Vel's Strategy Thread, Volume II

[DeepO]

Originally posted by Velociryx
Vel, please remind me, what did MRC stands for again?

MRC = Multi-Role City. The essence of the plan revolves around the worker crews improving far more tiles in a city radius than the city will be able to use for a long time (in all probability, for much of the game). By terraforming well "in advance" of use, I'm just a couple mouse clicks away, at any given point, to totally reconfiguring any city I control.

I gathered as much. Again very similar to me, except that I don't need to shift tiles around so heavily in all but one city: my capital.

The result of which, is that in general I'm running with fewer workers.

you don't have to build libs.
Very true, but Libraries come early in the game (during my "expansion phase," at a point when the only other culture generator is the Obelisk, and while it is cheaper, it only provides half the cultural output, and no other benefits, long term. So my decision is an easy one when picking between the two...I seldom build an Obelisk at all.

You're right, however I very rarely run into that in my games. I'm heavy on spreading whatever religion I get, which means easier expansions for all cities. Further, commerce cities focus on libs early on. It doesn't take long for them to reach a second expansion... at which point most of my territory will be available, even with a cultureless barracks city in the middle of my territory. Also, barracks cities tend to lay closely to the capital, and get covered with the capital's culture anyway.

the "rounding error" is where you *get* your greater return on invested hammers. Significant enough to be noteworthy, but not enough to offset the greater aggregate increase in hammers doing it the other way.

In thinking it true, I think I've got to agree with you. You see, I was assuming that as all my commerce was in e.g. 2 cities, I would not get rounding penalties in the cities that didn't have markets.

It's difficult to explain. In general, the less roundings, the better. That means that the more commerce in a city, the less you'll lose... this is not a hammer investment where you see the benefit (that too, but there is something else I'm trying to point out). It's purely that because of rounding I will get a higher gpt rate in my empire than you.

It's not exactly correct, though, and this has changed in my understanding since last post (so thx!): it's not because I don't need to build markets in production cities (they won't get to +1 gpt, so what's the point), that rounding doesn't occur... in a way, I'm still getting hit by the same losses as you are due to roundings, even if in my case those roundings appear in markets I didn't build.

Phew.

It does assume you build markets everywhere, though, completing them everywhere at the same time I complete one of them in my commerce city. I don't think you'll be able to do that in all games

And yes, it is circumstantial. However I don't think there is a situation where you will get more gold than a specialized empire. It might be that you picked your city's specialization wrong (barracks city in between the wines and gold), but nothing else.

But there's the rub. If you're only going culture heavy in "culturally specialized cities" (perhaps 30-50% of your empire), and I'm investing culture heavy everywhere, assuming a like number of cities, then I'm at least that much "ahead" in terms of total culture, making it relatively harder to take my cities, pushing the borders back further, etc.

Why? I've got culture specializations, I've got commerce specials. Both have very decent culture. My production centers focus on wonders, they'll have even more culture. The only cities not getting a lot of culture are the barrack's cities, which typically are in land, not next to a border, and get at least some culture from religions and theatres (and e.g. HE as well).

Like I said: where it's needed, I gravitate heavily to culture. Because I don't need to build units in there, and because I don't have to prioritize markets, I will get the best culture there is (artists!). Others will be more then enough: commerce cities go to 60% defense very quickly too.

WRT units...I use the f9 summary almost every turn. I don't have to be the biggest, I just have to be somewhere in the middle and maintain active diplomacy. If I do that, the warmongers will pick on the folks with the smaller military, and I can go on my merry way.

Well, every turn would be a bit much, but we've got exactly the same playstyle here. If I would risk speeding ahead (unlikely, but still), I will change my barracks cities to infrastructure builds too. Eventually, these will get a lib in many cases, but only after all happy/health stuff.

If I am falling behind, I try to get another barracks city going. Or perhaps have one of my MRC cities (I tend to keep one or two of them, for flexibility) build a barracks and some units, before continuing on commerce infrastructure.

A new city has nothing, and as such, is ill suited to any particular task. Given that, its first goal will be to focus on food, and I will use the fast-growing population as a means to speed-build basic infrastructure in place, chopping forests I don't need that are in the vicinity to further this process along.

Big difference to me here. In some cases, I'll do the same. In others, I will go for either production, or commerce early. Some cities, especially those founded later, go food-heavy at first, but that certainly is not true per definition for city #2 and #3!

Forests: If I spot I need to remove them (e.g. to let irrigation pass later on), I tend to go for them asap. If there is something like e.g. silks on forest before calender, I tend to use them instead. Commerce cities are greatly helped by chopping, the designation of which city is going to do what is an easy guide to where I need to leave my forests intact.

