Are Markets, Grocers and Banks worth it?

Windsor

Flawless
Joined
Feb 14, 2008
Messages
1,387
Location
Norway
So I got challenged by Fippy to get arguments against the three gold-buildings here. I argue that they're much closer to viability than he does.

So here it goes. Would you build Markets, Grocers and Banks if these changes where made:

- Building wealth used the vanilla formula
- No failgold from wonders, missionaries etc
- No overflow gold (current BTS patch, but changed compared to BUG)
- AI had the same willingness to pay cash as you have (you can't exploit the deity-AIs huge cash reserves).

From my point of view there's no doubt that these changes would make the MGB-buildings more than viable rather than the "maybe in the capital" fringe they're now.

So if anyone wants to make the argument that the MGB-buildings still would be too expensive I'd love to hear them.
 
hhhhh
 
they're still pretty bad... 150h for +25%, and even without failgold or currency money builds, you'll probably be able to keep your science slider at at least 70%... so 25% of 30% for 150h... a library is going to be a better build most of the time than a market, this doesn't change this. Although certainly you'd build them more often, but maybe you'd build 1 or 2 in a 20 city empire rather than the 0 or 1 you currently build.

even with this change, you still need a secondary reason to build a market or grocer... such as a use for the health or happiness bonus, or a desire to run merchant specialists, or building it on a shrine city. Banks could go in super commerce cities, but unless its producing 40+ commerce a turn, it's not a better investment than, say, a courthouse.
 
I know I've played vanilla at some point in my lifetime.... (searching)

We're talking 2:hammers: = 1:gold: and applying :gold: modifiers but not :hammers: modifiers? (i.e. market helps wealth, forge does not)
 
Premise:

From commerce, your city is producing G base gold per turn. (i.e. G = commerce * slider)
Your city will produce H hammers per turn over some time frame.
We use a time machine so that we get the effect of the market immediately if we build it.
We consider T turns.

If we build the market, the gold we get is

1.25 (GT + (HT - 150)/2) = (5/4) GT + (5/8) HT - 375/4​

If we don't build the market, the gold we get is

GT + HT / 2​

The market is better if

T > 750 / (2G + H)​

If we want to do better than break even at 30 turns, then we need

2G + H > 25​

e.g. a city with 15:hammers: per turn and 10:commerce: per turn will just break even after 30 turns if the slider is at 50%.

(and, remember, this was assuming that we got the market instantly, rather than spend the first 10 turns earning gold without the gold multiplier)

---

If the market didn't affect building wealth, then the comparison is just GT > 300, and you need G > 10 to better than break even at 30 turns.

For comparison, the BTS formula (with no forge/etc) leads to GT > 600, and you need G > 20 to better than break even at 30 turns.
 
It's also worth noting that the forge becomes less important if it's not modifying wealth/research -- the change to the vanilla formula makes the market steal some of the forge's thunder.

And really this whole thing is made more complicated by letting the commerce modifiers apply, since you need to reanalyze how it fits into and affects your whole economy. e.g. the city should build research if its multipliers are more sciencey than average, and you'd have to revisit the idea of earmarking some cities for building research and others for building wealth, with the corresponding building specializations.
 
The comparison above also presumes a fixed citizen assignment. In reality, if you make building wealth terrible, people won't do it except as a last resort -- you'd almost always rather run a scientist or work a hamlet than work a mine.

The 150:hammers: cost of the market shouldn't be compared with building wealth -- a more fair comparison would be more like 75:commerce: 75:food: from working a hamlet instead of a hammer tile (along with the associated cottage turns to develop it!), or 140 :science: 75:gp: from running a scientist, or somesuch. Or three catapults to help roll over your neighbor.
 
I will copy what I posted in general discussion:

In my typical game, courthouses are mostly skipped (I find people overestimate them) and almost always Markets, Groceries and Banks (except in the capital).

Why?
Because they have huge opportunity costs. If you build Wealth instead, you will often get a tech that can be converted to gold and beakers via trade with much higher multiplier (and you do also get a tech itself :)). For instance, Compass usually lands you CoL, Calendar and Construction.
Also, all of those buildings are built in an era where I get aggressive and then it is better (more efficient) to commit to units.
If tech trades are disabled (no fun) or you play Always War, Markets are worth far far more.

