Just beat the game at monarch difficulty for the first time...wonder vic'try (korea)

[Terra Nullius]

good explanation of pumps, my exalted mr eggplant

but...

in this case, we're talking about the capital. and i would still say that *generally, you want to keep the capital as high in pop as you can manage. there will always be competing priorities and pushing out settlers is certainly important.

signore paluda's example certainly shows a need to master the techniques of pumping, and you've provided a pretty decent short guide there.

my focus here is more to do with the economics. your explanation seems to be centering on the production issues, which is something i was aiming to steer ariano away from. my impression from the screenies is that he is focussing on rex and production - at the expense of building a decent economy - and hence running into trouble at expanding his game.

nuff said, i'm glad someone took the trouble to disagree with me.

the other part of the question which i didn't deal with before... 'should i be bee-lining for monarchy'

first, *personally, i disagree with bee-lining for philosophy in the first place. it seems like a bit of an exploit and will probably be gone once civ4 kicks in. if you're doing it for the fun of getting ahead and kicking some ai gluteus, ...by all means, go for it.
if you're trying to improve your overall play, i don't see any value in it. if anything i think it blinds people to important subtleties. each map is different. your decisions about which research path to take should be based on your specific situation at the time.

it's all a gamble, and some of them don't pay off. but i think you should get used to looking at the particular situation and then placing your bets accordingly.

don't just go after philosophy because you can beat the ai. similarly, don't make a rule of pursuing monarchy.

what specific reasons do you have for wanting to change to monarchy? taking a look at the map, i would expect a lot of people to choose republic far ahead of monarchy.

[vulture]

Originally posted by lebensraum
good explanation of pumps, my exalted mr eggplant

but...

in this case, we're talking about the capital. and i would still say that *generally, you want to keep the capital as high in pop as you can manage. there will always be competing priorities and pushing out settlers is certainly important.

signore paluda's example certainly shows a need to master the techniques of pumping, and you've provided a pretty decent short guide there.

my focus here is more to do with the economics. your explanation seems to be centering on the production issues, which is something i was aiming to steer ariano away from. my impression from the screenies is that he is focussing on rex and production - at the expense of building a decent economy - and hence running into trouble at expanding his game.

nuff said, i'm glad someone took the trouble to disagree with me.

I think the economics favours building more cities over growing existing ones, at least at first (as always, it gets more complex in real game situations). Say you have a city that is size 2, about to get to size 3. You have the choice between producing a settler or letting the city grow. If you carry on growing the city, you get to size 3 (working 4 tiles). If you produce a settler, you end up with two cities of size 1, together working 4 tiles. After another 10 turns they grow. If you stuck with a single city, you now have a size 4 city, working 5 tiles. Or with the settler, two size 2 cities, working 6 tiles. Your economics are in better shape, even before you take into consideration the long term benefits of more cities and more land (and more resources).

This is implicitly making a few assumptions though. Firstly, that the two cities in question have similar levels of corruption. Secondly, that it takes zero time to move to a new site and build a city (which is reasonable - the loss of 2-3 turns moving is compensated for pretty quickly, so it's a non factor). Thirdly that the cities grow at the same rate. Fourthly, that the cities grow at a constant rate regardless of size. Fifthly, that building marketsplaces or libraries isn't yet an option. And Sixthly, that you're able to defend both cities adequately.

Pretty much any of those can be a reason not to build a settler at some point. But very early in the game, building a settler is better not only for your long term potential, but also for your immediate shield and commerce production. (Not for culture production though).

Taking the points in order. Corruption. Taking two pop from your very productive capital to found a city that is going to be producing 1 gold and 1 shield until a courthouse can be built is obviously sacrificing quite a bit of production, for quite a long time. It can be worth it to grab a new resource, but beyond that, you're generally better off letting the AI settle there and taking the city when you can make better use of it. Building a settler in one highly corrupt city to found another highly corrupt city OTOH costs you very little in immediate and mid-term production, and is more often worth the effort.

