Specialists

[Thalassicus]

I think of economies as three possibilities:

Village Economy

• Construction is global (gold from one city can build stuff in another city)
• Village priority.
• Commerce tree filled.
• Specialists contribute to 0-2 cities.

Production Economy

• Construction is local.
• Mine/mill priority.
• Flexible policy strategy.
• Specialists contribute to 0-2 cities.

Specialist Economy

• Construction can be global or local.
• Farm priority.
• Specialist policies from each tree.
• Specialists contribute in some way to every city.

I'd like all three of these choices to be about equally interesting and useful. Which one we decide to pursue at the start of the game depends on our playstyle (tall/wide/conquest) and circumstances of the map terrain. I feel a gold economy is best for coastal or conquest games, for example. I'm not sure about the other two economy types since I don't play peaceful games often.

Specialist economies are usually tall empires, so food is often the most important yield. After Civil Service priorities are probably something like...

Yield Categories

1. High-food resources
2. 41 River grassland
3. 311 River plains
4. 2 Specialists contributing to great people
5. 31 Coast
6. 3 Dry grassland
7. 0 Specialists who will never make a great person (lone specialists creating a trickle of gpp)
8. 21 Dry plains | 22 Jungle
9. Forest, hills, river tundra
10. Dry tundra
11. Desert, snow, mountains

The order can vary depending on things like:

• The value of surplus food drops to near-zero when if reach our happiness cap.
• There's a limited number of river tiles, so unless we have a huge abundance of rivers we run out of those quickly.

In a non-specialist economy, #6 drops down two spots (specialists are useless in most cities).


The choice of three economies to pursue provides an additional layer of strategic depth to the game. Along these lines I've been gradually working at making specialists more useful in VEM:

• Specialist slots are available in the early game.
• National wonders have additional specialist slots.
• Several policies improve specialist yields to help them remain competitive with tiles (Commerce and Freedom trees).

Great People are also improved. Their tile improvements are better, and boost from techs. Great people are also more available than vanilla. These are some creation rates for Great People:

Market and National Treasury c5gold: Great Merchants):
3 - Vanilla
6 - VEM

Watermill and Heroic Epic c5production: Great Engineers):
0 - Vanilla
6 - VEM

Library and National College c5science: Great Scientists):
0 - Vanilla
6 - VEM

Temple and National Epic c5culture: Great Artists):
3 - Vanilla
8 - VEM

 

[Vanchelon]

Ye, I think it's a good change but.. Damnit now yer gonna force me to start that game with Ottomans that I've been putting off

 

[Txurce]

I liked the dynamic in Civ 4 where two types of economies were possible:

• Specialists
• Cottages

It provides an additional layer of strategic depth to the game. Along these lines I've been gradually working at making specialists more useful in TBM:

• Specialists provide higher yield.
• Specialist slots are available in the early game.
• National wonders have additional specialist slots.
• Several policies improve specialist yields to help them remain competitive with villages.
• Great person tile improvements are better, and improve with techs.
• Great person rate is mostly unchanged. I dropped the per specialist from 3 to 2, to compensate for the fact there's more specialist slots available.

I've been hooked for a long time on the flexibility that a specialist economy provides, especially since I tend to stay in the 4-6 city range when not warmongering.

I'm more curious as to what the advantages of a cottage economy are - in other words, why I would choose not to take the specialist route.

 

[Thalassicus]

A specialist economy (specomy? specon?) has several limitations:

• River tiles generally outpace specialists. Most early cities are near rivers.
• Golden Ages increase +1 of improvements, but not specialists.
• The earliest engineer/merchant buildings are situational.
• Nothing improves specialist until late game.
• Specons usually require more happiness.
• Ideal with National Wonders, which are best with a tall empire that goes Tradition.

Because of these reasons, a gold economy is better for conquest games (with the exception of Suleiman). Specons are a long term strategy ideal for tall empires that get National Wonders. A conquest game involves building lots of military to kill off at least one or two opponents by the Medieval era. It's difficult to combine those two.