Usually, I get some breathing room for Markets around T120, so when I calculate Market or no Market for let's say 30:commerce: city, I will be in position that I can run 50% slider without trades for gold, failgold, GM missions. With them, I am at 80% slider. So, my Market will generate 0.2*0.25*30=1.5:gold:. So, simple return of value says that you'll get your money back in 80T (if you had +25% OR bonus for building a building). Now, there is an opportunity cost which is map dependent, but if you invest into a building instead of simply wealth, you delay Oxford and Printing Press, you delay better units, you delay everything. Add at least 25% of opportunity cost. That is 96T now. since I typically win around T200, it will never have chance to pay for itself.
Ok, sometimes you get a lot of happy, but that is nothing you can't do with 3 warriors in Hereditary Rule, and those are the former spawnbusting guys you already have.
Bear in mind that 30:commerce: is a good commerce city.
In normal game, only 1 or 2 cities you build will pass this check for Market buildings and great AI cities you capture often keep markets, groceries and banks.

Courthouses also don't pass when I try evaluating them like that, however ORG leaders and large empires give them more value and also a tactical advantage of seeing research from high tech civ.
 
Also good commerce cities usually are not working hammer tiles (you much prefer cottages), and gathering 150h is annoying :)

Or whipping commerce tiles away for some time, very often it's a too big annoyance for gold buildings.

Even if all changes are made (no failgold, wealth nerfed and so on), i would rather run some merchants and get 1-2 great ones for gold.
You will use Caste at some point in normal games, and any city can be used (and commerce cities can continue their cottage work).

In multiplayer on big maps, gold buildings are of course popular.
But AIs are usually not forcing you into creating hammer intensive safety nets.
Bur. Caps give more return on everything, but normal cities are usually happy with the minimum need of buildings.
 
I like markets in commerce towns due to the multiplier and happy cap which is most important cause it allows more pop for uses. Grocers as needed when health kicks in and the aquaduct doesnt cut it anymore.

Banks not at all unless I have a 2x holy city with at least one shrine. And then just for it plus maybe 1-2 cottage towns. Though I will say late game I prefer to move off of binary research and settle between 20-80 or 80-20 gold to research once I hit 16+ cities. Though through my first 12 cities it is binary all the way.

I find for the courthouse, forge, market in commerce towns it is better to save 1 pop max whip overflow temples/units for building those buildings after a granary and libby are in.
 
Why would you do binary research at 80/20 and 20/80? :confused: It is more hassle than going all the way, and has less benefit.

For the record

Things binary research does for you

Exploit a fluctuating economy

If you are going to finish your first library in 5 turns, you can get a small boost to your research by accumulating gold for those 5 turns, and then run deficit research with the extra science multiplier.

Conversely, you want to do the opposite if you're about to finish a market or bank in a major :commerce: city -- you want to ensure all of your gold is spent by the time they finish.

This only matters when a significant portion of your economy is about to experience a change in its multipliers.

Reduces roundoff errors

An early game civilization with 15 :commerce: could:

  • spend one turn in 100% slider for 15:science: and 1 turn at 0% slider for 15:gold
  • spend two turns at 50% slider, which would be 7.5:science: 7.5:gold: per turn, but that gets rounded down to 7:science: 7:gold:, so you only make 14:science: 14:gold: overall

This is mainly relevant only when you have a really small economy.

Delays strategic decisions

If you are planning to research mathematics in 20 turns, planning to spend the first 10 turns at 0% slider and the next 10 turns at 100% slider means that if, eight turns later, you realize you desperately need sailing to protect your seafood, you've "wasted" very little research towards mathematics.

Trace gains from tech diffusion

This is like the fluctuating economy example above -- if you meet more civilizations and they research the tech you want to research, you get a slight boost to your science.

Gold reserve for emergencies/events

If an emergency or event occurs during the 0% phase of binary research, you could use the gold on that rather than for fueling later research.

Things binary research does not do for you

It does not make your libraries/markets/etc more efficient

When none of the effects from the previous section apply, then there is absolutely no efficiency to be gained by playing with the slider -- spending 3 turns at 0% research and 7 turns at 100% research will net you exactly the same :gold: and :science: as if you just spent 10 turns at 70% research.

All that matters to your multiplier buildings is the total amount of :science: and :gold: that flows 'through' them: binary research doesn't affect the total, just the timing.
 
I put them in my bureaucracy capital if I need money running binary research late game
 
The single greatest advantage of binary research is in reduced chance that once you get the tech, everyone else will already get it/trade for it too. You can wait and minimize risk, see where the game goes.
 
I put them in my bureaucracy capital if I need money running binary research late game
It shouldn't matter whether you're running binary research or not; you'll get the same amount of :gold: from a bank by keeping the slider at the point where you break even.
 