Skip number 2, since it is so rarely important.

Numbers 3 and 4: city growth rates. Building a settler in your capital can actually be used to speed up your growth. Imagine your captial can only ever generate +2 food, growing every 10 turns. But you found a city that has +3 food, growing every 7 turns. You use 2 pop to do this, but after 14 turns you can (in principle) get those two pop back from the new city (via a settler or two workers). At that point, you are at the exact same size, and with the same food in the food box, as if you'd never built a settler, yet you still have another city growing happily. Founding new cities that grow faster than the host city, and funneling the extra population back to the host via workers leads long term to faster growth for the original city - worth doing even without the extra benefits of having one more city and more territory. Cities that grow to the point where they have to start working hills or forests, and slow down their growth dramatically are often better off producing a settler to keep them at a size where they grow quickly again. Population is power.

Number five: markets and libraries. Taking population from a city with one of these to build a city without one, with worse corruption, is obviously a significant loss to your immediate economy. That's one good argument for settler pumps as well. The settler pump doesn't waste shields on these building early on, while the cities that do build them don't worsen your economy by building settlers.

Lastly: defending your cities. This is the one that most often limits my growth. If you are fortunate enough to get a 4 turn settler pump, it can be hard for all your other cities combined to produce enough military to defend all those cities. I sometimes find myself with undefended settlers, undefended cities and bribing the AI not to attack (would never try this vs a human player), until I can get enough camp cities running to churn out those spearmen fast enough. I probably ought to add a few of those settlers to some of the unit cities until they can keep up. On emperor, a 4 turn pump can out-produce my military, and out-expand the AI. Which is all you need really.

the other part of the question which i didn't deal with before... 'should i be bee-lining for monarchy'

first, *personally, i disagree with bee-lining for philosophy in the first place. it seems like a bit of an exploit and will probably be gone once civ4 kicks in. if you're doing it for the fun of getting ahead and kicking some ai gluteus, ...by all means, go for it.
if you're trying to improve your overall play, i don't see any value in it. if anything i think it blinds people to important subtleties. each map is different. your decisions about which research path to take should be based on your specific situation at the time.

it's all a gamble, and some of them don't pay off. but i think you should get used to looking at the particular situation and then placing your bets accordingly.

don't just go after philosophy because you can beat the ai. similarly, don't make a rule of pursuing monarchy.

what specific reasons do you have for wanting to change to monarchy? taking a look at the map, i would expect a lot of people to choose republic far ahead of monarchy.

As you say, there isn't a one size fits all strategy. The extra tech from philosophy is nice if you get it, but you need to be able to keep up with (or ahead of) the competition without it. You will get far more value from being able to research techs that no-one else knows and trading them than you will from getting to philosophy. You just have to find out what techs are a) good value and b) not a research priority for the AI (or that you can beat it to). Which depends on what you start with, and of course changes as the game progresses.

If you start with ceremonial burial, going for monarchy can be good. Mysticism isn't too useful as trade bait, but polytheism is, as is monarchy istelf. Except in unusual circumstances, almost all my decision about what to research are based around how useful a tech is to trade rather than whether it is intrinsically useful itself. Let the AI research those, and you trade for them. And since it's fairly easy to rob the AI blind in trade (well, kind of), you generally end up with the useful techs just as quickly as if you'd researched them yourself (plus you generally end up with most of the disposable cash in the world to boot).

[Terra Nullius]

habibi ganoujh,

i now understand why i have yet to see a handy rex calculator lying about the place. just spent the night playing 'round with spreadsheets to try and prove you wrong. i think i've sent myself cross-eyed.

the exponential nature of the thing is not well suited to spreadsheets. i should have just written a program to do the calcs instead. anyhow it's going to take some time before i get anything solid.