Good notes re: Barracks cities. In my way of planning the Empire, the units are the Barrack's city's "profits," while the paying maintenance they generate is simply that city's "price to play."

Yup. It applies to more of these specialization: e.g. missonary factories (normally in culture or commerce cities), specialists (food or commerce cities), workers/settlers (food, sometimes barracks. Early on many cities have to provide their own worker, but I'm trying to improve that in my games), science. All of these get shared, and thus 'rented out'.

PS: Arrian, I'm finally set up to play AU further... let's hope I can get another report out tonight

DeepO

[Aeson]

I agree with 1:1:1 for F/P/C as the most fair assessment on a general level. Then disregard the ratio totally and focusing on what the city and empire need most at the time. If it's going to be quantified, it has to be properly qualified.

For instance, the choice between 1.1.1 and 2.0.1 can favor either. If the goal is a WB to turn that 2.0.1 tile into a 4.0.3 tile... then the 1.1.1 tile may be favorable. I use 0.3.0 tiles quite frequently in coastal cities at size 1. Using them for Granaries can also be the best option. (As can using Caste System Artists to expand borders.)

When I say "it's situational" that doesn't mean random. It means in every given situation there will be a different weighting on F/P/C. Whereas approaching with a static weighting and sticking to it actually is random.

-------------

Specialization is very important for efficiency. Having flexible cities is also important, but they are almost a given as specialized cities will over time gain flexibility once their primary objectives are met, as well as specialized improvements (ie. Cottages for Commerce) gaining other implications as the game progresses (ie. Universal Suffrage for production).

It's more efficient to start specialized and end up flexible IMO.

[DeepO]

Now here's a classic.... I just spent 10 pages of text trying to explain something, and Aeson comes along and explains it better in 3 paragraphs

It used to happen all the time in GS

DeepO

[Blake]

Heheh yeah it's funny when someone takes "It Depends" as "It's random", which betrays a lack of depth of understanding, a blindness to the factors on which it depends on, so to speak.