Gandhi is also fantastic for a specialist game. High population, high food, and a good early UU to fend off attackers while getting things going. He's the more peaceful of the two, while Suleiman is better for a conquest specialist economy.

 

[Vanchelon]

Ye, I'd go Ghandi if I was going for a more continent style game, but atm I'm more looking towards starting an archipelago game as Ottomans. (on a modified huge setting, 156x86 ish with lower sea levels, also helps the AI perform better when theres ample land around on "ocean" maps).

Archipelago maps usually bring less rivers (ghandi less powerful), and coast tiles can work well with a "jack of all trades" specialist economy. You get decent food (atleast 3 per tile) and a basic gold yield which can take advantage of longer (and more) golden ages aswell.
It'll probably take a bit longer to kick off then Ghandi with lots of rivers, but should work out

I guess we'll see how well it works out

 

[Thalassicus]

I haven't played a serious Archipelago game... tell me how the AI starting units work out.

I give them a few triremes instead of warriors if an AI starts near lots of water. It should help them explore and defend better in the early game, in theory, but I haven't tried it much in practice.

 

[Txurce]

Because of these things, I believe a non-specon is better for most conquest games (with the exception of Suleiman). Specons are a long term strategy ideal for tall empires that get National Wonders, while a conquest game involves building lots of military to kill off at least one or two opponents by the Medieval era. It's challenging to combine those two.

Yes, I keep trying to thread that needle, and find myself generally being sidetracked and getting conquest off to too slow a start. Germany seems to offer the best opportunity to pull it off, so I keep trying some variation of this with them.

 

[Ahriman]

I think you are going in completely the wrong direction with the 8.6.16 changes.

It is very boring if specialists provide direct income that is directly competitive with yields from tiles. Among other things, it ends up making terrain unimportant, because you can always get whatever kinds of yields you need from particular specialists (eg: hammers in a grassland area from engineer specialists) and because they mean that the only real resource you need is food (since food lets you get specialists which means you get whatever you want).

IMO, around half the value from specialists should be from great people, not from their direct yields. Specialists should be useful only in, well, specialized or concentrated circumstances. It should be worth using specialists to get great people in a few high food concentrated cities that can support them, or using them to tweak a few specialist cities even further (eg: science specialists in a city with national college). It should not be worth using specialists in small cities purely for their yield effects without getting great people; you should nearly always be better off working tiles in these cities. The more you boost the universal value of specialists by increasing their yields, you start making the Freedom policies like half food and half unhappiness into things that are far too powerful.

If you think specialists are underpowered (and I don't think that is the case), then boost the value of great people further, don't just boost the direct yields from specialists.

The 8.6.16 changes also totally break Suleiman.

Basically, I thought specialists were pretty decent, I would revert the changes.

 

[Txurce]

It is very boring if specialists provide direct income that is directly competitive with yields from tiles. Among other things, it ends up making terrain unimportant, because you can always get whatever kinds of yields you need from particular specialists (eg: hammers in a grassland area from engineer specialists) and because they mean that the only real resource you need is food (since food lets you get specialists which means you get whatever you want).

IMO, around half the value from specialists should be from great people, not from their direct yields.

I agree with some of this in theory, but not necessarily in practice. At the end of the day, we need everything that GP's provide. For example, you're effectively saying that if I want more hammers, I could build a GE and plant a Manufactory - more gold, a Customs House. I could do that on grassland, and make terrain as unimportant as you fear it would be if Engineer slots gave hammers or Merchant slots gave gold.

To me this doesn't mean that your proposed approach is wrong - I just don't see how it leaves me in a meaningfully different place.

 

[Ahriman]

At the end of the day, we need everything that GP's provide. For example, you're effectively saying that if I want more hammers, I could build a GE and plant a Manufactory - more gold, a Customs House. I could do that on grassland, and make terrain as unimportant as you fear it would be if Engineer slots gave hammers or Merchant slots gave gold.

This is not the case, because to get a Manufactory or Customs House you have to focus on a particular specialist type in a particular city, and that that city will probably need a good food supply in order to maintain lots of specialists.