at this stage i'm using the following assumptions
1. all tiles produce 2 food, one shield, one commerce
2. settlers take three turns to found a city (i can't help myself)
3. one civ produces a settler every time it reaches pop3 (if it has sufficient shields)
4. the other builds a granary, then builds settlers at pop5
5. no special bonus tiles, no rivers, no nuffin'
6. first ring of cities (n<=4) no corruption
7. second ring of cities (n>4) some corruption (probably 20%)
8. city centre produces two food, one shield, two commerce
9. i must be off my nut

the only results i've got so far are very sketchy, but i seem to come up with this

after 40 turns
civ one has four cities and 196 gold
civ two has three cities and 190 gold

after 80 turns
civ one has 16 cities
civ two has ten cities
no results for commerce that i'd care to put in writing

roughly speaking,
civ one cities produce one settler every 20 turns giving
No. of cities = 2 exp(n/20)
where n=number of turns
(don't quote me on this, i still haven't thought it through properly)

civ two
builds a granary after 25 turns
builds first settler on turn 35
continues to build settlers every ten turns thereafter
all subsequent civ2 cities produce a settler every 20 turns

this is going to take a lot of polishing before it actually gives any worthwhile results

yours in tahini,
lebensraum

[Terra Nullius]

slow progress on the rex calc

the basic mechanics are done, but it needs a few extra features (like including the effects of corruption) and some tinkering to make it more presentable

the results are not quite what i expected
using the following scenarios
.i) build: warrior, warrior, settler,... (ie, 1<=pop<=3)
ii) build: warrior, warrior, granary, warrior, warrior, settler,... (3<=pop<=5)
iii) capital builds: w, w, granary, w,w, settler, ...
subsequent cities build: w, w, settler, ...

after 100 turns, i get
i) # of cities: 32
. commerce: 1550

ii) # of cities: 14
. .commerce: 1340

iii) # of Cities: 23
. .commerce: 1379

nb. i haven't debugged the whole thing yet, so this may not be completely accurate. but it seems to be about right. just thought people might like an update..

[Modo44]

I think your test is a bit too far from in-game reality. Early Granaries and tons of Workers simply win games. This is not a math thing, this is from exparience.


Try a Granary after the first Settler.

Try a Chop or two to speed up the early Granary, especially in food-rich cities. I know this is against your assumptions, but you do get different starts in Civ, and you often do get to Chop some Forests.

Something you fail to see is the fact that Granaries allow you to build various things (like units and basic structures) between Settlers. Cities without Granaries will often be stuck producing nothing but Settlers. This is most important for the capital, since there is no corruption to chew on that advantage.

Whether you want to or not, skillful Worker-management can help a lot to make the Granaries go online and productive quickly. This is a combination requiring skill and often careful planning, so looking at one without the other makes little sense.

[Terra Nullius]

yep, the effect of workers is something i still have to figure in. at this stage i'm using averages for all terrain values. the numbers can be altered, but as for adapting terrain values as the turns progress. hmm, that will take some work to implement.

forest chops.. can't do it yet but that gives me some food for thought

the main reason for the specific build order i chose (warrior, warrior, settler) is it allows the city to build on the same turn that growth occurs. of course anyone who does nothing but build warriors will get smashed in short order.

note that i have not included any figures for unit support at this stage.

Granaries allow you to build various things (like units and basic structures) between Settlers

not necessarily
in fact with the eg.s above, each city only gets to build two warriors ie. approx. twenty spare shields between settler builds

my conclusion is: growing the city is what actually gives you the freedom to fit in extra builds
for eg. build: w, w, s, barracks, s, library, settler, settler
my other conclusion at this stage is: it doesn't matter too much whether you build your settlers at pop3 or pop5. what is really important is to take advantage of any terrain bonuses, especially food

scenario iv)
capital builds: w, w, s, w, w, granary, w, w, s, ...
subsequent cities: w, w, s ...

after 100 turns
# of cities: 27
commerce: 1481

[Modo44]

Originally posted by lebensraum
in fact with the eg.s above, each city only gets to build two warriors ie. approx. twenty spare shields between settler builds

But, as you noted, this can get bigger if you let the city grow first. Making a Spear or a Sword between Settlers is pretty good. The difference is, if you make the city cruise between sizes 1-3, no room for additional builds remains, and you screw yourself big time. Remember this adds up with time. Plus, cities with a large food surpulus need to grow big enough for the production for a stable Settler-factory in the first place.

