What's the rationale for adding a Palace building with a national limit of 3?

mystikx21 said:

I gather the palaces/city halls from GEM have returned now however, and those do strike me at least as more haphazard and unnecessary, and are adding more culture, especially early. I don't get the rationale there, other than to spread out the palace yields in some way other than per-player bonuses?

Click to expand...

(mystikx21 posted this in the Culture thread, but IMO this question deserves its own thread.)

I'm not familiar with CEP, and the new Palace seems to me a rather substantial change from unmodded BNW. The mod's goal is to "set a high bar for adding new things to the game". Skimming through the old threads (at least those with a high number of posts), I wasn't able to locate discussions about the pros and cons of palaces. So: What's the rationale for adding a Palace building (separate from the Capital) with a national limit of 3?

EDIT: It seems Thalassicus introduced the idea of "City Halls" (Palaces) in this thread. The bulk of the discussion is about happiness from resources, though.

Was wondering about this myself. I thought it must be a bug seeing as the cost is lower than a monument and the yield provided is so huge they basically scream "push 2 settlers out ASAP" regardless of other considerations.

The basic concept was to distribute national happiness in a way that encouraged some early expansion, as distinguished from just sitting and growing your capital for a while. I'm not sure the palaces were the best way to do this, as it was possible instead to park sources of happiness in the various early game trees or via beliefs instead as a mechanic for this goal. The original plan was very quickly scaled down and adjusted anyway back in GEM (palaces ended up as 2 national happiness instead of 4).

There was then a system of putting in some city halls on top of that to increase city yields and avoid national yields or strong palace-centric yields (and thereby increase the value of having more cities) that was experimented with and ultimately removed as it wasn't very workable either. Apparently that's the part that made its way back in. I was less enthusiastic about that option, and it eventually went away anyway in favor of the national yields again. The lack of discussion surrounding it probably stems from that it wasn't a very long-lived change.

mystikx21: Thanks for the briefing!

I don't understand the need for further encouraging early expansion. Most of the time I'd try anyway to claim two or three good city sites, knowing that the new cities would be a liability in the short run (lower happiness, culture/research penalties), but pay off in the long run. It's also unclear to me why exactly national or palace-centric yields are a bad thing. Maybe there are compelling reasons for adding palaces/City Halls, but right now, I'm not feeling enthusiastic about this change.

There was then a system of putting in some city halls on top of that to increase city yields and avoid national yields or strong palace-centric yields (and thereby increase the value of having more cities) that was experimented with and ultimately removed as it wasn't very workable either. Apparently that's the part that made its way back in. I was less enthusiastic about that option, and it eventually went away anyway in favor of the national yields again. The lack of discussion surrounding it probably stems from that it wasn't a very long-lived change.

Click to expand...

I thought we had a pretty strong consensus that we could just put a +5 happiness policy into Liberty to encourage getting a 3rd city out early (replacing the golden age in meritocracy maybe?), the palace idea just doesn't seem necessary with that change.

That was my conclusion also, I didn't think was anything wrong with that as a policy effect rather than spreading it out. That also makes it more flexible, you can use it for extra happiness for golden ages, or you can expand more.

But I'm not sure if that's actually the reasoning now since these have more than just happiness.

I don't find myself opposed to early expansion either, with the short run costs. They're much lower than in Civ4. Where short-run costs of overexpansion could completely cripple an economy and tech rates, while short-run costs in Civ5 are basically running into happiness penalties, or infrequent golden ages until policies, luxury improvements, and buildings catch up. There really isn't much disincentive to getting a couple extra cities out to seize resources or luxuries and territory for blocking rivals.

The bigger disincentive is to get more cities out once you already have plenty of happiness for expanding, later in the game and could go wider because of available territory/resources/good city sites. And that's addressed by doing something like reducing the per city science penalty (alongside the culture cost reductions already in the liberty tree), using internal trade routes to bump a city up quickly, having available workers to quickly improve fresh terrain, and other possible modifications like raising the starting city size with a tech or policy, or giving some free cheap buildings, both of which reduce the micromanagement cost headaches of dealing with new cities in the middle or late game.

@lockstep

There are probably good reasons to avoid too many capital centric effects, mostly having to do with national wonders. It's a little more fun to be able to build diverse cities with specialized uses (eg this one gold/trade, this one production/wonders, this one units, this one science/specialists), and one way to do that is to make sure that there isn't one city that's obviously better for any of those goals every single game as the best place to put your national college and the foundry and the grand temple and so on.

