Yet another Early Growth Thread

[cv431410]

when you exapand, how fast should you build the cities?

Is it too fast? is it too slow?

You would agree with me:
0% research ==> expand too fast,
100% research ==> expand too slow.

You said it depend on what yoour objective is.

OK, give a particular objective (short term objectives for simplicity):

CS slingshot
win a peaceful space race
maceman rampage
calvery rush
maximum score at turn 100
maximum score at turn 200
...

how fast should you expand initially as measured by a particular variable (say research %, number of cities, ...)?

Now let us use this objective:

maximum score at turn 100.

Let us simplfy as much as possible:

Assumption: 1 cont each or you are in the back of inland sea, so you are safe from early attack, the goal to build your economy up asap.

I am a poor researcher so I call good researchers in this forum to do 10 experiments with the same start position. The experiment depends on your land, your resources, and your situation in the game. However, the results should still be very interesting

Background: If you watch high level AI, they are all expansion oriented, i.e. get a few cities up asap. These experiments are all about expansion.

Condition:
No barb
No AI
No unit building
No War
Please build 5 cities in all option;
Please stop expansion at 5 cities;
Please build 5 cities in the same locations in all 10 options;
Same research path;
All 5 cities will have the same building order in all 10 options;


Objective:
1 At the end of 100 turns, determine which of the 10 options is optimal;
2 At the end of 200 turns, determine which of the 10 options is optimal;



Let me list two extremes first:

1. 0% Research
2. 100% Research

1. 0% Research

In this approach, assuming you have enough forest and seafood/rice…, you will use slavery and chopping as fast as available. Build nothing but workers and settlers. Start building a warrior until size 2, switch to worker and slave asap. Workers start chopping. Build warrior or granary and wait city grow to size 2, and switch to worker/settler. If you start from ancient, it is quit probably possible to get 5 cities and 6 workers when you reach 0% research. Here you must stop, or your workers become suicidal.

2. 100% Research

Just build a super-city with 2, 3 workers and grow it to say size 5, 6, or even 7. It has forge, 3 mines, a library and is very productive. The research rate is always 100%. It has so much production that it produces a settler every 2 turns. At this point, you can produce 4 settlers in 8 turns and have these additional cities produce workers.

In a game of 100-turns, 130-turns, 150- turns, and 350-turns, which one is better? How do we justify which one is better than others? I have tried all of them and have no clue which one is better.

The other 8 options are 90% research, 80% research, …. For example:

(3) 60% Research
Slave/chop 3 cities + 4, 5 workers in the earliest possible way, maintain 60% + research, and take a break. When research recovers to 80% +, expand again.


(4) 80% Research
Slave/chop 1 cities + 2,3 workers in the earliest possible way, maintain 80% + research, and take a break. When research recovers to 90% +, expand again.

At 100 turns, write down your score. This should give a function between power and expansion rate.


Results

Option ...... 100 turns ...... 200 turns
0% ............ ? ............... ?
10% ...
20%
30%
40%
50%
60%
70%
80%
90%
100%


Please let us know.

 

[dime]

I am not a very advanced player myself and get my ass handed to me on monarch... that being said, I cannot understand what the purpose of this thread would be - especially not with having other people look into it.

First of all - being isolated is in any event a rare occassion and a not particularly wanted one - at least not if the AIs are in contact as this will undoubtedly boost their research.

Second - what's the point? See below:

Objective:
1 At the end of 100 turns, determine which of the 10 options is optimal;
2 At the end of 200 turns, determine which of the 10 options is optimal;



Let me list two extremes first:

1. 0% Research
2. 100% Research

1. 0% Research

In this approach, assuming you have enough forest and seafood/rice…, you will use slavery and chopping as fast as available. Build nothing but workers and settlers. Start building a warrior until size 2, switch to worker and slave asap. Workers start chopping. Build warrior or granary and wait city grow to size 2, and switch to worker/settler. If you start from ancient, it is quit probably possible to get 5 cities and 6 workers when you reach 0% research. Here you must stop, or your workers become suicidal.