Such blindness is common and the default. An excellent example is watching out for Buddhism/Hinduism not being founded in the opening moves, indicating those techs are still open and researching them gives an excellent chance of snagging the religion, even on higher difficulties - the lesson: that trying to get an early religion isn't always about blind luck, it's partly about being opportunistic, but you must learn to see the opportunity before you can take it. I picked that stratagem up in a AU thread.

~~~

Anyway, something I was thinking about was strategies and difficulty level. At what point is a strategy "invalidated" due to not working at high difficulty level? Is a strategy invalid just because it doesn't work on Deity? The measure of validity is reliability, the strength of any strategy which relies upon getting a wonder or religion depends on how reliably it can be done.
Raise the AI cheats enough, and they'll be able to beat you to wonders/religions even if your situation is absolutely ideal for it. Say for example having extremely easy access to Stone and being Gandhi and having a heap of forest. There is no way any player imaginable should be able to beat you to the Pyramids, but could an AI if they put their mind to it?

The CS Slingshot is a very tricky strategy from this perspective. It seems to rely on the AI never making the Oracle or Code of Laws a very high priority, this makes it nearly 100% reliable in Single Player on Monarch and Easier.
However, in multiplayer there is the possibility of another player building the oracle before you research code of laws, and using it to snag code of laws, metal casting, or another low powered tech. CS Slingshot is clearly not a reliable strategy in Multiplayer, even though it is on Monarch vs AI's.

I think that a large part of the strengths of strategies comes from minimizing risk. For example if you're Gandhi with Stone, building the Pyramids is not a high-risk strategy. But if you're non-ind non-stoned it's an extremely high risk strategy, to make a serious attempt you must chop extremely aggressively, the reward may still be high, but the risk is huge. The reliability and hence validity of a strategy depends on the situation. This I think is the key to CIV greatness, knowing what powerful strategies are sure bets. In this sense, CS slingshot is perfectly valid strategy, there ARE situations in which it is a sure bet.

A strategy with high payoffs is strong if there are recognizable situations which will make it low risk. THAT is what makes a strong strategy, not that it always works, but that in select situations you can rely on it working.

Furthermore, having a large arsenal of strong situational strategies becomes essential to be a master player, you must recognize the appropriate strategy which will be high-reward low-risk for the given situation.

There ARE situations in which CS slingshot is low risk high-reward, even in multiplayer. This would be a high resource, high-research Capital (probably fish resources and gold mine), with nearby terrain being very poor for founding extra cities - the Oasis in a Desert. CS Slingshot is low-risk because you have a good chance of researching Code of Laws before someone else completes the Oracle to pick a low-powered tech. It's high reward as you have an excellent capital to be boosted by 50%.

The contrary situation is one with a poor capital, say with no river and only 1 resource in radius. CS Slingshot becomes high-risk, it'll be difficult to research and build quickly enough. It becomes low-reward as it's a 50% boost to a poor city. The correct strategy for the situation would probably be settler-first.

I think the moral of this post, is firstly that a strategy cannot be said to be bad just because it doesn’t work all the time, it can only be said to be bad if you have no way of knowing how well it will work - its success truly being determined by a flip of a coin.
The second moral is that we do have a real strategy game on our hands here, one with multiple levels of strategy.

[Heroes]

Originally posted by DeepO

I am "talking about specific situations", quite true. The situation is that you build pyramid in the early game. Otherwise the base value of a specialist is 3 gold, not 6 gold, which makes specialist much less attractive comparing to working tiles.

Well, and I try to avoid that Pyramids prerequisite. The reason is that on higher levels, it becomes hard to consider the pyramids a given... and the higher you get in level, the more hereditary rule becomes important. Representation is a lot less attractive if you have to go for constitution first, or if at the time of reaching it all your cities bump into their happiness limits, not just your largest ones.

Which is why I consider your situation (with Pyramids and rep) a lot more specific than mine (without either of the two).

About incense on desert: surely it's better to have an incense on desert than nothing, but my point is, if you run representation and can hire specialist, you gain more by leaving that dear incense empty and hiring a specialist! It might change if leader is financial (makes working tile better), or you dearly need commerce but you can only hire an engineer, or, of course, you don't run representation. But you get the reasoning.

Yes, but I got that reasoning before too. And I object it: in most cases, a incense on desert is going to be more important to you than an extra scientist. Especially early on, when you might not even be able to use a scientist (no lib, or no caste system). 6 commerce is exactly the same as what a scientist under rep will bring you (which makes it the better choice as it also gives GSP) under 100% slider.

If your slider is not at 100%, part of that 6 commerce is going into gold. Now you can say: okay, but a merchant under rep generates 3 gpt+3bpt, so is obviously better. True, but again, this requires alot of stuff: a market or caste system, plus either the pyramids or constitution (and thus rep.). So, it's my feeling that you're talking about the exceptions here, where I try to follow the basic rule, in which your situation is an exception.

