Why you should go for Hammers early and often

[Martin Alvito]

There seems to be some sentiment that Trading Posts are the dominant improvement. There are problems with that idea in the early stages. Lots of people have noted the inability to buy Wonders, but there's another problem you should be aware of. The exchange rate between hammers and gold is pretty terrible, except when upgrading units.

You pay roughly 4 to 5 gold per hammer on a rush buy. What that means is that +1 HPT is strictly better than +2 GPT in the early game. No early city improvements can make rush buys competitive. It takes a mid-game Social Policy, a Wonder, and city improvements to make rush buys strictly superior. If you run up the Rationalism tree, you'll get +1 Science per TP, but that still means that you will take much longer to acquire your buildings early on. That's a problem. Monuments and Libraries are extremely strong buildings that you will want everywhere ASAP, and you will want Markets to feed a gold-focused strategy.

What this means is that you want to emphasize hammers early on. The Trading Post pays dividends immediately, providing +2 GPT now. The Farm is an investment which causes you to earn +1-3 HPT a few turns later, assuming a quality city location with some potential for Hammers. Cutting turns off the time needed to grow has a persistent effect; the extra food makes you grow faster now and later. (You still have to cycle through larger food boxes with the lower-output TPs, so don't treat that as a growth penalty when comparing Farms and TPs.) The effect is cumulative in a city with many quality food tiles; working a 3F tile that permits you to work another 3F tile improves the city's growth rate.

Focusing on Food initially yields you the oh-so-critical early Hammers you need to improve your tiles while still building up enough of a military to cope with an early attack. At the start, you generally want to prioritize improvements as follows:

- Luxury tiles for resale (get the first Maritime ally rolling, possibly rush a Worker if your capital is Hammer poor)
- Add +1 Hammer/Food with Pasture to Cows, Sheep and Horses wherever possible
- Add +1/2 Food to strong Food/Hammer producers with Farms (eg: Forest Deer, Wheat or River Plains, River Hills post-CS)
- TPs
- Farm Grassland River if you have a city full of natural Hammer tiles
- Mines at the end of the city's growth cycle

You may need to work a Mine in after you get the first Maritime ally (especially in the capital), depending on Hammer production from other sources. Mines with commerce specials are always worth working, though. +3 HPT and a large GPT bonus is worth sacrificing growth rate. You should Trading Post tiles you can't irrigate with CS or profitably Mine. Exceptions: a Jungle tile usually should get a TP, and Farming a Plains tile without rivers can make sense in otherwise Hammer-poor locations. Of course, you should overwrite any Mines not on luxuries once you choose to go into unhappiness.

Hammers are effective unless (until?) you are in extreme unhappiness. If you want to rush something, you can always have the city working on the next item on your priority list using its hammers. Plan ahead!

You'll find that it will take a few more turns to get the machine rolling this way, but you will end up with a much more functional machine.

 

[milk steak]

I'm not fully convinced of this. Gold allows you to do so many other things aside from simply buying production (CS allies, purchasing tiles, upgrading units). I think it fundamentally comes down to gold having so many more options than any other resource.

An easy way to get a strong army going is to simply spam Warrior units at the start of the game. They cost nearly nothing and with a strong gold economy, you can instantly upgrade them to Swordsmen once you unlock Iron Working in a significantly more efficient manner than straight up producing them.

Growth rate is easily boosted via gold as well (through maritime CSes). And the culture from an allied CS will outpace what you can produce yourself for quite some time without the maintenance cost associated with the +culture buildings.

Also, buying units allows you to skip so many turns in actual production. Sure, it's not as efficient as hammers when compared to a hammers to gold ratio, but you're forgetting about time. Why put 7 turns of growth stoppage building a settler when you can skip all that with a purchase? Why put 20 turns into building a Frigate when you can purchase it and use it in 1 turn? Why put so many turns into a building when you can have it begin contributing instantly (for example, +5 culture from a building -- if you buy it instantly, you get (turns to build) * 5 more culture than you would have otherwise).