There are some wonders changes from GEM that helped a lot with this, to make second, third, fourth cities very viable places for specialized use and ultimately much powerful than the capital. But that's mainly why the national yields are used instead of capital yields is to make diversification a more probable effect rather than just parking everything in one city.

Encouraging city specialization through wonder effects or, say, synergies between certain buildings and local resources, is certainly a good idea. I can also see the benefits of national vs. palace-centric yields -- e.g., allocating "free" science points nationally would somewhat thwart the tendency to make one's capital the "science city". But I really don't like the idea of bonus production/gold/science/happiness that's only collectible by founding a second and third city.

I'm also not a big fan of the palaces. It makes the game too easy. I would prefer more happiness in the Liberty tree. Liberty should be THE tree for expansion. Lots of early happiness comes from the Tradition tree with Monarchy. Plus, the Honor tree has plenty of happiness (and culture) with the policy giving +1 happiness and +2 culture per city. When I'm expanding peacefully, I often go for that policy, then pick the happiness in Liberty later on.

Liberty should have a policy for something like +5 happiness.

I don't mind the idea of palaces, but the items on the palace should be subtracted from national bonuses/capital bonuses, not just added "in addition to." I don't think the capital needs the capital bonuses AND the palace bonuses (for yields anyway). If you take the yield bonus of the capital (or lower it to 1 of each), then lower palace bonuses to 1 of each instead of 2 of each, that could work nicely to spread out the national/capital centric bonuses to early game growth without pumping tons of extra yields into the game.

It's important to have good choices, and a super capital reduces our options of where to build things. I can think of several ways to solve this problem.

• The best option is to turn starting yields into National Yields which have no source (like vanilla starting national happiness). This requires code changes to the yield library, however, which is complex, and I'm waiting on getting to that. I'll change back to this approach when I have time.
• The second option is to move yields into the basic per-city yields. I did some of this by increasing the base production and gold per city. However, this can end up rewarding wide empires too much.
• The third option is splitting the capital yields into multiple pieces, so each of the first few cities get some of the starting yields. This doesn't require code changes, and has less effect on the balance of wide vs tall empires than basic yields per-city.
• Some other method I haven't thought of?

The game normally just gives free yields to the first 1 city. The gold and science bonuses make it the best location for trade routes, production makes it best for unit and wonder construction, and happiness makes early golden ages come too easily. We have minimal choice where to focus trade, army training, and wonder construction. This discourages specializing other cities for those purposes.

Option C reduces the problem by lowering capital yields and splitting off some onto the first 3 cities. I considered giving the building to the first 3 cities for free, but there's not really a way to give free buildings in multiple cities right now, so it would require several hours of work to accomplish. I tried moving all the yields over to the 3-city building, but that made expansion too important. The current setup strikes a balance between the capital problem and expansion problem.

I'm open to alternatives if you can think of another way to solve these problems.

I agree that National Yields for Science, Culture, Happiness, and perhaps Gold are the best solution. I can also see the benefits of somewhat increasing the base production/gold of all cities.

With regard to substantially increased yields for the second and third city: While this offers me the long-term choice of not having to build everything in my capital, it also punishes/takes away my short-term choice of not founding cities no. 2 and 3 as soon as possible.

If my reading is correct that CEG will eventually switch to National Yields ("I'll change back to this approach when I have time"), I would really appreciate if the CEG change list straightened out that the current Palaces are meant as a temporary fix.

EDIT: This whole post was written BEFORE Thal posted in this thread (see above).





In case anyone missed the edit in the first post, here's the "official" explanation by Thal himself back in January.

One important thing to note: Not every starting bonus can be national - especially production needs to be tied to a city. We also need initial border expansion in the capital. Free national science might work well on the other hand.

I'm not strongly against palace buildings, but I fear that they might further encourage tall gameplay, even if they discourage super-tall capitals. But with them, we'll have 3 cities that are clearly stronger than the others, at least as long as palaces give yields with a mostly local effect.

In this thread we found out that there's currently very little reason to go wider than ~5 cities (this is a phenomenon seen in vanilla BNW already). Summary of possible reasons:

• We have a hard time improving our science with more cities due to the new per-city science penalty in BNW.
• We can't earn that much money from additional cities because there's no longer much gold coming from terrain (basically trading posts only).
• Culture is traditonally reduced for each adidtional city.
• You also can't use additional (young) cities very well for unit production either, because we need XP buildings first.

So the distribution of early yields should be seen in relation to wide-vs-tall balance IMO.