In the early ages, you really don't have many choices: farm, mine, cottage, that's it. But the base tile indeed makes difference. A hill on plain gets 4 hammer with mine, a hill on desert gets only 3. So a hill/desert is quite awful to work, I would rather hire a specialist or work a sea tile. Don't you agree?

First of all: I was talking about that base values of tiles is less important for your choice than you might think. In the sense that desert and tundra are never good choices, but that grass and plains is equal in their merits, only a bit more difficuly to improve sometimes. Hills are needed because they are the best source of hammers early on and you can't mine a flat grass, but otherwise they are not important at all.

Now, if you are going to compare grass/plains to desert/tundra, it's obvious: desert is less.

That doesn't mean that desert will always mean a specialist is better, and especially in the general situation (no rep) that is true. Desert hills are so important because they appear near fp, or coast. You need production there, and not more beakers... hence working a desert hill might be absolutely needed. I found that in general, over half of my desert hills are being worked asap. They stop growth (or at least hamper it), but I can get to food in some other way (or I don't need it as I'm maxed out already), while desert hills are typically in areas where there is no production around... its relative importance went way up.

Mind you, I'm in love with specialists, but if you want to get most out of them, you need to specialize them just like you need to do everything else. A scientist in a desert city is lousy, but add a scientist to your best science city (copernicus) and you gain a lot more.

For a new city, would you like to work a 1/1/1 tile or 2/0/1 tile? Without considering pop limit and granary, I would definately pick 2/0/1. From my formalism, 2/0/1 = 7 gold, 1/1/1 = 5.5 gold. From experience, working high food tiles first makes your city grow faster, and get more total output in the long run. An more extreme example is: will you pick a 0/0/4 tile over a 2/0/1 tile for a new city? I bet you won't, because it completely stops growth. If you agree with these 2 examples, you would agree 1 food > 1 gold.

I lost you. 1/1/1: f/p/c, right? In that case, early on I'll try to have a mix of 2/1/0 and 2/0/1. However, health is going to run out quickly, and I certainly do not have to keep a +2fpt city wide... which can let me put to 2/1/0 tiles to 1/2/0 or 1/1/1 or anything else which will let me trade in a food for a commerce or a hammer. as I will always try to maximize my commerce, and not my food (food is only needed in those circumstances you want to grow), I consider commerce more important than food.

As to picking a 0/0/4 tile: yes, most definately. The extreme example of this are gold mines on desert hills: 0/3/5 once improved, or 0/3/6 near river. Desert hills, stopping all growth... and still I will prioritize these asap, and choose the rest of my improvements around them. They take priority over cows...

You sound like you will make sure you will have food in your city, and once you don't know what to pick anymore on food tiles, you go for production tiles. Once you get all the best production, you go for commerce. This is what I'm reading when you post 1 food = 1.5 hammers = 3 commerce (and not gold!). Am I right in this?

This was the Civ3 way... CIV is different. I said all of these were similar in value, meaning 1f=1h=1c. Depending on the situation, I will either pick food over commerce, or commerce over food. Early on, you might be tempted to think food is everything, but I dispute that. By focusing on commerce instead of on food, you can most of the times gain a serious advantage. As there are also other situations, I don't say 3f = 1.5h = 1c (the opposite of you), I average it out and say that the 3 types are equal. Relative to you, I put more emphasis on commerce.

DeepO

Oh, man, you are posting so much ...

It seems that I said "if A, then B", and you said "if not A, then C". Well, I generally agree your "if not A, then C" things. You know, in science, it's common to have this kind of thinking: you take some simple conditions as base, get the accurate solutions under these conditions, then add the complicated conditions, and find the differences they make. For example, to understand the motion of earth, you first consider the mass of sun as infinity, ignore all other planets, therefore earth is running an exact elliptic trajectory. Then you consider the effects of the finite masses of sun and other planets, and get all kinds of variations to this ellipse. I mean, there are many many possible situations. If you want to get the strategies under each situation, you quickly become dizzy. The best ways seems to be: first analyze an idealized (simple, but not necessarily quite common) situation, fully understand its inputs and outputs, get the reasoning, then apply the reasoning (not the concrete conclusions) to other situations.

Let me summarize my A's and B's:

1. For sake of simplicity, assume you get pyramid, you can run representation and slavery, leader is not financial or philosophical, city has granary and has no problem of happiness or health, city is growing, you can hire the proper specialist in city. Every point of these can change, but you can get the corresponding change in conclusions quickly.

2. Now the question I want to analyze is: given these assumptions, which option is best to use a citizen, work a tile, specialist, or pop rush?

3. To answer this, I need to get some numerical value for each option. Assuming the born value of each type of specialist is equal, I get 2 hammer = 3 gold, and each specialist is worth 6 gold + 3 GPP. Every specialist needs 2 food, therefore 1 food = 3 gold (+1.5 GPP).

4. Therefore come the conclusions (only under the assumptions, of course). The most important one is: try to make some improvement on any worked tile (well, an exception is oasis on desert, you can't put any improvement, but its 3/0/2 base value is good enough to work).

5. Under these assumptions, in general you want to make city grow to its happiness or health limit as fast as possible, then begin to work the low food tiles or do pop rush. Just "in general", because for example a gold mine is much better than a normal 2 food tile, according to my numerical estimation.