Well - it is very possible to go to 0% research as long as you have gotten writing and agriculture, as you can farm and use specialists for all research... this would actually probably give you better early reasearch than building cottages as they take a fair amount of time to grow - especially if you are playing a philo leader which could use the great people to lightbulb techs

2. 100% Research

Just build a super-city with 2, 3 workers and grow it to say size 5, 6, or even 7. It has forge, 3 mines, a library and is very productive. The research rate is always 100%. It has so much production that it produces a settler every 2 turns. At this point, you can produce 4 settlers in 8 turns and have these additional cities produce workers.

In a game of 100-turns, 130-turns, 150- turns, and 350-turns, which one is better? How do we justify which one is better than others? I have tried all of them and have no clue which one is better.

By the time you have a city large enough to churn out settlers at the pace you describe, you could theoretically churn out settlers to your heart's content, but there would be no land to settle. You either get them out early or not at all (with the one caveat if you somehow manage to block off the AI from some land). Otherwise go for military conquest instead

The other 8 options are 90% research, 80% research, …. For example:

(3) 60% Research
Slave/chop 3 cities + 4, 5 workers in the earliest possible way, maintain 60% + research, and take a break. When research recovers to 80% +, expand again.


(4) 80% Research
Slave/chop 1 cities + 2,3 workers in the earliest possible way, maintain 80% + research, and take a break. When research recovers to 90% +, expand again.

At 100 turns, write down your score. This should give a function between power and expansion rate.


Results

Option 100 turns 200 turns
0% ? ?
10% ...
20%
30%
40%
50%
60%
70%
80%
90%
100%


Please let us know.

This is absolutely ridiculous. It could only be done replaying the same map, and still, you'd have variances because of e.g. AI building or not building wonders, you yourself building wonders, possibilities of war etc. Any player that believes that the way to go is optimizing expansions simply based on current science rate levels needs to try to get a better understanding for the game. I can be at a 40% science rate level (theoretically overexpanded) and see a potential city with strategic resources that I can claim in order to stop the AI from getting e.g. Iron... which means he'll soon be a sitting duck. In this context, research rate is of course utterly meaningless and the appropriate rate of expansion will be different in each game... it will also depend on which route you take in research... if you go for the upper half of the tree (Currency, CoL) you'll be able to expand more due to courthouses and markets but will need to do it through settlers. In the lower half of the tree, you get strong military techs that means you can be very successful on the attack, with pillaging and city conquering money "paying" for your too high science levels... you run a deficit but the conquest pays for it

/Andreas

 

[cv431410]

This is absolutely ridiculous. ....
/Andreas

The point is this:
when you exapand, how fast should you build the cities?

Is it too fast? is it too slow?

You would agree with me:
0% research ==> expand too fast,
100% research ==> expand too slow.

So what should it be?
70%? why 70%?
80% why 80%?

You probably would say even with this experiment, it is still hard to say. Ture, but it is a start.

Obviously, there are many, many variables, which are hard to control. So you will need do all you can to control these variables, such as:

Remove wars from the picture;
Remove AI from the picture
One cont each;
Use the score as the only measurement of growth (to start with);
Use the research rate as a measurement of expansion (to start with);
Build 5 cities in all options;
Stop expansion at 5 cities;
Build 5 cities in the same locations in all 10 options;
Same research path;
All 5 cities will have the same building order in all 10 options;

...

Obviously there are alternative to measure the expansion. For example,

how many cities do you have at turn 10, 20, 30, 40, ...? but that 's more complicated than the proposed one.

Get something simple and easy to use, so in the game you can say "I am growing too fast" or "I am growing too slow".

Note you are not fighting with AI here, you are simply try to expand optimally.

 

[dime]

I understand where you are coming from - however, I still don't see the point. I mean, if you didn't have to worry at all about war, this would mean you could safely ignore all military techs early and beeline for CoL / Currency... and then continue for Calender, Civil Service, Monarchy, Guilds, Banking, Drama (all techs that give health, happiness and $$$ enhancing civics and buildings) which in effect would support expansion.

In reality, however, this scenario doesn't exist. You WILL need military (to address barbs, to potentially rush a neighbour etc). If there is a neighbour capital 20 tiles away, forget about the settlers (or build one city to get the copper) and just rush him to take him out and get more land. If you have juicy land, just expand like hell...