The two quotes below seem contradict each other a bit...

Originally posted by lebensraum
my conclusion is: growing the city is what actually gives you the freedom to fit in extra builds

Originally posted by lebensraum
my other conclusion at this stage is: it doesn't matter too much whether you build your settlers at pop3 or pop5. what is really important is to take advantage of any terrain bonuses, especially food

The reality is, you need to grow to an optimum size that alows a sustainable cycle with little or no use of the lux slider. It can mean combining multiple cities to share food bonuses.

[Terra Nullius]

Originally posted by Modo44
The two quotes seem contradict each other a bit...

sorry, i was being vague again

i) a bigger city or a city with granary does not produce more shields between settler builds.

ii) it is more important to make use of terrain bonuses than to stick to a rule such as ... always build settlers at pop 3/ pop 6.

The reality is, you need to grow to an optimum size that alows a sustainable cycle with little or no use of the lux slider.

omigawd... now he wants me to figure happiness into the equation as well!


i have actually thought of a device that would allow worker actions into the picture. the tricky part is including enough detail to make the results meaningful, but not so much that it becomes overburdened and difficult to use.

the idea is not to create the perfect tool that will go out and play the game for you, just a quick reference that allows fairly accurate comparison between different rexing policies

ps. scenario v)
capital: w, w, s, granary (with chop), w, s, w, w, s
others: w, w, s, w, w, s, ...
27 cities, 1557 commerce

caveat rectus:
the following is not worth reading unless you are having a really, reeaaally slow day at work

re: the conclusions cited above:

i) basically, what i was trying to say is;
case 1) build settlers on the turn that a city grows:
only (approx.) twenty shields will be generated between settler builds
# ie. city size is not a significant factor in determining what you can fit in between settler builds.
(applies to either pop3 or pop5 in this eg.(it may not apply exactly to other pop sizes))

case 2) growing the city, instead of producing settlers straight away:
a larger city will generate more shields
# ie. if build order = ..., library, settler, settler
then time to complete (pop5 city) < time to complete (pop3 city)
your overall rate of settler production is not necessarily affected since you can pop out consecutive settlers once the improvement is built

the distinction here is that;
if you are constantly building settlers, then having higher pop does not allow you to fit in anything extra between builds
it is only when you pause settler production to complete an improvement that having the higher pop is beneficial

Granaries allow you to build various things (like units and basic structures) between Settlers

in other words; no, a granary does not allow you to build things between settlers, but if you stop building settlers to complete an improvement, then a granary will help you to complete the job faster

oh my cleo, look how much space i took to explain such a minor distinction, .. am i really that an#l?
(no bw, you may not respond!)

[Modo44]

Originally posted by lebensraum
in other words; no, a granary does not allow you to build things between settlers, but if you stop building settlers to complete an improvement, then a granary will help you to complete the job faster

It does.

For cities that can not become 4(or less)-turn Settler-Pumps, you will quickly get lots of production with relatively low food. A Granary can mean up to a whole Settler's worth of production added (Mined BG...), and that is without slowing down the REX. Even if you only get a Curragh or a Warrior "for free", that is still an enormous advantage simply because the early moves matter more.

A 4-turn Settler Pump can only be done with a Granary (except if you get ridiculous numbers of tiles with food bonuses), so there is no going around it, right?

Of course, the short-term slowdown will show. But, again, you need to figure in the Workers the Granary will give you. The point is, once online, you get much faster land improvement, and your REX speeds up much faster because of it. Your calculations are still missing that crucial factor.

As for build order, try: Warrior, Warrior, Settler, Granary, Worker, [Warrior][Archer] (depending on production/food ratio due to first Worker), Settler.