If you agree with my reasoning, then easily you can get the variations under different situations.

1. If you can't run representation, then a specialist is only worth 3 gold + 3 GPP, much less useful. Therefore a 0/0/6 tile (e.g., desert incense) will become good to work.

2. If leader is financial, then it's your best interest to make every tile generates at least 2 gold. This makes specialist (and pyramid) less useful too.

3. If leader is philosophical, then a specialist generates 6 GPP. To make best use of them, you would like to get pyramid at all cost.

4. If your city has no granary yet, you'd better work high hammer tiles or do pop rush first to build granary asap.

5. If your city is reaching happiness or health limit, you get the options: work high food tiles to do pop rush, or work low food tiles to balance food, or hire specialist.

6. If your city has great trouble in growth ... Emm, why did you make that city?

7. If your city can't hire specialist yet, well, begin to build a corresponding building.

8. A very concrete point: suppose your city has quite some floodplains and desert/hills, and you can do pop rush, would you like to work those desert/hills (0/3/0)? To me, 0/3/0 seems too few, I would rather work the high food tiles and transfer food into hammer. Of course, this could change if your city gets forge, factory, power, etc. (because these don't help pop rush).

9. If my capital has a gold resource, I will work it asap, because that's too much commerce. However, if the comparison is not so extreme, like 2/1/0 vs. 2/0/1, then it's hard to say that you gain sth. from emphasizing commerce. Because your palace generates 8 gold anyway, one more gold adds just 1/8, while one more hammer adds probably 1/4 or even 1/2.

At last, I don't mean any concrete strategy is the best, but I do want to find some rational way to understand and suggest strategies. If my post can stimulate people's thinking, that's good enough.

[Cort Haus]

2. If leader is financial, then it's your best interest to make every tile generates at least 2 gold. This makes specialist (and pyramid) less useful too.

3. If leader is philosophical, then a specialist generates 6 GPP. To make best use of them, you would like to get pyramid at all cost.

These points highlight a dilemma I find with playing Elizabeth. Cottages to use Financial, or food & specialists to use Philosophical - assuming terrain allows either approach? This impacts early research choices if a beeline requires that some worker techs must be deferred.

As I've hardly ever built the Pyramids, strategies that revolve around their ownership as a pre-requisite seem a bit irrelevant to me. I prefer Oracle strats because almost any civ on any start on most levels has a realistic chance of getting that wonder vs the AI. Pyramids is very situation-specific, I find.

[DeepO]

Originally posted by Blake
Such blindness is common and the default. An excellent example is watching out for Buddhism/Hinduism not being founded in the opening moves, indicating those techs are still open and researching them gives an excellent chance of snagging the religion, even on higher difficulties - the lesson: that trying to get an early religion isn't always about blind luck, it's partly about being opportunistic, but you must learn to see the opportunity before you can take it. I picked that stratagem up in a AU thread.

I posted something along those lines, where me going for budhism (while hinduism would have seemed the better option) meant I was almost certain I was going to get one of the two religions on Monarch, without me starting with mysticism. Reading the info messages can be very important: you see when AIs are builderish or warrish from their civic choices. Wonders completing give excellent reasons to go to war, etc.

Aeson is the king of opportunity, though. He managed a CS slingshot on deity, simply because he got a tech gift and spotted the opportuniy to use it well.

A strategy with high payoffs is strong if there are recognizable situations which will make it low risk. THAT is what makes a strong strategy, not that it always works, but that in select situations you can rely on it working.

Excellent idea, and I agree in full.

Strategies being reliable in recognizable situations is important, and that you can recognise those situations, too. Sometimes, this means you need to go scouting farther than's safe. Sometimes you ned to go for certain techs to spot e.g. resources, before you really need them. Basic strategy is important, but the ability to assess your situation is as important

It's true that risk increases for a couple of strategies on higher levels, some will be outright impossible, and others gain importance. Specialization is a classic here: the higher up in level, the least you can afford to spil. More efficiency --> more specialization.

CS slingshot is the opposite: the lower in level, the surer it is. That doesn't mean you can't use it on higher levels, sometimes you can offset the AI bonusses (e.g. by going commerce heavy in your capital, with 2 scientists and academy, you will outresearch Emperor AIs. But you'll need some health resource to have 2 comfortable scientists, which means you can use a worker too... all things that are expensive to build when you want to have the Oracle).

Sometimes, this means that by chosing to focus on one path, your neglecting the other. Tech vs military, food vs production... unbalancing can get you in front in one area, but you'll need some way of getting even.

Classic examples:
- use military to gain land without growin into it.
- use commercial power to gain tech, leading to more advanced (but fewer) units
- use commercial power to upgrade instead of build
- use growth to spam cities, denying others to settle there (barbs!). Less military needed in some cases
- the whole horizontal vs vertical growth thing: horizontal growth leads to better food and production, but less commerce. You can offset this by e.g. focusing on harbors and CHs. And you'll gain loads of resources, which you might be able to sell: no health needed as all your empire is small, so let's sell that resource for as much as possible. Once needed, cancel the deal.