The rule of thumb - add a city whenever you can
a) be pretty sure it'll pay for itself early
b) claims a strategic resoure or a resource which is limiting your city growth (e.g. if you're at the happiness cap with some of your cities and can add gems)
c) it blocks off land from the AI, to give you more time to settle it later (though at higher levels the AI will build galleys early, so may not work)

To give an example why the experiment is a tough one... In a recent game playing as Mehmet, I got a start in the coast with predominantly grassland on a piece of land 2 tiles broad, between the ocean and a big (12 tiles) lake... I also had fish, cow and pigs in the fat cross, plus alot of forest and hills. Perfect production city that could double as GP farm if I have nothing to build. I proceeded to scout around and found Kublai close, and alot of juicy land... get this; primarily grassland, some floodplains, a total of 4(!!!!) gold mines (requiring three separate city locations), all with food resources nearby. One of those cities also denied Kublai horses, which given his UU was priority #1 (he may still have horses somewhere else tho...). So - I quickly rushed settlers, workers and defenders, not at all thinking about wonders, and focused on getting each new city a library for holding off any cultural pressure + getting a solid border to fill in the land in between afterwards (Kublais capital was too far away for an early rush and I was missing copper anyways). In this scenario, I built 6 cities almost instantly. The fact I had 4 gold running, as well as some early cottages on flood plains and some sea resources ensured I had no problems whatsoever doing this and maintaining a solid research pace... research rate was ~70% (meaning I could have probably expanded MORE) since remember this is 70% out of much more base commerce than I would have had with only 3 cities. So - rather a 60% rate with 6 solid city locations than 80% with only 2 cities.

Now - for comparison, should you elect to play a highlands map with few food resources and little grassland and end up in the middle of the map, then of course your expansion rate will have to be slower. It will be more difficult to work any gold/silver/gems mines without enough food and you'll likely have fewer cottages in cities with only plains/hills as cottages will be "food negative"... hence, expanding too quickly will kill science.

I am just mentioning these examples to highlight why I think the test is pointless... depending on what type of starting location you get, the answer may be go for 3 cities early or go for 8 cities early.

Typically on any map, I'll try to get 4 cities early. I will tend to put them close to each other without alot of "empty space" in between to avoid distance maintenance, apart from when I really want a specific city location to claim resources, or to block the opponent. If I am playing a leader that favors the CoL route (either philo/spiritual going for caste system or financial) and lack rushing opportunities (no UU, no copper or no convenient target) I will usually expand even more as I know courthouses will be up shortly. After this, I try to always have a settler available for whenever an opportunity presents itself... typically, if a city location IS good, it'll pay for itself shortly. If it isn't so good, then maybe there is no point in settling it immediately, as it'll cost you more than it'll contribute.

Another point worth mentioning - I used to be very focused on settling locations before, trying to deny the AI from getting them as they'd invariably put up a city in the "wrong" tile and spoil it... of course, if you need to take and raze the city and then rebuild you might as well do it yourself from the start. With the new patch, however, I feel the AI is increasingly putting cities in better locations (still one tile from the coast sometimes tho) and then it may be as well to let the AI build the city - you can always take it later. Hence, the decision criteria for expanding is easy - can the city contribute short-term to more commerce and give resources (either that I don't have or that I can trade) alternatively, does it deny the AI of a strategic resource that weakens it militarily?

/Andreas

 

[DaveMcW]

The answer is 42%.

What was the question again?

 

[cv431410]

In reality, however, this scenario doesn't exist. You WILL need military (to address barbs, to potentially rush a neighbour etc). If there is a neighbour capital 20 tiles away, forget about the settlers (or build one city to get the copper) and just rush him to take him out and get more land. If you have juicy land, just expand like hell...

Take out as many factors as you can, No barbs.

The rule of thumb - add a city whenever you can
a) be pretty sure it'll pay for itself early
b) claims a strategic resoure or a resource which is limiting your city growth (e.g. if you're at the happiness cap with some of your cities and can add gems)
c) it blocks off land from the AI, to give you more time to settle it later (though at higher levels the AI will build galleys early, so may not work)

Again, take (c) AI out to suimplify the experiment.