Also, there are many ways to acquire gold very early in the game. Selling all luxury resources at the start will allow you to buy a settler quite early.

 

[Martin Alvito]

It's a question of opportunity costs in the early going. Your supply of gold is finite. You would prefer to buy food with the initial pool if at all possible. 500G -> 6 Food per turn by the time you can buy an ally, assuming you went Worker -> Settler or Scout -> Worker -> Settler. Given decent tiles to work, that's going to return substantial Hammers and Gold that you otherwise wouldn't have.

Rush buying a Settler out of the gate costs nearly as much as buying a maritime ally. That's not a viable play, because you get more Food from the ally than you give up wasting turns building the Settler. While it potentially gets the new city on the ground faster, both cities grow far more slowly than if you bought the Food. The Maritime move eventually laps the Settler purchase. It's also worth noting that you only require 250G to maintain the ally at the next investment, instead of 500G to acquire it. So you end up ahead on gold and time.

If you're talking about rush buying a building for +5 Culture per turn, then we're not talking about the same time period here. What I'm saying is that you want Hammers right out of the gate, because Gold in the early game cannot compete with the efficiency of Hammers. You're going to overwrite to Trading Posts once you're teched up, prepared to puppet a victim, and ready to start pushing the Freedom tree.

For most civs, upgrading spammed Warriors to Swords is the second priority, and it's usually best financed with luxury sales. The two civs with strong spear UUs (Persia and Greece) want to skip that step and make UUs instead with Hammers.

 

[milk steak]

Hm, I think our strategies are fundamentally different. I beeline Theology to unlock Patronage (via GL slingshot) and steal a Worker from a CS (the negative diplo hit is instantly removed with Aesthetics). I tend to favor purchasing a Settler over gifting a Maritime CS since a 2nd city will greatly improve gold output over simply having additional growth in a single city. Also, I generally don't have Philanthropy (+25% influence from gold) unlocked with Aesthetics just yet. Settler seems to be a more efficient use of gold at that point.

I tend to build Warriors in my 2nd city to avoid early DoW and to be able to upgrade them later to Swordsman if I need.

Edit: And by spamming Warriors, you can instantly upgrade them to Iron Working once it unlocks. It's easy to churn out Warriors and you'll have a huge Swordsman army very quickly. No Iron? Ally with a City State that does have it!

 

[Shillen]

The trading post logic is thus:

Trading post gets you enough gold to ally maritime city states which give you enough food to work the hammer tiles.

I'm not sure I entirely agree with it but that's the logic they're using.

 

[Martin Alvito]

Hm, I think our strategies are fundamentally different. I beeline Theology to unlock Patronage (via GL slingshot) and steal a Worker from a CS (the negative diplo hit is instantly removed with Aesthetics)...

I'm playing Deity on small land maps, where GL slingshot is just a bad idea. I've tried to make it work, but it's just too slow. GL slingshot feels much like CS Slingshot on Immortal did in Civ 4. You can do it, and you will get CS from a good start, but the opportunity cost is enormous. You need a military if you don't want to eat an early DOW when you're defenseless, and you won't have one running GL slingshot.

Staying in Classical for a while makes the initial CS buy inefficient as compared to GL slingshot, but there's no help for it. You feed the beast 250 more gold when the time comes and like it.

Ganking a Worker from a CS is great whenever you can swing it. Beats choprushing the second Worker in city 2.

The one thing I will say for small Deity maps is that you learn to play the early game efficiently. Mistakes are consistently punished by death. Good early play consistently yields a winning game, since the AI is bad at combat.

Trading post gets you enough gold to ally maritime city states which give you enough food to work the hammer tiles.

I understand that. The argument I'm making is that it's strictly more efficient to ally first and work the hammer tiles. Every time you +1 HPT, you're gaining long-term ground on +2GPT. If you sell luxuries aggressively enough, you can keep the ally chain going while developing Hammers rather than Gold.