DeepO

[DeepO]

Originally posted by Heroes
It seems that I said "if A, then B", and you said "if not A, then C".

Not really. I'm saying: default to C, but perhaps in case of A you can do B. You don't have to, though.

You're still working under the assumption that pyramids/representation is the ideal situation, and in case you can't get it, you need to change your strategy thinking around. That might be true on Noble, but quickly stops being true above that level... like I said, hereditary rule is my current favourite civic early on. And I don't need the pyramids to get it: by the time you can build it, I've researched monarchy myself.

1. For sake of simplicity, assume you get pyramid, you can run representation and slavery, leader is not financial or philosophical, city has granary and has no problem of happiness or health, city is growing, you can hire the proper specialist in city. Every point of these can change, but you can get the corresponding change in conclusions quickly.

"for sake of simplicity"
Okay, for sake of simplicity, assume none of these are required, and certainly not a given

Really, you're trying to extrapolate a specific situation to the general rule, and if you are a scientist (I know more then the basics about research, I've fairly recently stopped my PhD as I couldn't bear it anymore), you should know that is asking for disaster.

3. To answer this, I need to get some numerical value for each option. Assuming the born value of each type of specialist is equal, I get 2 hammer = 3 gold, and each specialist is worth 6 gold + 3 GPP. Every specialist needs 2 food, therefore 1 food = 3 gold (+1.5 GPP).

Have you ever considered that you're looking at specialists, who indeed have a certain ratio assigned to it, but these were balanced off the ratios of city improvements?
In improvements, food is the easiest, production second. Commerce is the most difficult to get. So, for specialists, food is less abundant, hammer a bit more, and nearly every one generates a decent amount of commerce (gold or beakers).

Add to this that specialists only generate beakers or gold, and lack the flexibility of commerce.

So: by looking at only one appearance of those f/p/c ratios, you're distorting the picture. Moreover, you do so in a very specific situation, and try to extrapolate it to thw whole.

That will never work, I'm afraid.

5. Under these assumptions, in general you want to make city grow to its happiness or health limit as fast as possible, then begin to work the low food tiles or do pop rush.

No, that's another specific situation. Consider this: early on, the race for techs is very important, take the religions as example. In that context, you want instant commerce, and not more commerce later on, after you've grown.

Just another example of where your situation is not ideal anymore, and again is too specific.

1. If you can't run representation, then a specialist is only worth 3 gold + 3 GPP, much less useful. Therefore a 0/0/6 tile (e.g., desert incense) will become good to work.

0/0/6 tiles are already quite well. Look at what both I and Aeson are saying: the choice between a 2/1/0 or 2/0/1 and a 1/1/1 is so important early on, that in many cases we'll pick the extra commerce. That's a 1 food for 1 commerce exchange, not even a 2 food for 6 commerce exchange.

2. If leader is financial, then it's your best interest to make every tile generates at least 2 gold. This makes specialist (and pyramid) less useful too.

I've got more chance of being FIN than I have running a pyramid fueled rep.

3. If leader is philosophical, then a specialist generates 6 GPP. To make best use of them, you would like to get pyramid at all cost.

Sure. This is the other inbalance. rep. is important to PHI leaders... it is not a necessity, though, I always run a lot of specialists, whether I'm PHI or not.

4. If your city has no granary yet, you'd better work high hammer tiles or do pop rush first to build granary asap.

Nope. Many cases you go the opposite way and freeze your city. Also, the use of a granary is highly dependent on food availability: in many cases I build it for its health bonus, and not its growth bonus.

8. A very concrete point: suppose your city has quite some floodplains and desert/hills, and you can do pop rush, would you like to work those desert/hills (0/3/0)? To me, 0/3/0 seems too few, I would rather work the high food tiles and transfer food into hammer. Of course, this could change if your city gets forge, factory, power, etc. (because these don't help pop rush).

how many times have you been able to poprush in desert towns? Slavery quickly gets turned off once caste system appears: it's no good having a fp city without the possibility to use specialists.

9. If my capital has a gold resource, I will work it asap, because that's too much commerce. However, if the comparison is not so extreme, like 2/1/0 vs. 2/0/1, then it's hard to say that you gain sth. from emphasizing commerce. Because your palace generates 8 gold anyway, one more gold adds just 1/8, while one more hammer adds probably 1/4 or even 1/2.

If everyone is getting 8 bpt, and you're gaining 9bpt, you will reach a tech first. Being first is all that counts.

DeepO

[Heroes]

Originally posted by DeepO


Not really. I'm saying: default to C, but perhaps in case of A you can do B. You don't have to, though.

You're still working under the assumption that pyramids/representation is the ideal situation, and in case you can't get it, you need to change your strategy thinking around. That might be true on Noble, but quickly stops being true above that level... like I said, hereditary rule is my current favourite civic early on. And I don't need the pyramids to get it: by the time you can build it, I've researched monarchy myself.

"for sake of simplicity"
Okay, for sake of simplicity, assume none of these are required, and certainly not a given

Really, you're trying to extrapolate a specific situation to the general rule, and if you are a scientist (I know more then the basics about research, I've fairly recently stopped my PhD as I couldn't bear it anymore), you should know that is asking for disaster.