Why would you do binary research at 80/20 and 20/80? :confused: It is more hassle than going all the way, and has less benefit.

For the record

Things binary research does for you

Exploit a fluctuating economy

If you are going to finish your first library in 5 turns, you can get a small boost to your research by accumulating gold for those 5 turns, and then run deficit research with the extra science multiplier.

Conversely, you want to do the opposite if you're about to finish a market or bank in a major :commerce: city -- you want to ensure all of your gold is spent by the time they finish.

This only matters when a significant portion of your economy is about to experience a change in its multipliers.

Reduces roundoff errors

An early game civilization with 15 :commerce: could:

  • spend one turn in 100% slider for 15:science: and 1 turn at 0% slider for 15:gold
  • spend two turns at 50% slider, which would be 7.5:science: 7.5:gold: per turn, but that gets rounded down to 7:science: 7:gold:, so you only make 14:science: 14:gold: overall

This is mainly relevant only when you have a really small economy.

Delays strategic decisions

If you are planning to research mathematics in 20 turns, planning to spend the first 10 turns at 0% slider and the next 10 turns at 100% slider means that if, eight turns later, you realize you desperately need sailing to protect your seafood, you've "wasted" very little research towards mathematics.

Trace gains from tech diffusion

This is like the fluctuating economy example above -- if you meet more civilizations and they research the tech you want to research, you get a slight boost to your science.

Gold reserve for emergencies/events

If an emergency or event occurs during the 0% phase of binary research, you could use the gold on that rather than for fueling later research.

Things binary research does not do for you

It does not make your libraries/markets/etc more efficient

When none of the effects from the previous section apply, then there is absolutely no efficiency to be gained by playing with the slider -- spending 3 turns at 0% research and 7 turns at 100% research will net you exactly the same :gold: and :science: as if you just spent 10 turns at 70% research.

All that matters to your multiplier buildings is the total amount of :science: and :gold: that flows 'through' them: binary research doesn't affect the total, just the timing.

If you only have 1 way multipliers yes it is slower by about 5%. But if you have both gold and science multipliers you get more total overall commerce % increase than with just gold only or science only. Granted it doesn't kick in till late in the game but once I have 15+ cities when I need to have some org religion time to build in the needed infastructure to new cities.

And I only switch to 80-20 late game when I have a big empire, early and mid it is 100-00.
 
If you only have 1 way multipliers yes it is slower by about 5%. But if you have both gold and science multipliers you get more total overall commerce % increase than with just gold only or science only. Granted it doesn't kick in till late in the game but once I have 15+ cities when I need to have some org religion time to build in the needed infastructure to new cities.

And I only switch to 80-20 late game when I have a big empire, early and mid it is 100-00.
I can't figure out what you're trying to say here, so I'm responding to a guess.

Consider a hypothetical scenario where you have the choice between taking ten turns with

  • Running the slider at 50%
  • Alternating between 80% and 20%
  • Alternating between 100% and 0%

and you have a city with 80 :commerce: and, say, 75% :science: multiplier and 50% :gold: multiplier.

If you run the slider at 50%, you get 70:science: and 60:gold: per turn, for 700:science: 600:gold: total.

If you run 5 turns at 80%, you get 112:science: and 24:gold: per turn. Then if you run 5 turns at 20%, you get 28:science: and 96:gold: per turn. That adds up to 700:science: and 600:gold: total.

If you run 5 turns at 100%, you get 140:science: per turn. Then if you run 5 turns at 0%, you get 120:gold: per turn. That adds up to 700:science: and 600:gold:.

See? You get the same amount of :gold: and :science: whichever way you decide to do things.
 
The single greatest advantage of binary research is in reduced chance that once you get the tech, everyone else will already get it/trade for it too. You can wait and minimize risk, see where the game goes.

Yup that's mainly why i got into using it.
Delaying your own decisions has great benefits against AIs, waiting what they do first.
 
There is one disadvantage of binary research that I haven't seen discussed: when you have a large amount of gold on hand, and either you or the AIs have currency, they may demand gold from you.
 
It's mainly an early game (pre-currency) thing with binary, once you might get gold from AIs each turn i rarely do 0% ones.

If my next tech is valuable, i can do more harm than good by gathering too much gold ;)

What i sometimes do then is one lower science turn, with gold calculated for my next turn being 100% again.
(but without collecting more gold than needed for that)
 
Top Bottom