The long and the short of it is: Hammers make stuff you don't have to rush. Getting enough Hammers fast enough permits building enough stuff to stay alive while you pump money from luxury resales into city-states. After that, you end up building more stuff than the guy who went for +GPT.

 

[milk steak]

Yea, I got screwed by an early DoW once while building the GL. So to combat that, I now buy the Settler ASAP and have the 2nd city produce a lot of Warriors to up my military rating (and to upgrade them to Swordsman/Longswordsman later if I'm going for Conquest).

 

[King Jason]

I think this might wind up being the CE vs SE debate because I think gold becomes better in the long term. Not only does your income ramp up, but with Big Ben or Mercentalism (or both), you can buy things in the same time it'd take to build them.

Personally, I just do both. Hammers are good early, but I certainly set myself up a beast of a gold city that'll drag in tons of GPT.

The true beauty of GPT apart from it's varied uses, is that it's "production" you can put anywhere. I have had massive jungle-only science cities with not a production plot workable, the gold comes in handy to buy buildings here.

Also, I like to share borders between my cities a lot. It's rare you're gonna work every tile in a radius, so for example, I had another jungle science city who's immediate tile range were all jungles. It's northern border of 2 tiles length were hills, but this was the immediate border of the other city. Which was my highest production city usually spamming wonders or units.

On the turn my science city needs to build something ~ I swap the tiles; the northern city is working the jungles, the southern city takes the hills. There's very little loss of production and science in total, and the city get's to build what it needs in a timely manner.


So really, there's all sorts of things you can do.

 

[Seanner]

You left out the granary! It's the best building there is! It effectively converts into a free mine of some kind (3H, possibly with 5G) or 3S for 1G. The water mill is second best with a 2G cost. If we are going with the logic that growth is everything (certainly a 36 size city can do more than a 1 size city), then buildings that improve growth better be at the top of the list.

Btw:
Later the bestest best of all the best buildings has to be the seaport in one of those nifty 3 fish cities that come up from time to time. That building converts to 6H, and you'll always be working those tiles anyway! That and the ability to use a harbor instead of 15 hexes of million worker turn road building makes coast cities generally superior. Technically, their 2050 AD capability is less due to more useless tiles, but most cities will never see size 36.

So I'd like to add to your thesis that one should build granaries in cities, and they should make these cities near fish.

 

[Martin Alvito]

Gold absolutely becomes better in the long term. This is indisputable. However, I find that I usually capture a large source of labor around the time this becomes an issue. Usually I just convert them into coins, but at the right stage of the game I use them to overwrite tiles with TPs. This gives back some of the ground that I gained earlier with a Hammer focus through increased maintenance costs, but the earlier time advantage is always well worth it.

The price of the flexibility of Gold is its inefficiency. Hammers are strictly better at making things early on, and you can almost always substitute things Hammers can make for Gold. The main exception is when you want a tile faster than your culture can assimilate it, or when the AI is idiotic about determining which tile to add. You know, when there's that tasty luxury on your extended borders, and the AI insists on adding a lousy Grassland tile instead.

Tile sharing is classic Civ gameplay. One pile of Mines in the right place can fuel two cities with other specializations.

Five Granaries is always dominated by buying another Maritime ally. The ally gives you a resource, which always either yields +5 (or more) Happiness or frees a resource for resale. Granaries are too expensive for what they do, and are an afterthought in games with a decent population of Maritime city-states.

Growth is not everything, or the only thing. It's useful up to the point at which you can no longer sustain it without incurring greater costs than the growth provides. The optimization problem we want to solve is how to most quickly get the most efficient empire up and running. Until the Happiness cap problem is resolved, we know what to do once the empire is functional. Pop a civ, annex, and rush buy everything except Big Ben.