Have you ever considered that you're looking at specialists, who indeed have a certain ratio assigned to it, but these were balanced off the ratios of city improvements?
In improvements, food is the easiest, production second. Commerce is the most difficult to get. So, for specialists, food is less abundant, hammer a bit more, and nearly every one generates a decent amount of commerce (gold or beakers).

Add to this that specialists only generate beakers or gold, and lack the flexibility of commerce.

So: by looking at only one appearance of those f/p/c ratios, you're distorting the picture. Moreover, you do so in a very specific situation, and try to extrapolate it to thw whole.

That will never work, I'm afraid.


No, that's another specific situation. Consider this: early on, the race for techs is very important, take the religions as example. In that context, you want instant commerce, and not more commerce later on, after you've grown.

Just another example of where your situation is not ideal anymore, and again is too specific.


0/0/6 tiles are already quite well. Look at what both I and Aeson are saying: the choice between a 2/1/0 or 2/0/1 and a 1/1/1 is so important early on, that in many cases we'll pick the extra commerce. That's a 1 food for 1 commerce exchange, not even a 2 food for 6 commerce exchange.

2. If leader is financial, then it's your best interest to make every tile generates at least 2 gold. This makes specialist (and pyramid) less useful too.

I've got more chance of being FIN than I have running a pyramid fueled rep.

3. If leader is philosophical, then a specialist generates 6 GPP. To make best use of them, you would like to get pyramid at all cost.

Sure. This is the other inbalance. rep. is important to PHI leaders... it is not a necessity, though, I always run a lot of specialists, whether I'm PHI or not.

4. If your city has no granary yet, you'd better work high hammer tiles or do pop rush first to build granary asap.

Nope. Many cases you go the opposite way and freeze your city. Also, the use of a granary is highly dependent on food availability: in many cases I build it for its health bonus, and not its growth bonus.

8. A very concrete point: suppose your city has quite some floodplains and desert/hills, and you can do pop rush, would you like to work those desert/hills (0/3/0)? To me, 0/3/0 seems too few, I would rather work the high food tiles and transfer food into hammer. Of course, this could change if your city gets forge, factory, power, etc. (because these don't help pop rush).

how many times have you been able to poprush in desert towns? Slavery quickly gets turned off once caste system appears: it's no good having a fp city without the possibility to use specialists.

9. If my capital has a gold resource, I will work it asap, because that's too much commerce. However, if the comparison is not so extreme, like 2/1/0 vs. 2/0/1, then it's hard to say that you gain sth. from emphasizing commerce. Because your palace generates 8 gold anyway, one more gold adds just 1/8, while one more hammer adds probably 1/4 or even 1/2.

If everyone is getting 8 bpt, and you're gaining 9bpt, you will reach a tech first. Being first is all that counts.

DeepO

DeepO,

I say some assumptions are idealized, not meaning they are common or should be achieved at all cost, but just they are simple to understand. There are so many possible situations that I don't think any strategy is "default". The best you can have is a line of thinking, to quickly get new method for new situation. These assumptions are seldom all true, absolutely. And I don't mean you should try to realize these assumptions. All I say is that it's an EXAMPLE of rationalize a strategy, and what one really needs is the way of rationalization, not any concrete strategy.

Excuse me, I don't understand your point about "In improvements, food is the easiest, production second. Commerce is the most difficult to get. So, for specialists, food is less abundant, hammer a bit more, and nearly every one generates a decent amount of commerce". The best guess I have is: it's easiest to get food by working tile, production 2nd, commerce hardest; to compensate these, specialist generates most commerce, less production, and zero food. So, what derivation do you draw? Sorry, I still don't see how do you determine when to work a tile, when to hire a specialist, and when to do pop rush. Could you please be more concrete on this? Thanks.

About slavery and caste system. Some time ago I thought caste is great, but now I think it's not so great. First, it's medium upkeep, higher than slavery's low upkeep. Second, I don't need that many specialists anyway, most citizens will just work tiles. Third, losing the flexibility of pop rush is ... sad. Therefore, I would like to run slavery at least till emancipation becomes available. How to deal with a high food city? Under slavery you can pop rush some buildings (lib, forge, market, etc.) to hire specialists, which looks sweet to me.

I don't object to the idea of emphasizing commerce. The only thing I need is to have a rational and concrete outline of when and how to do it: work tile? which improvement? specialist? pop rush? caste system? Every action could find some situation which makes it all right, but to be efficient, we'd better know the principle.

What does "fishing" say? Give a man a fish, you feed him for a day; teach a man fishing, you feed him for a life time.

[Velociryx]

It's more efficient to start specialized and end up flexible IMO.

I would agree with this statement. And if you're in a position of having to make every single hammer count in order to survive, then this IS the approach to take.

I do not believe that makes it the default choice in every situation, because as soon as you're away from "the survival line" and have some breathing room....yadda yadda yadda, as above....LOL

-=Vel=-

[Velociryx]

If everyone is getting 8 bpt, and you're gaining 9bpt, you will reach a tech first. Being first is all that counts.

With regards to tech, I would say this is true only in a few key situations.

I don't have to be "first" to research Priesthood, for example, to get the Oracle. I just have to be first to finish the Oracle....which takes production, not commerce. I don't even have to be the first IN production on the f9 screen to get it, if I'm willing to use pop and chop to get me across the line.

