Rebalancing Proposal

[wobuffet]

I agree that it's a no-brainer when pursuing a cultural victory, regardless of multipliers. I don't know why the multipliers make it "far too extreme," unless you think this particular factor makes CV's too easy compared to other VC's.

Culture modifiers are easily the most extreme: in any typical, halfway-successful Culture game, your capital (with the Hermitage and GWs like the Louvre or Sydney Opera House, etc.) will have something like a +250% modifier by turn 250 or so... not to mention that putting the city on "Culture" yield/government mode funnels pretty much all of the imported culture to that city, too.

I agree that it's a no-brainer when pursuing a cultural victory, regardless of multipliers. I don't know why the multipliers make it "far too extreme," unless you think this particular factor makes CV's too easy compared to other VC's.

This is definitely part of it, although I'm hopeful that fixing the Landmark producing base 12 bug after some SP(s) should help that.

Food from citystates has always benefited from modifiers in cities (such as We Love the King Day). If culture from citystates is too strong, I'd prefer to reduce the effect rather than reverting to a situation where the two yields operate under different rules.

I actually think any of the yields imported from City-States should be flat, unaffected by domestic multipliers. It makes sense to me that if Civ Y replaces Civ X as Kuala Lumpur's Ally, the benefits accruing to Y should be the same as those X just enjoyed (the fish or books going to one empire are just going to another now).

As a side note, I think it's not too implausible that imported food benefits from city-specific buildings (food is food, and efficient food distribution systems apply equally to homegrown and foreign origin foods), but that culture imported from foreign lands does not benefit from buildings/policies specific to home culture (Lhasan culture is not French culture).

That's my point - I think the early cost of RAs is probably okay, because the alternatives appear better/equal to RAs. However, a declaration of friendship does make RAs dramatically better, so if we can get a DoF then the RA is certainly valuable. It's also worth pointing out Seek's point is primarily aimed at the early game, when we don't have modifiers yet.

This is a good point. Still, I don't imagine many circumstances where 50 for an early-game RA (typically +6% or more Science production for 30 turns) is not very easily worth it.



That said, at the moment I'm more concerned with checking out how the new AI gold spending and the new Opportunities work.

 

[Seek]

Consider it from the perspective that RAs convert a set amount of gold into a set amount of science, as Zaldron and orangecape pointed out. From that angle the value depends on our science income. I hadn't thought about this perspective until recently. I want to thoroughly discuss and test it, which is why I've been experimenting with RA costs proportional to science income.

This graph shows the income and cost from a research agreement. It assumes our science is equal to the AI's. If we have the GL/NC, our personal science is higher but not the AI's, so the values on the left side of the scale are lower in practice.



Consider a game where it's taking 15 turns to earn classical techs and we do not have a DoF. We could spend gold on:

• +5/turn from five RAs costing 500 over 30 turns.
• +5/turn from a Library in a 5 pop city for 490* over 30 turns.
• +9/turn from friendship with a cultural CS for 500.

If we earn the CS's friendship for 17 turns, that's 150 from the citystate, 150 from the RAs, or 150 from the Library. Is this a good setup? I honestly don't know, which is why I want to discuss and test it.


-----------
* I believe a Library costs 430 + 2/turn... but I had to calculate this manually and can't verify it, since I'm unable to log into the game right now.

I don't think this representation is accurate because

1. The Library is 1/2 so the actual yield is only 2-3. (The footnote is accurate regarding Library cost).
2. There is still the vanilla RA effect which comes at the end of the RA granting a not- insignificant boost, so for argument's sake let's say it's usually around 30, +1 per turn.
3. While RAs might grant 1/turn if we sign them right after discovering Philosophy, I would argue that it's not unreasonable to say that they grant closer to 2/turn once the 30 turns have passed. But to be fair I won't include this in the summary below.

So more accurately: 500/30 turns would yield 75 from a Library, at least 300 from 5 RAs, and 150 from a CS. (Also worth noting is that unless playing on Pangea it's rare to encounter more than 4 civs before the Medieval era and if making a gambit for the Great Library it's counterproductive to buy a Library.)

RAs seem much more efficient if we take these factors into consideration, and since science is so especially important in this phase of the game and the cost so negligible, it's a no-brainer to sign any and all RAs we can. 300 may be too high (that's just what seemed right to me) but I strongly think that they should cost 200 *minimum* in the Classical era.

 

[Txurce]

(Also worth noting is that unless playing on Pangea it's rare to encounter more than 4 civs before the Medieval era and if making a gambit for the Great Library it's counterproductive to buy a Library.)