EDIT: To reiterate/clarify, the argument is that Hammers are simply more efficient early on. Gold is flexible but inefficient, except when buying city-state allegiance. The implication is that the player that focuses on early Hammer tile improvements, and acquiring more improvable tiles through Food improvements, will outproduce the player that focuses on early Gold tile improvements and grind that player into dust. A limit exists; that limit is determined by the point when the city can no longer acquire further Hammers. At that point you want to start putting up Trading Posts.

One exception exists: you can almost always get enough gold to fuel your allegiance machine through reselling luxuries to AIs. If for some reason you cannot, you had better start on the TPs earlier. Multiple Maritime allies are crucial for any strategy.

 

[fugazi]

If you know what you want to build and get to utilise it, hammers are great. If you don't have any specific needs, gold is better as it can be used for everything - buying units, buildings, extra tiles, friendly/allied status with city-states, resources from other civs, unit upgrades, etc.

I do share part of your opinion though - hammers are severly underrated by most. Every city I build gets at least one mine. That way they can at least produce a little bit, saving my gold for more useful purposes than buying every little building I want.

 

[Seanner]

I don't believe you are giving granaries enough credit. They provide 1G -> 2F conversion, effectively giving a mine or a specialist -- say, 3H or 3S.

So how is a 1G -> 2F -> 3H conversion too expensive? Yes alliances are better, but that doesn't make granaries not worth it after you get alliances. Unless you are saying that the 100H is too high, which doesn't seem obvious. At say, 11H a turn (3 mines + city), done in 9 (add one H from a plain or something last turn), and costs you 55F (9 turns * 2F per mine lost * 3 mines + 1 from last turn plain) worth of growth. It recovers the food in 55/2 = 27.5 turns. Of course, 27.5G of maintenance was lost meanwhile, and more will be lost yet, but the 1G -> 2F -> 3H or 3S was the whole point of building it anyway. After you lose 27.5G, you are buying 3H for 1G (or 3S), an excellent conversion in either case (civil service only costs 133G at that science rate!).

The only other possible argument against granaries then, in this hypothetical situation, is that those 9 turns could've built some other building, like 4/5 of a market. At the smallish size city the granary would've been built at, it's questionable whether a market would provide much benefit anyway. If built on a river, and 4 farms were being worked adjacent, and perhaps 2TP somewhere, the city might make 9G (4 + 2*2 + 1). The market will yield 9 * 1.25 = truncated to 11G = 2G increase. Now the numbers obviously go up when you work minted gold mines, at the expensive of growth, so the market should come later on in the city's life when it's grown..which it will do faster when it has a granary.

Another other critical resource perhaps then..library vs granary:

80H / (11H/turn) = ~7 turns, with output of, at size 6 again, 3S (+1 for every 2 citizens..1 * 6/2 = 3), and also costs 1G. The same conversion rate as a granary therefore, but again scaling better (and cheaper to build). The thing is, just like the market, its scaling is directly dependent on city size -- improved by granary. And a guy working a specialist slot in it can be fed by a granary. Food can do more than just convert into science..it can convert into whatever tile or specialist you put it on, or it make more dudes to convert.

These calculations are somewhat arbitrary, but obviously I cannot do every possible permutation. Intuitively though, I can't see how a granary is a bad investment.

Is anything so worth building right NOW, besides perhaps a monument, that it's actually a mistake to build a granary (lighthouse doesn't count obviously..similar but cheaper effect and you either have more than 1 fish or not)?

 

[fermionorama]

"after you get alliances" ... meaning you're already allied with every maritime city state in the game? That doesn't happen very often for me, but I play larger maps, and tend to increase the # of city states in a game.

 

[Martin Alvito]

So how is a 1G -> 2F -> 3H conversion too expensive? Yes alliances are better, but that doesn't make granaries not worth it after you get alliances.

Libraries are a higher priority because you need to make Great Scientists early and often to brute force your way up the bottom of the tech tree to keep the conquest machine rolling. Monuments are a higher priority because you need border pops and Culture for policies. Units are a higher priority because they convert to luxuries and pop.