Everything you said are true. But, all what you are saying is qualititive, not quantitative. What I am trying to get is "measurement", not feelings.

To give an example why the experiment is a tough one... In a recent game playing as Mehmet, I got a start in the coast with predominantly grassland on a piece of land 2 tiles broad, between the ocean and a big (12 tiles) lake... I also had fish, cow and pigs in the fat cross, plus alot of forest and hills. Perfect production city that could double as GP farm if I have nothing to build. I proceeded to scout around and found Kublai close, and alot of juicy land... get this; primarily grassland, some floodplains, a total of 4(!!!!) gold mines (requiring three separate city locations), all with food resources nearby. One of those cities also denied Kublai horses, which given his UU was priority #1 (he may still have horses somewhere else tho...). So - I quickly rushed settlers, workers and defenders, not at all thinking about wonders, and focused on getting each new city a library for holding off any cultural pressure + getting a solid border to fill in the land in between afterwards (Kublais capital was too far away for an early rush and I was missing copper anyways). In this scenario, I built 6 cities almost instantly. The fact I had 4 gold running, as well as some early cottages on flood plains and some sea resources ensured I had no problems whatsoever doing this and maintaining a solid research pace... research rate was ~70% (meaning I could have probably expanded MORE) since remember this is 70% out of much more base commerce than I would have had with only 3 cities. So - rather a 60% rate with 6 solid city locations than 80% with only 2 cities.

Now - for comparison, should you elect to play a highlands map with few food resources and little grassland and end up in the middle of the map, then of course your expansion rate will have to be slower. It will be more difficult to work any gold/silver/gems mines without enough food and you'll likely have fewer cottages in cities with only plains/hills as cottages will be "food negative"... hence, expanding too quickly will kill science.

I am just mentioning these examples to highlight why I think the test is pointless... depending on what type of starting location you get, the answer may be go for 3 cities early or go for 8 cities early.

Typically on any map, I'll try to get 4 cities early. I will tend to put them close to each other without alot of "empty space" in between to avoid distance maintenance, apart from when I really want a specific city location to claim resources, or to block the opponent. If I am playing a leader that favors the CoL route (either philo/spiritual going for caste system or financial) and lack rushing opportunities (no UU, no copper or no convenient target) I will usually expand even more as I know courthouses will be up shortly. After this, I try to always have a settler available for whenever an opportunity presents itself... typically, if a city location IS good, it'll pay for itself shortly. If it isn't so good, then maybe there is no point in settling it immediately, as it'll cost you more than it'll contribute.

Another point worth mentioning - I used to be very focused on settling locations before, trying to deny the AI from getting them as they'd invariably put up a city in the "wrong" tile and spoil it... of course, if you need to take and raze the city and then rebuild you might as well do it yourself from the start. With the new patch, however, I feel the AI is increasingly putting cities in better locations (still one tile from the coast sometimes tho) and then it may be as well to let the AI build the city - you can always take it later. Hence, the decision criteria for expanding is easy - can the city contribute short-term to more commerce and give resources (either that I don't have or that I can trade) alternatively, does it deny the AI of a strategic resource that weakens it militarily?

/Andreas

Can we take out of the following as a starting point:

AI;
Units (produce them only as happiness fillers);
...

We can add AI and units back later when we get some quantitative results first.

You mentioned 70%, 80%. That are very good numbers and most people are using numbers between 60% - 90% as general guidline. My point is, can we

(1) Justify 60% - 90% are indeed good? (maybe we are way off; maybe at begining, 0% is best, );
(2) Can we narrow it down a bit, to say 75%?

 

[cv431410]

The answer is 42%.

What was the question again?

Now, that is an answer.

Whether is correct or not, we will need to see your justification.

If you can show something like this:

Option ...... 100 turns ...... 200 turns
0% ............ 500 ............... ?
10% ............... 510
20% ............... 520
30% ............... 530
40% ............... 540
50%............... 530
60%............... 520
70% ............... 510
80% ............... 500
90%............... 500
100% ............... 500

then I will take your answer.

 

[dime]

We can add AI and units back later when we get some quantitative results first.