There is a huge difference between the degree to which terrain matters between the case where specialists themselves give good yields and the case where you only get the terrain-nullifying yields for a precious few great people.
There is a big difference between being able to get good production in every city because I have an engineer in each one (and one engineer per city is never going to produce a significant number of great people) and between a situation where I concentrate engineers in a single city, which gives me a great engineer, that I can then use to get a manufactory in a single city. The latter case is much more about specialization, the former case is more about getting generalized good yields everywhere.

Also don't forget that the super-improvements aren't the only ability that great people have.

Great people are interesting and different from yields from working tiles; they're an investment now for a payoff later, they have multiple abilities which you have to choose between, they can have empire-wide effects, they can give their benefit to a different city than the one that produced them.
Specialist yields aren't really that different from tile yields.

 

[Txurce]

1. This is not the case, because to get a Manufactory or Customs House you have to focus on a particular specialist type in a particular city, and that that city will probably need a good food supply in order to maintain lots of specialists.

There is a huge difference between the degree to which terrain matters between the case where specialists themselves give good yields and the case where you only get the terrain-nullifying yields for a precious few great people. There is a big difference between being able to get good production in every city because I have an engineer in each one (and one engineer per city is never going to produce a significant number of great people) and between a situation where I concentrate engineers in a single city, which gives me a great engineer, that I can then use to get a manufactory in a single city. The latter case is much more about specialization, the former case is more about getting generalized good yields everywhere.

2. Also don't forget that the super-improvements aren't the only ability that great people have.

Great people are interesting and different from yields from working tiles; they're an investment now for a payoff later, they have multiple abilities which you have to choose between, they can have empire-wide effects, they can give their benefit to a different city than the one that produced them.
Specialist yields aren't really that different from tile yields.

1. I see your point, but am framing my perspective from Thal's goal of making specialist economies viable. So when I talk about a Manufactory vs hammer yields, I'm presuming both approaches being buffed probably beyond the level you prefer.

2. I agree that GP are more interesting than yields. This applies to GS as much as anything else - I wish bulbing them were more balanced, because it's more fun. But if we're pursuing a viable specialist economy, then yields seem much easier to balance.

 

[Ahriman]

1. I see your point, but am framing my perspective from Thal's goal of making specialist economies viable.

My point is that the goal of making a pure specialist economy viable is not a good one.

Specialists should not have direct yields competitive with yields from working tiles; their yields should be similar only once you include their great people points. Ignoring tech and social policy modifications, you should always be worse off using a specialist than you could get from a tile if that specialist never contributes towards getting a great person; you can get 3 hammers from a mined hill, you should get 2 hammers from an engineer. The engineer becomes valuable because you can use it in a place even without the hill, and because it also contributes GPPs.

Should using specialists in some key cities be at least competitive with using no specialists at all, once you take the great people benefits into account? Absolutely. But that is very different from trying to make specialists alone give yields equal to those from tiles. They didn't in Civ4, I think we should keep it that way.

This way, they are interestingly different. Working tiles is better if you have good terrain, and if you want to concentrate on yields today. Specialists are weaker in the short term, but arguably stronger in the long term when they are used to generate great people. They are a form of investment.

 

[Txurce]

My point is that the goal of making a pure specialist economy viable is not a good one.

Specialists should not have direct yields competitive with yields from working tiles; their yields should be similar only once you include their great people points. Ignoring tech and social policy modifications, you should always be worse off using a specialist than you could get from a tile if that specialist never contributes towards getting a great person; you can get 3 hammers from a mined hill, you should get 2 hammers from an engineer. The engineer becomes valuable because you can use it in a place even without the hill, and because it also contributes GPPs.

Should using specialists in some key cities be at least competitive with using no specialists at all, once you take the great people benefits into account? Absolutely. But that is very different from trying to make specialists alone give yields equal to those from tiles. They didn't in Civ4, I think we should keep it that way.

This way, they are interestingly different. Working tiles is better if you have good terrain, and if you want to concentrate on yields today. Specialists are weaker in the short term, but arguably stronger in the long term when they are used to generate great people. They are a form of investment.