This mirrors my games.

RAs seem much more efficient if we take these factors into consideration, and since science is so especially important in this phase of the game and the cost so negligible, it's a no-brainer to sign any and all RAs we can. 300 may be too high (that's just what seemed right to me) but I strongly think that they should cost 200 *minimum* in the Classical era.

At 300g I may or may not sign one RA (and no more) in the Classical era. At under 200g, I would mortgage Boardwalk and Park Place to rush a trireme and sign as many as I possibly could. In terms of making me weigh the choice seriously, the sweet spot is on the higher end of 200-250g.

 

[Zaldron]

Well, I think things are more interesting the more situational they are. At present, buying from most if not all of the cultural CSs is a no-brainer when pursuing a Cultural victory. Yes, this might be the case even without multipliers affecting CSs' culture contributions, but their existence makes it far too extreme.

As a result, it's much, much less (comparatively speaking) worth it for a player not pursuing a Cultural victory to buy cultural CSs' favor. I find, for example, that having a Militaristic CS Ally or two is great when I'm playing a peaceful game. When I'm playing a Cultural game, my capital has ridiculous multipliers that it seems unfair to apply to culture incoming from CSs. (On a related note, I'd like to see more base and smaller/no multipliers on Culture buildings [or at least NWs and GWs], to make the "Culture = small/tall empire" dynamic less extreme and offer the player more flexibility.)

If you change CS culture to not be affected by multipliers then it becomes a no-brainer to *never* by cultural CSes after the early-mid game, because your other culture sources dwarf the un-modified culture you get from CSes.

Note that even with culture multiplying, for a non-cultural pursuing civ a cultural CS provides approximate the same level of culture increase that it does for a culture-pursuing civ.

I personally haven't observed the final thing you noticed regarding small vs large empires. As long as I pick up the liberty policy and build (or buy with the extra money from more cities) a few culture buildings, new cities don't slow the culture rate at all. Thal even showed in some circumstances that additional cities up to about 10 continue to improve the rate at which you attain policies.

 

[wobuffet]

If you change CS culture to not be affected by multipliers then it becomes a no-brainer to *never* by cultural CSes after the early-mid game, because your other culture sources dwarf the un-modified culture you get from CSes.

That's a scaling issue then: what you're saying is, culture from cultural CSs should increase faster with era to account for generally higher multipliers. I agree.

Note that even with culture multiplying, for a non-cultural pursuing civ a cultural CS provides approximate the same level of culture increase that it does for a culture-pursuing civ.

This is an empirical claim. I can't go ingame and test it at the moment, but I'll check out the numbers next time I play. When I'm not pursuing a cultural victory, Opera Houses and Museums (and even Temples) are far from a priority for me and don't get built until significantly after I research the corresponding techs (if at all).

 

[Txurce]

This is an empirical claim. I can't go ingame and test it at the moment, but I'll check out the numbers next time I play. When I'm not pursuing a cultural victory, Opera Houses and Museums (and even Temples) are far from a priority for me and don't get built until significantly after I research the corresponding techs (if at all).

And this is a good example of a point I tried to make earlier: that you are basing much of your argument on how you play. Others, as you've probably noticed by now, play differently. For example, I always build all the way to Broadcast Towers by the end of my Science games.

 

[wobuffet]

Very possible, Txurce. My main point is, we as players have more interesting options if CSs are of similar value regardless of which victory condition we're pursuing.

If we had to choose between cultural CSs that either [A] gave a flat amount (say, 10 * era) culture per turn or increased its Allied empire's culture output by some fraction (say, 20%), we both agree that world A would be much more fun. We just disagree on how close to world B multipliers take us (which, as you correctly point out, depends largely on playstyle).

 

[Zaldron]

I just realized that I misstated what I wanted to say. Instead of "Note that even with culture multiplying, for a non-cultural pursuing civ a cultural CS provides approximate the same level of culture increase that it does for a culture-pursuing civ."

I meant something more like "Note that even with culture multiplying, for a non-cultural pursuing civ, a cultural CS provides approximately the same percentage of culture increase compared to what it does for a culture-pursuing civ. Note that total gain for the cultural civ will be more, but the increase to the rate of policy acquisition will be about the same. To elaborate further, one civ may go from 20 to 18 turns per policy while another goes from 10 to 9."

EDIT: My concern is avoiding the case where a cultural civ goes form 10 to 9 turns per policy while a non-cultural civ goes from 20 to 10 turns per policy.