My suggestion:
Give higher per-city bonuses for every city instead of just the first three (or just the capital). Of course, not as much as the three palace building give now. We don't have to fear that wide is too strong any more, Trade routes and per-city penalties took care of that!

Of course, if we don't have free early yields and every city is just the same, we have the problem of very slow game starts and high dependence on map luck. So here's an alternative. It might sound weird initially, but it might be worth a try!



Alternative:
We get free yields for the capital initially as in vanilla, from the one and only palace. So map luck is less of an issue and our games don't have a boring, slow start. But: The palace effect is limited to the first 50 turns! After that, an inevitable event happens that cancels the bonuses.


Alternative 2:
If you think that's too frustrating and feels like a penalty:

We could get a new national wonder that unlocks approximately at the time we found our 3rd city or so. It gives some basic yields for every city, but it nullifies the bonuses from the palace! We might call it "Organized Administration" or something. It should be valuable enough that you want to get it soon. It should feel like an upgrade, even if we lose some yields in the capital.



What do you think? I'm pretty sure the idea is not yet good enough to implement, but what do you think about the general concept of capital bonuses that are time-restricted? As I said, this slightly awkward switch would only be necessary for those yields that don't work well as national income.

It's important to have good choices, and we normally don't have a good choice of where to build wonders, since the capital is usually best at everything. I can think of several ways to solve this problem.

The best option is to turn starting yields into National Yields which have no source (like vanilla starting national happiness). This requires code changes to the yield library, however, which is complex, and I'm waiting on getting to that. I'll change back to this approach when I have time.
The second option is to move yields into the basic per-city yields. I did some of this by increasing the base production and gold per city. However, this can end up rewarding wide empires too much.

Click to expand...

The identified problem (don't want the capital to always be best for wonders) is only relevant to production. It does not justify creating palace buildings that give science or gold or happiness.

Solutions A and B both seem better than Solution C - which seems to be what you're doing. I understand that A might take some time, but given that wide empires are fairly weak at the moment, B seems better if needed. Remember that new cities cause more unhappiness than in BNW, its ok if they have slightly higher yields.

While I understand that sometimes temporary fixes are needed for critical problems, it's best if these are "behind the scenes" things, like yields on dummy buildings, rather than in-your-face things like creating new buildings that aren't in the vanilla game. The solution seems worse than the problem. I'd rather just have the capital slightly better for wonders in the interim. It's not that big a problem, and it doesn't affect the early game balance of techs and buildings by creating new things to spend production on that don't need techs.

but what do you think about the general concept of capital bonuses that are time-restricted?

Click to expand...

Honestly, it feels artificial and gamey - again, worse than the problem that is being fixed. The capital will nearly always be a slightly better city because it is founded first (so it will be bigger) and is guaranteed to be in an ok position with regards to resources and luxuries.

If we use option B to distribute the bonuses across every city (instead of the first 3), the early game is less balanced, and wide empires are much more powerful. It seems like building the building causes some of the anxiety, so I'll give it to us for free in the next version.

If we use option B to distribute the bonuses across every city (instead of the first 3),

Click to expand...

Which bonuses are we talking about? It helps to be specific. +1 production makes some but not a huge amount of difference. Most other bonuses don't seem relevant for wonder production (though I guess science from the palace makes you want to build any science multiplying wonders there). +1-2 production doesn't really seem to make that much difference in where you want wonders.

I'm also a bit confused, because I thought I remembered that GEM already had changes that tinkered with gold and science as national yields (and I don't see why we'd want any free culture at all beyond the 1 on the capital's palace. Monuments and the culture from various social policies should be meaningful).
Is the problem the time it takes to integrate those into the BNW code yield library?

It seems like building the building causes some of the anxiety, so I'll give it to us for free in the next version.

Click to expand...

Thanks! Yes, it is adding new buildings that causes the anxiety. Buildings cost production, and so adding new investment decisions into the early game makes a significant difference.

Thalassicus said:

If we use option B to distribute the bonuses across every city (instead of the first 3), the early game is less balanced, and wide empires are much more powerful. It seems like building the building causes some of the anxiety, so I'll give it to us for free in the next version.

Click to expand...

Even if Palaces no. 2 and 3 are given for free, I'd prefer them not to give happiness (switch it to the capital) or culture (drop it altogether, and adjust the policy cost formula). The latter would improve early game balance by restoring the significance of monuments.