I tend to agree with you conceptually. But being specific, your point seems focused on GE's and hammers. Do you also see a problem with the other GP yields, vs taking a different approach to achieve similar gains?

With regard to Engineer slots, if there were so many that they matched strong geographical locations, I would totally agree. But are there enough early-game Engineer slots to unbalance a given city site? If I built a Water Mill, Smithy and Forge, for example, I would probably do so on a site that has lots of hills - and would definitely mine them as well. There may be a problem with too many hammers in total as a result - but I can't imagine not building the mines in this case (or building multiple hammer buildings in a grassland-heavy location).

 

[Ahriman]

Do you also see a problem with the other GP yields, vs taking a different approach to achieve similar gains?

The effect is most clear with hammers, because you have tiles that yield only hammers and a hammers-only giving improvement (mine) so the comparison is easiest to see, and because nowhere-to-get-hammers is the most obvious way in which terrain might be weak.

But I think the same kind of thing applies to other specialists; merchants vs trading posts, for example.

but are there enough early-game Engineer slots to unbalance a given city site?

If it is a no-brainer to work the engineer slots instead of a mine then we have a problem. Yes, we definitely can get enough slots to do this. I'm also not sure why you are focusing only on early game.

If I built a Water Mill, Smithy and Forge, for example, I would probably do so on a site that has lots of hills

Why? A water mill is arguably more useful in a city where hammers are hard to come by, and a smithy or workshop give little synergy with hammers, so both of these are useful in a city that isn't focusing on hammer production and will never build a military unit (but would like enough hammers to construct buildings locally). A watermill is great in a specialist city where you want lots of food for the specialists, where you'll build gardens, and where you might want hammers for building culture and science boosting buildings.

But it seems that we agree on the general point; it is a mistake to increase specialist yields/decrease their GPPs, and thus make them more direct substitutes for working tiles.

 

[Txurce]

Why? A water mill is arguably more useful in a city where hammers are hard to come by, and a smithy or workshop give little synergy with hammers, so both of these are useful in a city that isn't focusing on hammer production and will never build a military unit (but would like enough hammers to construct buildings locally). A watermill is great in a specialist city where you want lots of food for the specialists, where you'll build gardens, and where you might want hammers for building culture and science boosting buildings.

But it seems that we agree on the general point; it is a mistake to increase specialist yields/decrease their GPPs, and thus make them more direct substitutes for working tiles.

I said Water Mill, Smithy and Forge (not one or two) all in one city as an example of a city focused on production. You can substitute (or add) a Workshop or Factory or Foundry. I can't imagine not mining almost every hill I have, and probably wouldn't build a more expensive production building in a city that would take a long time to build it. That's why I don't see this as a game-warping issue. But to be fair, I do see how - late in the game - I may have to choose between working a slot or a tile in a given city, and then choosing the slot. This works against the conceptual separation which I would ideally also prefer. But I do find the idea of a specialist economy an interesting option, and that seems to be achievable only by having the slots give more substantial yields. I'd prefer to test for the unwelcome results that you predict (essentially ignoring terrain to focus solely on food).

 

[Ahriman]

I said Water Mill, Smithy and Forge (not one or two) all in one city as an example of a city focused on production.

So you're not bothered by the possibility that all your non-production cities (with only 2 of these structures say) will still be unambiguously better off with engineers than working non-river mines?
I do see that as a game-warping issue.

I'd prefer to test for the unwelcome results that you predict

The unwelcome result I predict is for example the fact that in every city with an engineer specialist slot, I would rather use an engineer yielding 3 hammers and 2 GPPs than I would work a mined hill for just 3 hammers.

Does this really require playtesting to observe?

Similarly, if a merchant gives 4 gold and 2 gpps, then aren't I clearly better off using this than working a non-river trading post?

 

[Txurce]

So you're not bothered by the possibility that all your non-production cities (with only 2 of these structures say) will still be unambiguously better off with engineers than working non-river mines?
I do see that as a game-warping issue.