 

[Thalassicus]

I don't imagine many circumstances where 50 for an early-game RA (typically +6% or more Science production for 30 turns) is not very easily worth it.

I believe it's 100... or at least, it is in my development version, which will be released once I have access to the game again.

The Library is 1/2

I made it 1 per 1 as part of the science-scaling in v131.1... or at least it should be. It's a bug if it's still the old value.

That end-of-RA bonus is so annoying. I wish I could get rid of it... keeps messing up game balance. I'd forgotten about it and you're right, that does unnaturally skew the results.

Zaldron has a good point. If we consider the effect one cultural citystate has on the # turns to get a policy, I suspect it would be about the same for either cultural or non-cultural victory strategies. Players focusing on a cultural victory have much more culture to start with, so the citystates increase that by less of a percentage.

 

[wobuffet]

I believe it's 100... or at least, it is in my development version, which will be released once I have access to the game again.

I couldn't remember whether they initially cost 100 or 50, but since the proposed change was to halve all RA costs, I figured I was covered either way.

Players focusing on a cultural victory will have much more culture to start with, so the citystates increase that by less of a percentage.

That's exactly what I want! I'm just lobbying for a larger percentage decrease for culture-heavy civs than what we currently have.

Zaldron has a good point. If we consider the effect one cultural citystate has on the # turns to get a policy, I suspect it would be about the same for either cultural or non-cultural victory strategies.

If gaining my first cultural CS ally reduced my time to next policy from 20 to 18 turns, that means my culture output increased by 20/18 = 11.111%.
If gaining my first cultural CS ally reduced my time to next policy from 10 to 9 turns, that means my culture output increased by 10/9 = 11.111%.

I think this % should be smaller for empires that are already producing lots of Culture. Obviously I haven't been very clear about what I mean so far, so to be explicit, here's what I'm saying:

Status quo:
(1) Pursuing a cultural victory? Cultural CSs are a must-have and give you a huge edge (esp over AI civs who are also going for a cultural victory)
(2) Not pursuing a cultural victory? Typically cultural CSs are meh, especially if you don't have many cultural buildings up yet to multiply/modify the imported Culture. (11.111% of some initially small domestic Culture yield is, obviously, small.)

What I think would be better:
(1) Pursuing a cultural victory? Cultural CSs are somewhat helpful, but not always the best use of gold – esp. by later points of the game.
(2) Not pursuing a cultural victory? You can either get some relatively cheap culture (in terms of and/or ) by putting up the early Culture buildings you've been putting off, and/or grab a couple cultural CSs. What's best depends on your short- and medium-term gold and building plans.

In essence, flat rewards would pose more interesting, situational decisions to the player.

 

[Thalassicus]

I disagree with #2 in the status quo section. Cultural citystates are my favorite CS type for conquest games:

• Conquerors typically have negative happiness, reducing the value of food from maritime citystates.
• We can more quickly produce units than militaristic citystates, of a better quality, and control what we get.
• Cultural buildings are a lower priority than military training and production.

In an early conquest game I typically have a Monument and garrison in each city. With five cities and a palace that's about 25. Since a medieval-era cultural citystate provides 20, each CS increases policy rate by about ~80%, depending on circumstances from the game.

If we have a comparable culture game with monuments and temples in each of say... 3 cities, plus a National Epic in the capital, we have about 38. A cultural citystate with half the yield going to the capital would increase that by 23, only a 60% increase.

The same thing can be extrapolated to midgame scenarios where the main multiplier becomes available (opera houses). Also, a peaceful empire will typically have more citystate allies, and each additional citystate gives diminishing returns (40 to 60 is a larger % jump than 60 to 80).

 

[Txurce]

I disagree with #2 in the status quo section. I feel cultural citystates are the most useful CS type for conquest games.

I probably get 80% of my culture from cultural citystates in a conquest game.

I made a similar point earlier about science games. Depending on how I'm doing on pop, I sometimes favor Cultural over Maritime. I would rate them 1a and 1b rather than 1 and 2. (And of course often have one Militaristic for defense. I find the current state of the CS to work really well.)

 

[wobuffet]

Thanks for the math, Thal. I suspect that flat Culture rewards from Allied cultural CSs equal to maybe 10–15% of domestic Culture output, assuming we're playing a decently successful Culture Victory game so far, would work well. (I have no problem with the +80% Culture per CS figure during a conquest game quoted above.)