You mentioned 70%, 80%. That are very good numbers and most people are using numbers between 60% - 90% as general guidline. My point is, can we

(1) Justify 60% - 90% are indeed good? (maybe we are way off; maybe at begining, 0% is best, );
(2) Can we narrow it down a bit, to say 75%?

I think you are missing my point completely... WHY DO YOU NEED A QUANTITATIVE MEASURE??? Outside of you being a physics student and obsessed with numbers I just don't get it... I mean, you could get a quantitative number, but it would only ever be applicable to that specific scenario, which would btw. be a brutal bore to play 50 turns of, let alone 500... the correct quantitative measure is highly scenario dependant... you could run that test and for all I know it gives you the result of 81%... so now what? Will you load a real game and try to be as close as possible to 81%, believing that will give you the optimum empire city growth rate? Or, will you sit back, feel good about having proven the elusive number, the answer to all questions?


The point I though I made is that 90% will be correct in some scenarios, 60% will be correct in some and <20% will be correct if you are focusing on a specialist economy and doing it correctly... the science %-rate is USELESS as a stand-alone measurement for whether or not you expand at an optimal rate as it depends on both the circumstances and the approach you take... you can go for a cottage spam strategy with a financial leader which would suggest 70%+ science rate is preferential or a specialist economy focus with a philo leader, when beyond writing and caste system the science rate if correctly played makes very little difference for your research rate. The suggested test takes a ****-load of time and will with certainty produce inconclusve results which will not be applicable to any real-game situation...

/Andreas

 

[cv431410]

I think you are missing my point completely... WHY DO YOU NEED A QUANTITATIVE MEASURE??? ...
/Andreas

Ok, I got it, you will play by feelings. That's good for you: a simple life is better.

I have been playing a game with my friend for 100 turn only in custom_cont, one cont each, no Barb. Because it is 100 turns, there is no war either. (It is just a game, no need to comment on a particular setting. If you insist to ask, the point is find out who is the overall best builder for the first 100 turns. satisfied?) If they get A SCORE OF 1001 and you get 1000, they win.

The score is determined by population (50%), land (25%), tech (15%) and wonders (10%). Early expansion is, therefore, extremly important.

Tweaking for optimal growth, therefore, is important for me in my competition. Playing good means losing the game; playing optimally is the only way to win my game. You are not in this situation so you do not have this requirement, good for you.

... run that test and for all I know it gives you the result of 81%... so now what? Will you load a real game and try to be as close as possible to 81%, believing that will give you the optimum empire city growth rate? Or, will you sit back, feel good about having proven the elusive number, the answer to all questions?

/Andreas

If someone can prove 81% is a optimal number for one particular situation, then that serves a very good guidance for other situations. For example, when situation changes, 40% is probably not a good number.

...
... the science %-rate is USELESS as a stand-alone measurement for whether or not you expand at an optimal rate as it depends on both the circumstances and the approach you take...

In the early turns, the science rate is closed linked to the expansion speed. As I said earlier, there are more complicated measures, but I need a starting point. Once a framework is set up, it can be fine turned later.

 

[miller4242]

The answer is 42%.

What was the question again?

Good One,Dave!!!!

 

[pigswill]

Firstly I would suggest that you use total beakers with break even finances(tax/trade equals expenses) as the measure of an effective economy.
Second; there's going to be no hard and fast rule; you could have a map with several gold resources when it may be optimal just to run gold-mines the entire game; you could have another map where you run riverside cottages and athird game where you run scientists using food resources.
Third: a closed 100 turn game is going to play totally different to a 'normal' civ game and what works for the 100 turn game won't help at all in a normal game and vica versa.
I'm afraid to say that I think its unlikely that you'll find that many volunteers to run the test games given those parameters; but you never know, there's a lot more people that read the forum than post regularly so you could be in luck.

 

[Stolen Rutters]

Total beakers is the right number. Percent doesn't mean much. 80% with one city generates the same research at 40% with two identical cities. Granted the cost of maintenance is different but you will research exactly as fast.

My rule of thumb has me counting a worked gold mine as one extra city paid for, iron as another city paid for, a silver mine as another one, three hamlets as a fourth. I use the capital or the designated commerce city pushing all my early research and develop the rest for military or more marginal commerce growth.