The unwelcome result I predict is for example the fact that in every city with an engineer specialist slot, I would rather use an engineer yielding 3 hammers and 2 GPPs than I would work a mined hill for just 3 hammers.

Does this really require playtesting to observe?

Similarly, if a merchant gives 4 gold and 2 gpps, then aren't I clearly better off using this than working a non-river trading post?

I'd agree if those were the circumstances from the start. But in a game I'm going to build plenty of mines (and theoretically Villages) before I have more than one slot. Then I'll reach a stage where I have tiles and slots, but can fill both. Eventually I will reach a point where I have to make a choice between them, and you're right that I will then choose slots over tiles. But I can't be sure that this will happen early or often enough to warp the game. (By warp I mean lead me to convert most of my mines to farms, etc).

 

[Thalassicus]

@Ahriman
I think I understand where the confusion is - there were two changes in this patch which might seem related, but are independent of one another.


1) Policy reliance

Here's the potential yield of a Merchant between .15 and .16:

.15 Merchant
521

.16 Merchant
511

(Not counting world wonders, which aren't always accessible.)

I did not increase the total potential yield of specialists. I reduced it and made it less reliant on policies. In the defensive buildings conversation you said things should be useful on their own and not require a specific policy - that's what I did here.


2) Great person availability

There are more specialist slots available in TBC than vanilla, which increases the availability of great people. I reduced this great person availability back closer to vanilla levels.


If we revert to .15, it would increase policy dependence (something you said you don't want) and increase great person abundance.

If you're concerned about engineers specifically, I could reduce their particular yield. Keep in mind the only engineer slot available for some time requires a river. River mine yields are better than engineers:

3 - Engineer
31 - River Mine
41 - River Mine with Engineering

There's also the disadvantages I listed earlier about a specon empire. Blast Furnaces and Smithies are generally on the wrong side of the tech tree for this until the late Medieval era, since we can survive with Oligarchy and archers while going for the top half of the tree.

 

[Ahriman]

Here's the potential yield of a Merchant between .15 and .16:

The total potential yield with all possible boosting policies is not really the relevant feature. The relevant feature is the base yield of the specialist, which you increased. Most of the time, you won't have all possible specialist-boosting policies *or* all possible trading-post boosting policies.

In the defensive buildings conversation you said things should be useful on their own and not require a specific policy - that's what I did here

No you didn't. You made specialists superior to tile yields even without any policies.
They *were* useful in previous versions, because they produced great people, which are valuable. Now, they are too strong, and some policies are arguably too weak.

In contrast, walls weren't useful.

If we revert to .15, it would increase policy dependence (something you said you don't want)

I think policies should support different playstyles; I think it is ok for specialist-boosting policies to boost that kind of playstyle.
In contrast, the wall-boosting policies don't really boost a playstyle; "build walls" isn't really a playstyle, there isn't really any synergy other than building the structures.

There are more specialist slots available in TBC than vanilla, which increases the availability of great people. I reduced this great person availability back closer to vanilla levels.

I think it is unfortunate that specialist slots are so widely available, for exactly the same reason as I find it unfortunate that strategic resources are so widely available in vanilla. Once a constraint is relaxed too much, it is no longer meaningful.
I would cut basically everything that gives 2 specialist slots down to 1, and I'd consider removing a few other slots.

If you're concerned about engineers specifically, I could reduce their particular yield.

I would revert the yields of all the specialists, so that they remain inferior to working tiles (and not just river tiles).
As I said before, much of the benefit of specialists should be from great people; a GPP should be worth somewhere between 1/3 to 1/2 of an actual yield.
I would sooner increase the GPP income from specialists to 4 (or reduce GPP costs for great people) if you thought specialists were underpowered (which I didn't think they were).

What you have done is the opposite; you have made specialists less about great people, and more about their direct yields, which I think is unfortunate, and you have made them too valuable overall.

 

[Thalassicus]

I think we're just going for different goals.

I liked the dichotomy in Civ 4 where there were two types of viable economies. I know there's a desire to branch out into new things, but in this case it's something I feel Civ 4 got right. I like the flexibility and improved strategic options it provides.