Just normatively speaking, if you've already done all you can in your own Cities to maximize the amount of homegrown Culture you're generating, it seems to me that shoveling Gold into buying out 4 Cultural CSs should increase your total output by no more than, say, 50% or so. I'd calibrate the Culture per turn from Allied cultural CSs using some such standard.

edit:
The deeper problem I have with cultural CSs at the moment is that buying them all out is both easy and incredibly effective when playing cultural games; the human player's ability to recognize cultural CSs as important and maintain them as Allies when pursuing a Cultural victory makes achieving such a victory far easier than I (admittedly subjectively) feel it should be. The AI gold spending fix has a good chance of eliminating these concerns altogether.

The other half of this, from my point of view, is that I think tying Cultural victories so closely to the number of cities means in practice that all Cultural games end up feeling very drab and similar: 3–5 cities, tons of population in each, tons of buildings in each, and tons of GWs, NWs, etc. in the capital. I strongly feel that [A] reducing or eliminating Culture modifiers attached to Wonders (e.g., making the National Epic +10% Culture empirewide instead of +25% in that city) and increasing base Culture building yields and such would allow the player more flexibility and, by extension, more fun when pursuing Culture victories.

 

[Seek]

I believe it's 100... or at least, it is in my development version, which will be released once I have access to the game again.

Yes, right now it's 100/Cla, 300/Med, 1000/Ren, 3000/Ind, 6000/Mod.

I made it 1 per 1 as part of the science-scaling in v131.1... or at least it should be. It's a bug if it's still the old value.

Really! I had no idea - the notes only mention a hammer cost increase and I probably haven't looked at the tooltip since release. I understand where you're coming from wanting to link RA cost with gain, but the availability of gold plays a large factor in what we consider worth spending it on. Since generally by the time we enter the classical era we've met all the nearby civs (and sold ~1000 worth of goods) the 100 cost is negligible. Heck, in my most recent game I signed one with Monty even though it's almost guaranteed that he will dow me before the RA finisher comes through, just because it was so cheap - there was no reason not to!

I strongly feel that reducing (the importance of) modifiers and increasing base Culture building yields and such would allow the player more flexibility and, by extension, more fun when pursuing Culture victories.

I'm not sure I follow you here - it's my understanding that the multipliers allow for culture victories in wider empires (albeit slower ones than small/tall empires) where that would be impossible in vanilla - can you expound on this? I would think that going in the opposite direction would grant more flexibility, perhaps removing one culture from the Temple and adding a small multiplier around 10% (maybe lowering the Opera House multiplier in response) - I think this would make the building more desirable in cities without the affected luxury resources, give slightly greater weight to cultural CS and give greater flexibility to expand more while gaining policies at a reasonable rate.

 

[Txurce]

I understand where you're coming from wanting to link RA cost with gain, but the availability of gold plays a large factor in what we consider worth spending it on. Since generally by the time we enter the classical era we've met all the nearby civs (and sold ~1000 worth of goods) the 100 cost is negligible. Heck, in my most recent game I signed one with Monty even though it's almost guaranteed that he will dow me before the RA finisher comes through, just because it was so cheap - there was no reason not to!

Thal, 100g in the Classical is much too low. Just compare it to 3000g in the Industrial, where I would usually be able to sign no more than two over the duration, and never two at a time. (And this could be even lower now, given the nerfs.) Oddly, the opening is one of the few times when I feel I have a good bit of gold - partly because I only have one or two cities to spend it on. I enjoy having to decide where to focus: rushing a building, bribing a CS, or buying an RA. As I mentioned earlier, at 100g I wouldn't even think twice; it would be all in for RA's and the resultant tech lead that would allow me to blast my neighbors (even playing science).

 

[wobuffet]

I'm not sure I follow you here - it's my understanding that the multipliers allow for culture victories in wider empires (albeit slower ones than small/tall empires) where that would be impossible in vanilla - can you expound on this? I would think that going in the opposite direction would grant more flexibility, perhaps removing one culture from the Temple and adding a small multiplier around 10% (maybe lowering the Opera House multiplier in response) - I think this would make the building more desirable in cities without the affected luxury resources, give slightly greater weight to cultural CS and give greater flexibility to expand more while gaining policies at a reasonable rate.

Sure. I was a bit unclear: I'm talking about the various modifiers attached to National Wonders (National Epic's +25% and the Hermitage's +50%) and Sydney Opera House (+50%) as well as things like the Piety Social Policy that, among other bonuses, gives +10% to any city with a World Wonder. In cultural games, all these add up to stacking all your Culture in one city: (almost always) your capital.