Build Library +25% , work two scientists for +6 and +6 GPP, pop Great scientist 18 turns later (normal speed), build academy +50% . Once I power tech to currency and code of laws, I can suddenly afford six or seven more cities to help me keep pace.

This can happen with or without knowing people. I can pick up techs with smart trades but an isolated start runs about the same. I just have to build those extra cities instead of capturing them. I will beelines to certain techs to best use my terrrain.

 

[dime]

Ok, I got it, you will play by feelings. That's good for you: a simple life is better.

I have been playing a game with my friend for 100 turn only in custom_cont, one cont each, no Barb. Because it is 100 turns, there is no war either. (It is just a game, no need to comment on a particular setting. If you insist to ask, the point is find out who is the overall best builder for the first 100 turns. satisfied?) If they get A SCORE OF 1001 and you get 1000, they win.

The score is determined by population (50%), land (25%), tech (15%) and wonders (10%). Early expansion is, therefore, extremly important.

Tweaking for optimal growth, therefore, is important for me in my competition. Playing good means losing the game; playing optimally is the only way to win my game. You are not in this situation so you do not have this requirement, good for you.

Well, now, that is something completely different. This piece of information is vital... now I know that;
a) you are not a nut-job that wants irrelevant information
b) you have explained what is important in the game

I won't run the suggested test since again I believe the result would not be useful if I don't play the exact same MAP (for the specific reason of e.g. different abundance of gold etc.) but I would give some suggestions...

I assume it is 100 turns on NORMAL speed... this means you get a bit into the tech tree, but not super deep. Now, considering how much emphasis is given to population in the score and how little is given to tech and considering you are focusing on the early game, these are my suggestions:

a) if having a choice, pick a leader with one or preferably two of the traits Philosophical - since you should run SE
Charismatic - since most likely happiness will be major growth constraint
Imperialistic - faster getting settlers out
Expansive - faster getting workers out + health bonus
Early GPs to lightbulb tech could be a real boon here, as that would enable you to very quickly get e.g. CoL or something similar (for courthouses). Charismatic, however, would let all your cities be one size larger which can in itself be a difference maker

b) run specialist economy. With the short time span, cottages won't come into their own. So, you want to farm everything and restrictively use the whip when at happiness/health cap. Also, it will ensure that you at every given time maximize your population. Running scientists specialists will in the early game give a higher research rate than early cottages anyways, and they will also give early GP-points. Given the circumstances, you should definitely use all GPs for lightbulbing (wonders give comparably little points it appears, and the effects will be limited to a short time period) tech and hence you will want to focus on only getting engineers and scientists if possible...

c) You have three principal choices for getting a great start... I would lean towards b-lining for writing (to get libraries and hence scientists on line) and then go after the worker techs. Potential other approaches would be b-lining for Priesthood, building Oracle while researching writing and getting CoL from the Oracle... this would get critical techs for supporting rapid expansion early (e.g. courthouses, libraries, caste system to run unlimited scientists)... of course as soon as you have several size 3-4 cities that can run scientistst up, you also have solid research... There is a potential to do a slingshot for CS, but I assume you will play multiplayer within the same game, so unless he decides to completely forego the Oracle I guess there is little chance of this happening!?

d) Wonders should be considered only from a points perspective with few exceptions... Pyramids would be helpful for specialist economy, but comes on line too late I think... I do play on Epic most of the time, tho, so can't really comment on how many turns it takes to build on normal... if we're talking 15 turns, then fine... otherwise not. The two wonders you may want to prioritize for usefulness are Stonehenge (since it will push your borders, and hence your land area) and Oracle if it can be gotten early enough to yield a good growth tech (CoL... potentially Mathematics)... HOWEVER, given the points criteria, it would be advisable to set up an early production city that pumps out wonders, prioritizing the cheaper ones first

e) Again, not sure how deep into the tech tree you get in 100 turns on normal BUT I would say that you probably aren't going to research that far... what you really need are:
- Worker techs
- Pottery for granary
- 1 religion would be nice - synergies with going for early Oracle... reason to push the happy cap
- Priesthood for temple
- Writing for library
- Mathematics for acqueducts
- Currency for markets (happy + $$$)
- Calender for happy resources
- CoL for courthouses
- Civil service (for chaining farms and bureaucracy)
- Monarchy for HR (happy)
Note that this completely foregoes Alphabet (noone to trade with, GL online too late - only useful if wanting Drama for theatres), All the techs at the lower end of the tree past bronze working, Meditation, Monotheism (Org.rel will only be costly and you won't focus on building a lot of infrastructure), HR, Sailing (potentially interesting if you want Lighthouse for many sea resources)