Hmm... better wait for a bit before changing this. I'm pretty sure we'll get several bug reports that city yields are weird if those buildings are made invisible
And I personally prefer seeing what happens, and I prefer choosing where to put the 3 palaces.

(Also, the more I understand how much work all of these seemingly minor changes are I feel bad for the effort that ends up not being appreciated afterwards if people don't like the change.)

Maybe we can find a more elegant solution together? People will surely pay more attention to this thread now that you actively participate in it. We might have a solution several people agree on within 2 or 3 days.





Are you sure we have to fear wide as much? We really haven't found much of a strategy for building more than five or so cities, except some Mayan ICS strategy (based on the Pyramid UB and some beliefs boosting shrines AFAIK).




What about a hybrid solution? A mixture of 3 invisible palaces and additional yields for every city?

In 3.4., the palace gives 2 prod/science/culture/gold IIRC.
What if we make the new invisible palace give half of these yields and give the other half to every city?

So this solution would give the 3 first cities exactly the same yields as what they get in 3.4. - it would only mean that all later cities get "half a palace", too.

This shouldn't be an immense boost to wide, but enough to make it slightly more attractive and to get later cities going a bit faster.
The palaces would be invisible as Ahriman likes, and people like me wouldn't mind not being able to choose the location, because they're so minimal in their lategame effect.

As far as I can tell, the identified problem is that we don't want to put too many yields on the capital, because then you want to focus on the capital too much. Fine, that's a reasonable problem, but a fairly small one, and not a high priority one.

Maybe we can find a more elegant solution together? People will surely pay more attention to this thread now that you actively participate in it. We might have a solution several people agree on within 2 or 3 days.

Click to expand...

The elegant solution is for early science and gold yields to be national yields, rather than coming from the palace. I think there is agreement on this. The only consideration is on what is needed in the interim before this is technically feasible. My main answer is nothing, just leave any yields on the capital's palace, it's not that big a deal to have them in the capital rather than national yields, and it is less disruptive to the overall economy.

In 3.4., the palace gives 2 prod/science/culture/gold IIRC.

Click to expand...

I guess I just don't see "why" at all. Why are extra science, culture, and gold needed? If you want more science, build farms, grow your pop, go for writing, and build libraries. I think it's ok if every civ has basically the same science rate in the first few dozen turns, basically coming from national income rather than from cities or pop.

Early game culture should come from monuments or social policies (or pantheon), there's no need for an extra culture building. There shouldn't be too much gold early on or it undermines production, and gold can come as a national yield and then from villages and caravans. Too much gold undermines the importance of trade routes, and of getting caravans up and protecting them.

The palaces would be invisible as Ahriman likes

Click to expand...

Making them physical buildings is only half the problem, the other problem is that I don't see why we need extra gold/culture/science at all. They're not solving the capital-centric problem, they're just overall yield boosts. Why do we want that? BNW early game feels pretty good - better than G&K did, because of better barbarian AI that mobs more, and caravans.

General thoughts on the situation

1) If they stay, they should be for free, as with GEM's happiness providing palaces. Adding a no-brainer choice to build something, and to expand, isn't a good design effect. The alternative here would be to water them down so they're more of a choice building, which would defeat the purpose of making them exclusive to small number of cities I imagine.

2) If they stay, culture and happiness should not be yields or effects they provide, at all. I'm very skeptical that science or gold should be involved either (those should be national yields, eventually). Happiness should come from policies or beliefs and later buildings, and adding culture sources is likely to be extremely disruptive. Food or production are to me the only relevant local yields that could be considered, and both are already higher per city as it is. I'm not entirely sure what the need for a "building" (dummy or otherwise) is as a result.

3) Adding gold to the raw cities (any city) is maybe fine, but I'd prefer making roads more valuable by upping city connection values if there's some problem with not enough gold being around early on. If the problem is gold to pay for new buildings, then I'd say the problem is new buildings.

4) I am extremely skeptical that the wide-tall balance would be significantly altered toward wide if there are more per city effects available. There's already extra food and production and gold now for every city than in vanilla but there are also more mechanics involved preventing wide expansion from being as profitable in the form of science and culture penalties and happiness management for much of the game. The balance is already heavy toward tall in my opinion. So long as any yields provided are modest (1-2 production isn't very heavy for example), it's not going to cause any serious damage in balance.

5) I think we can do other things to deal with later game expansion, by ameliorating the science penalty or adding higher population to later game city starts, and so on. Adding a free building effect to raise the marginal yields could be achieved by maybe adding low tier buildings per new city rather than adding an invisible building.