I say that Granaries are an afterthought because I never find reason to build them in the first hundred turns. Other builds take priority. Once you have a functional army and the basic improvements, and you're looking for additional builds, it's time to think about Granaries. But not until then.

Yes, if you play on huge maps, Granaries should be deferred even further. The great thing about a high concentration of city-states is that you get to cherry-pick your luxuries when selecting your allies.

 

[Evalis]

I think everyone playing this has gotten so used to exploiting the AI that it's standard strategy. What if you are playing human opponents though? No one in their right mind will give you 200 or 300 gold for a luxury resource, let alone at the beginning of the game when they don't need it.

And how about fighting over maritime city states! You'll probably plow far far more than just 500gold into those guys trying to keep them allied to you. The computer is horribly horribly stupid in combat. The only enemy that posses a threat is nobunaga since his 1 hp units can still inflict serious damage, which the ai has a habit of continuing to march at your borders, so I don't really think it's fair to run comparisons of tile yeilds based on exploitations you can do to the computer.

Personally I'm of the mindset that production and food tiles are indispensible in the early game if you don't have an ai around to cheese.

 

[strollen]

Libraries are a higher priority because you need to make Great Scientists early and often to brute force your way up the bottom of the tech tree to keep the conquest machine rolling. Monuments are a higher priority because you need border pops and Culture for policies. Units are a higher priority because they convert to luxuries and pop.

I say that Granaries are an afterthought because I never find reason to build them in the first hundred turns. Other builds take priority. Once you have a functional army and the basic improvements, and you're looking for additional builds, it's time to think about Granaries. But not until then.
.

I agree both Libraries, and monuments are higher priority as are settler if you have a good city location. There is nothing wrong with granaries per se it it is just an opportunity cost argument. On the other hand if you have hammer rich city, which seem to be rare, and not enough good food sources than building a granary makes sense.

All that said the Aztec hanging gardens is well worth building a size 8 or 9 city gets +5 from it and my capital was getting 6+ food by 1500 or so.

 

[Martin Alvito]

I think everyone playing this has gotten so used to exploiting the AI that it's standard strategy. What if you are playing human opponents though? No one in their right mind will give you 200 or 300 gold for a luxury resource, let alone at the beginning of the game when they don't need it.

I can't see the ability to use the AIs as cash cows sticking around for very long. All you have to do to fix it is recalculate the offer on the basis of the AI's present happy state. On Deity it always has huge happiness surpluses early. It should pay for a thirty turn resource contract accordingly.

If for some reason city-states weren't available, that emphasizes Food above and beyond all other concerns. Early on, pop is life. You'd substitute your gold into rush buying non-combat units, and Hammers would indeed take on even greater importance.

In a multiplayer game, the ally game becomes tactical. You want to acquire resources and deny the adversary his wherever possible. In a 1v1 hypothetical, the stronger commerce civ generally should pick fights over allegiance, expecting to win them. The weaker commerce civ therefore needs to wait for the stronger civ to make its move in the ally game.

As you add more players, the calculations get more complicated.

 

[Bibor]

Of course hammers are the most important. Food and gold can come from many sources. You do want that catapult out in 10 turns, not in 27, right?

 

[SammyKhalifa]

Sorry to go on a slight tangent here, and noone will like this really, but one overarching theme I've seen here is that maritime City-States really should be nerfed. They're central to every single strategy--something in that tells me their bonus is too powerful.

 

[tflst5]

this strategy fits very well with the Russia civ (extra hammers for horse+iron). I did something similar to this in my last game. Having a massive amount of hammers allowed me to do so much more early on than many of the gold starts i have tried recently. Instead of allying and sinking money into maritime city states, i just built granaries and mills. Instead of buying culture city states i build monuments. The maintenance they cost me was far less than the maintenance i was paying to keep my city states happy in my other play throughs.

The hammers let me build more units where i would normally be buying them. The extra horse and iron from the Russia ability gave me lots of surplus to trade. And in the end my gold stockpile was actually higher - since i had fewer things that necessitated purchase. I could use it to fill the gaps.