had that new city from the 1st turn into your current policy. How is that fair?

Well, if you regain those cities, and it shouldn't be difficult, there is no loss anywhere. I don't think it'll be as big a problem as you think.

 

[LAnkou]

well, the problem they want to solve is:
accumulating a lot of culture with many cities, then selling all of them to get back to really low amount of SP needed

Solution they put in place:
-amount of SP needed never go down

situation where amount of SP go down (and if we want to keep it that way for that particular situation):
-lose a city (keep)
-sell a city (change)
-gift a city in a peace treaty (keep)
-liberate a city/CS (keep)
-destroy a city (keep)
-wonder that makes SP cost cheaper (keep)
-SP that makes cost cheaper (keep)

Solution i propose to the problem
-trading a city outside of a peace treaty don't lower SP cost

New exploit:
-accumulate a lot of culture
-declare war
-at peace treaty, give all cities but capital
-pray that the civ doesn't redeclare war while you build project utopia

Honestly, i'm not sure it's the most effective strategy for culture victory

 

[Slowpoke]

Edit: Whoops, wrong thread

 

[André Alfenaar]

Maybe it helps to concentrate on the question what should happen to the SP cost when founding a new city. The answer to the question what should happen when losing one city automatically follows.

For instance the situation in which it takes 10 turns to get a new SP. If the player founds a new city in turn 8 then the cost of this SP will increase to 13. The same would happen if instead a player founds a new city in turn 1. This doesn't feel right. Instead, the number of turns that the city wasn't there sould be accounted for. So when founding a city in turn 8 the SP cost should increase from 10 to 11 -not 13-.

So reversily: when the SP-cost is 11 and one loses a city in turn 8 the SP-cost should decrease to 10 turns.

 

[PieceOfMind]

Maybe it helps to concentrate on the question what should happen to the SP cost when founding a new city. The answer to the question what should happen when losing one city automatically follows.

For instance the situation in which it takes 10 turns to get a new SP. If the player founds a new city in turn 8 then the cost of this SP will increase to 13. The same would happen if instead a player founds a new city in turn 1. This doesn't feel right. Instead, the number of turns that the city wasn't there sould be accounted for. So when founding a city in turn 8 the SP cost should increase from 10 to 11 -not 13-.

So reversily: when the SP-cost is 11 and one loses a city in turn 8 the SP-cost should decrease to 10 turns.

Exactly. This is what I propose in my previous post. It can be made to work the same way (gameplay wise) as altering the culture rate as Celevin proposes.

 

[alpaca]

I agree with both the analysis and the solution you propose. In fact, maybe the most important advantage to your proposed system for me personally is that you could easily compare between different games and game states. What I would do is displaying something like 1000*culture/z versus base policy cost. The 1000 would make sure you don't lose too much info due to rounding and don't have "ugly" floating point numbers.

 

[Balnor]

I agree with the OP's suggestments, yet for simplicity's sake I would just fraction the total numbers instead of the income/turn as you did. This would also solve the issue some people had with your culture output decreasing when you gain more cities.

To put it into concrete terms : Whenever you gain/lose a city, the game should determine your current progress to the next policy in %, next determine the threshold from your new number of cites, and finally change your accumulated culture to stay true to the percentage you had before.

Simple example :

Say you were at 50/100 towards your next policy. You now add a city that would raise your cap towards 200. The game would now just carry over the fraction towards your new threshold =>
50/100 is 50%; 50% of 200 is 100; => your new culture is now at 100/200.
I know the actual values differ, just used these to keep it simple.
Finally, instead of displaying the 50/100, the UI should just show the percentage towards the next policy, in my example 50%.
This way players would not get confused by the total numbers switching all the time when they add/lose cities, yet they can still clearly see what progress they are making (you could keep showing the total numbers on mouseover or something).
The nice thing about this is also, when playing without that feature where you have to pick SP the turn you get them, these percentages could increase to over 100%... when you hit 200% you would know that you now have exactly enough to unlock 2 policies if you so choose.

Easiest and most elegant fix IMO, does not allow any abuses with gifting cities, and the system doesn't encourage you to settle your cities only the turn after a SP gain as it does now. It is basically the same system the OP uses though (whether you fraction the income by # of cities or always try to keep the total % true when thresholds change should play exactly the same), so not trying to take any credit from him here.

 

[DSChapin]

OP is really the best solution: it works well mathematically, and also solves some other existing problems like the fact that you want to wait to settle until the turn you buy a policy.

It's conceivable that reducing your apparent rate of culture income might confuse people, but I don't think it would be too unclear. Heck, they could give the factor a name (like "corruption"), and it would be no more unjustified than the present increasing-cost-per-city.

 

[Celevin]

Well, if you regain those cities, and it shouldn't be difficult, there is no loss anywhere. I don't think it'll be as big a problem as you think.

What if I want to keep my empire at 10 cities though, and not at 13? Maybe I'm going for a culture victory. In the post-patch game, I would (as you suggest) I'm better off keeping my spoils. Then in another war the same problem can happen again, and again. It's like I'm being forced to increase my empire size.


@PieceOfMind:

Checked over your numbers, that works as well for ironing out those threshholds. It was a lot to take in . Let's ignore the "no policy saving" change right now, and say I have a tonne of culture built up (enough for at least 5 policies). What happens when I sell all but 1 of my cities with your proposed change?

Since each policy bought will cost different amounts, we'd have to proceed with caution. It could be done through a counter of "this is how many policies you can buy" instead of saved culture. Due to this, I think your proposal would require a UI change. Even if I am forced to buy policies on the spot, it would still require this change, as selling 20 cities can allow me to go from 0 buyable policies to 3.

I enjoy that your system gives no more nasty culture levels than normal unlike mine. What I mean is I intend on altering culture values with a double when my culture output's only maybe 3, which can have nasty UI results. I also love that your system is convex. My inner geek is *proud*


@Balnor:

Excellent post. Your idea's just right along the same lines as PieceOfMind, but uses percent values instead. I think people might get a bit confused when they build a new city and see both their culture and culture required levels jump though . Very elegant! And it works very well with stored policies.


Both of these and my proposal are all viable solutions. Considering the damage level of not micromanaging your culture is at the same damage level as science overflow, I would be a very happy person to see any of these implemented.

 

[Zachriel]

OP is really the best solution: it works well mathematically, and also solves some other existing problems like the fact that you want to wait to settle until the turn you buy a policy.

It's conceivable that reducing your apparent rate of culture income might confuse people, but I don't think it would be too unclear. Heck, they could give the factor a name (like "corruption"), and it would be no more unjustified than the present increasing-cost-per-city.

It has the advantage of leaving culture cost the same throughout the game. The player will immediately see the change in accumulating culture, and the mouseover could provide the dilution factor.