Hope this is somewhat helpful

/Andreas

 

[dime]

And of course, which I didn't mention, expand like crazy. If you have no constraints from barbs, opponents, just churn out those settlers and workers quickly... you don't need alot of research to be productive and since research isn't valued that highly (whereas pop and land area is) get those settlers out ASAP and settle everything... as your cities grow, your economy will improve and there is really little to lose by being behind in tech if you can manage to trade it for land/pop - bigger landarea will also mean more resources which will drive up happiness/health caps... just make sure:

- that you let your settler/worker producing cities grow to the happiness cap
- that you have enough workers to especially hook up resources but also farm, farm, farm
- that every new city gets a granary (for pop growth) and a culture producing building (for land area growth) ASAP...

Another thought - given that landarea is relatively important and the risk associated with falling behind in science is limited, one potential exploit would be to pick a creative leader and then space your cities a bit further apart than in a normal game. That would definitely give you an advantage in getting a big land-area at the expense of income, as shortly the land-area between cities would be filled up... so maybe have 6 tiles between cities rather than 4 (assuming total available land isn't a limiting factor).

/Andreas

 

[cv431410]

Well, now, that is something completely different. This piece of information is vital... now I know that;
a) you are not a nut-job that wants irrelevant information
b) you have explained what is important in the game

I won't run the suggested test since again I believe the result would not be useful if I don't play the exact same MAP (for the specific reason of e.g. different abundance of gold etc.) but I would give some suggestions...

I assume it is 100 turns on NORMAL speed... this means you get a bit into the tech tree, but not super deep. Now, considering how much emphasis is given to population in the score and how little is given to tech and considering you are focusing on the early game, these are my suggestions:

Speed - fast
Era = ancient or classic.

If you start classic, that's more than enough to get to liberalism. 80 turns will be enough to get to liberalism.

a) if having a choice, pick a leader with one or preferably two of the traits Philosophical - since you should run SE
Charismatic - since most likely happiness will be major growth constraint
Imperialistic - faster getting settlers out
Expansive - faster getting workers out + health bonus
Early GPs to lightbulb tech could be a real boon here, as that would enable you to very quickly get e.g. CoL or something similar (for courthouses). Charismatic, however, would let all your cities be one size larger which can in itself be a difference maker

I would select:
Philosophical - since you should run SE
Spiritual - because the civis is changed a lot.

b) run specialist economy. With the short time span, cottages won't come into their own. So, you want to farm everything and restrictively use the whip when at happiness/health cap. Also, it will ensure that you at every given time maximize your population. Running scientists specialists will in the early game give a higher research rate than early cottages anyways, and they will also give early GP-points. Given the circumstances, you should definitely use all GPs for lightbulbing (wonders give comparably little points it appears, and the effects will be limited to a short time period) tech and hence you will want to focus on only getting engineers and scientists if possible...

Extremly good advise, I will take it.

c) You have three principal choices for getting a great start... I would lean towards b-lining for writing (to get libraries and hence scientists on line) and then go after the worker techs. Potential other approaches would be b-lining for Priesthood, building Oracle while researching writing and getting CoL from the Oracle... this would get critical techs for supporting rapid expansion early (e.g. courthouses, libraries, caste system to run unlimited scientists)... of course as soon as you have several size 3-4 cities that can run scientistst up, you also have solid research... There is a potential to do a slingshot for CS, but I assume you will play multiplayer within the same game, so unless he decides to completely forego the Oracle I guess there is little chance of this happening!?

This is the point I have been struggle with a for a long time: early growth or early wonder. I basically have decided I should go entirely for early growth and give up wonders all together.