[Terra Nullius]

lordie, it's almost done!!!

i should also be able to account for settler pumps and such once i beat the pole into submission

now, you know how i hate to disagree with you modo,
but...
scenario 1)
build settlers at pop3
[IMG]M:\Civilsation\Rex Calculator\modo_pop3.jpg[/IMG]

scenario 2)
build granary, then w, w, s,...
[IMG]M:\Civilsation\Rex Calculator\modo_granary.jpg[/IMG]

now i'll admit both of these scenarios are artificial, however, as you can see, they both result in producing two warriors or equivalent between settler builds

the basic difference seems to be that
scenario 1)
takes 20 turns to grow two pop points
takes 5 turns to produce a warrior
takes 10 turns to stack away the shields for a settler

scenario two)
takes 10 turns to grow two pop points
takes 2 to 3 turns to produce a warrior
takes 5 turns (at pop5) to store the shields for a settler

there is also the issue of wasted shields to consider.

generally however, if your city is rexing as fast as it can. it still only has say, 25 shields to spare in between settler builds.

[Modo44]

Originally posted by lebensraum
generally however, if your city is rexing as fast as it can. it still only has say, 25 shields to spare in between settler builds.

Only if you assume no BG, food bonus or other things are in the city radius, and not growing over size 6 (i.e. no river/Lake present to make building Settlers useful). Yes, in this "clean" situation you "only" get 25 Shields and the added trade income, which will still allow you to over come your cruising-at-size-3 enemy.

But, in the real game, you can often get faster growth for a pump or increase production to exceed the 25 Shields. Again, this is a problem of this simplfication. If you can make it to size 6, here is one bonus. If you have Bonus Grassland around, here is another. This adds up.

But no matter the number of tiles or city size — the Granary always gives you a powerful long-term increase in productivity.

[Terra Nullius]

Originally posted by Modo44
But no matter the number of tiles or city size the Granary always gives you a powerful long-term increase in productivity.

i was soo close to having to agree with you there.

all except the always...
i think the analysis shows that in the abscence of corruption or food bonuses, granaries are next to useless.

granaries reduce the effect of corruption

intuitively, this makes sense to me. population in the core is simply more powerful than pop on the outskirts. if you take away the corruption, there is no longer a benefit to be had. you might as well just keep popping out settlers at pop3 'til you cover the map.

allowing for corruption, it makes sense to increase population in the core, because it is worth more to the economy and more productive.

granaries do not (of themselves) allow faster rex-ing
granaries do not (of themselves) allow more improvements to be built
granaries do reduce the effects of corruption by increasing the power of the core cities

in real game terms, i think this suggests different stategies to the common wisdom. to me, it suggests that where corruption is minimal and no food bonuses are present, low pop and no granaries is the more powerful way to rex.
granaries should not be built and settlers should probably be built at pop3 until the core cities are laid out and corruption becomes a significant factor.

(cue: it's only a paper moon)

i certainly agree that some of the original assumptions may be flawed. in a settler pump, the average yield per tile (from terrain) is actually 1.25 shields(plus one shield from the city centre)
real cities probably vary between about 0.75 and 1.25 per tile.

haven't had a chance to test the effects of corruption yet. i'll focus on polishing up the last few things and then people can run the tests and decide for themselves.

[Terra Nullius]

ps. not that i have anything against granaries myself.

i have done wonders before even doing a granary as the very first build. popped a hut with cultural expansion - got a settler. chopped forest. irrigated a cattle tile just inside the (expanded) cultural boundary.

with a start like that, who needs warriors?

[Modo44]

Originally posted by lebensraum
i was soo close to having to agree with you there.

all except the always...
i think the analysis shows that in the abscence of corruption or food bonuses, granaries are next to useless.

Thing is, only your capital actually is free of corruption...

But you know what, you almost convinced me, at least for this special case (no food bonuses, no Chops). I see that very early slow growth (while building Granaries) might indeed hamper the REX if the opponents are near enough to stop you. There are still problems that need to be solved to make me go your way.