Here's some screenshots from a Culture game that I recently completed (turn 262, just after building my Broadcast Towers): the tooltip's numbers are off, but arithmetic shows that my capital has a +243% Culture modifier (1581.3/461 = 3.43), while my other non-puppet cities have only +118% (237.1/108.8 = 2.18). All my landmarks are in the capital as well (156 from terrain, although that's probably too high because of a bug that's been reported since), not to mention World Wonders, factors like the Palace, and indirect things like the Museum (per-population yields favor the capital, because of the capital emphasis for food bonuses from maritime CSs as well as capital-favoring SPs).

Spoiler :

This leads to, in practice, almost all Culture coming from your capital city.

Spoiler :

Modifiers that are attached to regular buildings (not Wonders) do not affect the wide/tall balance, as far as I know. However, modifiers attached to Wonders and/or the Capital make culture games much slower if you go wide.

 

[Seek]

Ah, okay - I misunderstood you and thought that you were arguing against the added multipliers from normal buildings such as the opera house and museum.

 

[Zaldron]

Sure. I was a bit unclear: I'm talking about the various modifiers attached to National Wonders (National Epic's +25% and the Hermitage's +50%) and Sydney Opera House (+50%) as well as things like the Piety Social Policy that, among other bonuses, gives +10% to any city with a World Wonder. In cultural games, all these add up to stacking all your Culture in one city: (almost always) your capital.

Here's some screenshots from a Culture game that I recently completed (turn 262, just after building my Broadcast Towers): the tooltip's numbers are off, but arithmetic shows that my capital has a +243% Culture modifier (1581.3/461 = 3.43), while my other non-puppet cities have only +118% (237.1/108.8 = 2.18). All my landmarks are in the capital as well (156 from terrain, although that's probably too high because of a bug that's been reported since), not to mention World Wonders, factors like the Palace, and indirect things like the Museum (per-population yields favor the capital, because of the capital emphasis for food bonuses from maritime CSs as well as capital-favoring SPs).

Spoiler :

This leads to, in practice, almost all Culture coming from your capital city.

Spoiler :

Modifiers that are attached to regular buildings (not Wonders) do not affect the wide/tall balance, as far as I know. However, modifiers attached to Wonders and/or the Capital make culture games much slower if you go wide.

I actually dislike the 10% for world wonders because it's so non-deterministic. I feel however that percent modifiers for national wonders are an integral part of making tall empires viable. It's interesting to note in your second screenshot that even though your capital was generating a fair majority of your culture, the other cities were still all improving the rate at which you attain policies, which is really the bottom line at the end.

I believe what often makes wide games seem much slower in culture is that the focus is typically science instead of culture. If you focus on a cultural victory that's still easily attainable going wide. With the proper buildings and policies it may even be FASTER for the wide empire (so much more raw culture).

 

[wobuffet]

It's interesting to note in your second screenshot that even though your capital was generating a fair majority of your culture, the other cities were still all improving the rate at which you attain policies, which is really the bottom line at the end.

I don't think they were, actually, but the Culture formulas are confusing. Can you show your math?

It looks like it's either a non-compounded +15% or +30% per city without the Liberty policy (see links), meaning that the 2nd city would have to yield 15% or 30% of the capital's Culture output. But 237/1581 = 14.99% < 15%. Add in that with fewer cities, more City-State-imported Culture is funneled through the capital's very high multipliers, and it seems clear that having more Cities is not helping my SP acquisition rate.

http://forums.civfanatics.com/showthread.php?t=438415
http://forums.civfanatics.com/showthread.php?t=451033
http://forums.civfanatics.com/showthread.php?t=419620

 

[Zaldron]

I don't think they were, actually, but the Culture formulas are confusing. Can you show your math?

It looks like it's either a non-compounded +15% or +30% per city without the Liberty policy (see links), meaning that the 2nd city would have to yield 15% or 30% of the capital's Culture output. But 237/1581 = 14.99% < 15%. Add in that with fewer cities, more City-State-imported Culture is funneled through the capital's very high multipliers, and it seems clear that having more Cities is not helping my SP acquisition rate.

http://forums.civfanatics.com/showthread.php?t=438415
http://forums.civfanatics.com/showthread.php?t=451033
http://forums.civfanatics.com/showthread.php?t=419620

If I understand correctly for the last few months it's an additive 15% per city without the liberty policy or 10% with the policy. That said, I misread your capital's culture income as being lower and your math that you broke even at 15% seems correct. With the liberty policy you would come out ahead on the additional cities, not to mention the extra science/gold.