 

[cv431410]

And of course, which I didn't mention, expand like crazy. If you have no constraints from barbs, opponents, just churn out those settlers and workers quickly... you don't need alot of research to be productive and since research isn't valued that highly (whereas pop and land area is) get those settlers out ASAP and settle everything... as your cities grow, your economy will improve and there is really little to lose by being behind in tech if you can manage to trade it for land/pop - bigger landarea will also mean more resources which will drive up happiness/health caps... just make sure:

- that you let your settler/worker producing cities grow to the happiness cap
- that you have enough workers to especially hook up resources but also farm, farm, farm
- that every new city gets a granary (for pop growth) and a culture producing building (for land area growth) ASAP...

Another thought - given that landarea is relatively important and the risk associated with falling behind in science is limited, one potential exploit would be to pick a creative leader and then space your cities a bit further apart than in a normal game. That would definitely give you an advantage in getting a big land-area at the expense of income, as shortly the land-area between cities would be filled up... so maybe have 6 tiles between cities rather than 4 (assuming total available land isn't a limiting factor).

/Andreas

That's really the heart of this thread: would you agree with me:

(1) If you research is 100%, you expand too slow?
(2) If your research is 10%, you expand too fast?

If you do, then what is the rate?
If you do not, why not?

 

[abuaftab]

One point that I picked up was about 0&#37; research. If I was to put research down to 0 and use scientist specialists would that be as good as 100% research or thereabouts? I've never thought about let alone tried this.

 

[Stolen Rutters]

A few points.

If you decide to run SE, research rate is uncoupled from tech speed, since you are not getting most of your research from your base commerce but rather your specialists, a different research pool. Your research rate at 10&#37; (or even zero) means that you are paying for a large empire (more scientists powering ever faster research). This is, of course, assuming that you are setting your research rate for the most neutral cash flow position.

The most effective Specialist economy runs representation (+3 per specialist doubles the research rate of the scientists and even turns merchants into researchers). You need Pyramids or else the other player will crush your score if he beats you to it. You have a +3 happy cap bonus for the first 4-6 cities... with the five city limit, that will really give you 12 to 15 more base happy pop than despotism depending on map size. Pop is 50% of your score and being 15 points behind on base happy cap will be rough.

100 turns from a classical start is a bit too far away to pick up constitution. If you lose the Pyramid race, you will need to use Hereditary Rule and Free Religion, or end up with most of the happy resources to make up that happy cap deficit, but the extra three science per specialist Representation will give the opponent a leg up if they are running the SE + Rep (with The Pyramid). You would have to hope they aren't powering through a Liberalism race of their own since they WILL be teching faster than you, all things being equal. You will just have to be better at using your available resources in that case.

How does religion work on a later start like classical? do you start with an ancient tech religion or are those bypassed? Maybe do a liberalism rush through the religious techs to get earlier free religion and more religions on your side if the opponent builds the pyramids first. With population size such a big part of the score, they may be doing the same.

With representation from the Pyramid, SE will dominate the early game on tech speed AND max pop. edit - and the possibility of an eary great engineer can get whoever owns The Pyramid another instant wonder.

 

[cv431410]

It is true that research % means less in SE than in CE; but wait, I am not reaching the state of SE yet.

In a 100-turn game, you want cities up ASAP, because later cities do not contribute much.

The early game is all about production of settlers and workers, say 6 cities and 8 workers. At this stage, I am not yet developed a SE. At most I will have 1 or 2 library. Even in those cities, I could not use 2 specialists because I have to grow the city.

Under the circumstance, the research % is a good measure for how fast I expand.

You should accomplish 6 cities + 8 workers in 22 turns with 0% research, in 40 turns with 60% research, or 60 turns with 80% research (I am making these numbers up). Obviously some options are better than other at turn 100.

 

[cv431410]

How does religion work on a later start like classical? do you start with an ancient tech religion or are those bypassed? Maybe do a liberalism rush through the religious techs to get earlier free religion and more religions on your side if the opponent builds the pyramids first. With population size such a big part of the score, they may be doing the same.

Change civic ASAP as the game starts; if you are lucky, you will get one. Otherwise, start with CoL.

If that fails, a GS will pop Philosophy.

If that fails, research or pop theology.

If that fails, get artists for devine right.

If that fails, you are a really bad player.