You care only for landgrab in your REX considerations, not taking care of anything else. Meaning, economy and military get thrown out of the window. This is a major flow. You probably get many cities, yes, but very small, weakly defended, and weakly populated ones. In other words, you are dead (or soon becoming), except against a low-level AI. Here is an idea. Try checking more stats of your civ (trade, production, population, military, Workers) at "end point" (a set date, maybe different dates to allow us to compare it with games we play), depending on the REX scenario.

You did not find out how Workers and their abilities fit into the eqation. How do Granaries support adding new Workers? How do Roads connect your civ for army and Settler movement?

You failed to show us which build pattern is faster in the long run. You failed to show if/when one becomes better than the other. Another idea. Maybe you already have a spreadsheet, and could make some graphs? Use more options (minimizing Shield loss due to other builds, if you can): Granary, Settlers at size 3; Granary, Settlers at size 5; Settlers at size 3; Settlers at size 5; Settler, Granary, Settlers at size 3; Settler, Granary, Settlers at size 3. This would make for a good comparison, methinks.

You have not checked how Chops, food bonuses (like playing an Agricultural civ ), production bonuses (Mined BG can dramatically increase the in-between-Shields you can get with a Granary), and other variables (like player's skill with Workers) add up. Granted, those would be hard to do.


I prepared a test scenario for everyone to play with. It allows you to check how Worker management, Roads, and the need to Irrigate figure into your REX (and they do, believe me). It is a standard map, filled Plains and a very short River (so you can Irrigate), with a secluded 1-tile island for one AI. You need to select the Zulu as your opponent to be sure you get the right starting spot. Enjoy.

I think we can call the threadjack complete now, eh? ;-)

[Terra Nullius]

Originally posted by Modo44
Thing is, only your capital actually is free of corruption...

true, but the same effects should play out where corruption is negligible

There are still problems that need to be solved to make me go your way.

i certainly hope so

You care only for landgrab in your REX considerations, not taking care of anything else.

agreed, but it's a complex game, so you've got to pick a limited set of variables and even then the model won't cover all the sneaky little things we do to punish nasty ai's

Try checking more stats (trade, production, population, military, Workers) to allow us to compare it with games we play), depending on the REX scenario

now that's ambitious!
at this stage, average production and build order are inputs, commerce and pop are the outputs. anyway, good ideas let's see what sort of response we get from this version first.

You did not find out how Workers and their abilities fit into the eqation.

agreed, this is a tough one. i want to keep the thing fairly simple overall. nobody wants to spend hours entering numbers into a spreadsheet when they could be playing the game. it'll take some research to make this work.

How do Granaries support adding new Workers?

that'll have to wait for the next version

How do Roads connect your civ for army and Settler movement?

oops, just remembered, one last change to make before i post it

Maybe you already have a spreadsheet, and could make some graphs?

soon, my preciousss, sSsoon, ... achh, hiss. gllm

Use more options (minimizing Shield loss due to other builds, if you can

nope at this stage, there is no allowance for the effect of courthouses. personally, i think this is fair. by the time courthouses are built, the game is already well underway. the variables (in terms of intelligent use of workers, chops, so on) become so many that the model would become hopelessly inaccurate or really tedious, i'd rather just fire up another game and have some fun being an uninformed idiot.
my focus for some time has been on the first three cities if you get that right, you've got a game. so for me this is more to do with optimising those first few cities.

Granary, Settlers at size 3; Granary, Settlers at size 5; Settlers at size 3; Settlers at size 5; Settler, Granary, Settlers at size 3; Settler, Granary, Settlers at size 3.

you want fries with that?

Chops

easiest way to do it at this stage is adjust the cost of granaries to 50 shields i'll post some results on that when i get a chance

food bonuses (like playing an Agricultural civ)

done - excess food can be specified for different cities

production bonuses

also done, city centre production and terrain averages can be specified

and other variables (like player's skill with Workers)

sure, just spank my barb and call me soren.

Enjoy.

rex vs commerce calculator

I think we can call the threadjack complete now, eh?

jack??!!!

what jack? i don't know jack!