This is why I do not value commerce as much as food/production in the early game. There's just not that many compelling advantages to being "first" to research most techs. A few, sure....most obvious being those that give you a free Great Leader, or a free tech....in those instances, being first is of course, of paramount importance. But aside from founding a religion early on, there really aren't too many important "I MUST BE FIRST" techs in the early tree. What's interesting here, is that you don't even have to found a religion early on for it to be effective....let someone else found it, let them spend a GP on creating the temple, and then go take it from them. In that instance, being first did not help them one whit, except for the fact that it focused your first attack.....which, come to think of it, still didn't help THEM....

-=Vel=-

[Riotamus]

Originally posted by DeepO

In thinking it true, I think I've got to agree with you. You see, I was assuming that as all my commerce was in e.g. 2 cities, I would not get rounding penalties in the cities that didn't have markets.

It's difficult to explain. In general, the less roundings, the better. That means that the more commerce in a city, the less you'll lose... this is not a hammer investment where you see the benefit (that too, but there is something else I'm trying to point out). It's purely that because of rounding I will get a higher gpt rate in my empire than you.

Actually, are they roundings or truncations (i.e. is 1.6 treated 2 or 1)? In the former case there should be no net gains or losses regardless of the distribution scheme.
Or maybe the game actually carries out the calculations at full precision, and then distributes the results over the cities (rounding upward or downward individually in each city)?

[DeepO]

Originally posted by Heroes
I say some assumptions are idealized, not meaning they are common or should be achieved at all cost, but just they are simple to understand. There are so many possible situations that I don't think any strategy is "default". The best you can have is a line of thinking, to quickly get new method for new situation. These assumptions are seldom all true, absolutely. And I don't mean you should try to realize these assumptions. All I say is that it's an EXAMPLE of rationalize a strategy, and what one really needs is the way of rationalization, not any concrete strategy.

First of all, please don't get me wrong: I'm not saying your approach is incorrect. Examples you know, and have played through are very important to judge a certain situation. But we've got a different vision here: I try to have some kind of average situation from which to start, and for my feeling you're starting to specified. Drawing conclusions from that is dangerous, and that's all I want to point out.

There is a default strategy, though. At the start, all situations are similar. Early on, you won't have the pyramids (yet), nor e.g. markets. In that early situation, food is as important as production or commerce...

Later on, when you gain certain techs/buildings/civics, and when the game circumstances change, the relative importance shifts. You correctly describe such a situation: in your situation you will find food more important than commerce, as you aim for that ratio. People who, OTOH, aim for a higher commerce importance (because they have modifiers on production and food, thus don't focus on commerce modifiers) have opposite ratios. That's why I'm saying extrapolation is dangerous.

Take the example of someone aiming for early war. He will set his paths to that, and will focus on production. Growth is unimportant: you will get he extra food from taking over cities, instead of growing on your own. In such a situation, he would e.g. say the ratio is 1:3:1 (f:c).

Now, point out to him, that in some situations commerce is more important to production, as you don't need a lot of units if they are advanced. He says: "okay, that's true, so the ratio in those games is 1:3:3". This is the danger: because he is starting from his specific situation, he doesn't realize that a 2:1:3 situation might be a better way of prioritizing his assets.

I've bumped into this thinking myself in CIV too many times: I'm a builder at heart, and a MMer (but a lazy one). Because I start out with this mindset, it took me a long while to better play in warfare. Even if I set out to conquer, I was still too builderish. I didn't emphasize enough that food is totally unimportant early on in warfare, but commerce might be a bigger aspect of it than production. So, I put my ratio at 2:3:1, where it really needed to be 1:2:3 (or 1:2:2).

Now, it's easier for me to think it through: because I played nearly only test games (forcing me in new directions each time), and because I forced myself to win every type of game on each level before going up, I do have some kind of default behaviour from which to start. And I modify the default based on what type of game I play (diplo-war-build-religion-tech). It's easier to change your thinking pattern around, IMHO. And adaptability is needed if you want to do well in CIV.

The best guess I have is: it's easiest to get food by working tile, production 2nd, commerce hardest; to compensate these, specialist generates most commerce, less production, and zero food. So, what derivation do you draw?

You're correct in your guess, that's exactly want I wanted to say.

Derivation: For 'default' behaviour, you try to average out both extremes: the specialists 'ratio' is directly opposite to the tile ratio, so the default ratio is somewhere close to 1:1:1. Going tile heavy, or going specialist heavy is what really makes the difference, but that's a choice depending on situation, and one you've got control over.

Going specialist heavy, is just as picking what to research or which civics you use one of the way of tweaking the relative importance to something you can use.

1. Assume you're at war.
2. You want more production, any production, as you need units now.
3. So, you put a higher importance on hammers: your ratio shifts from e.g. 1:1:1 to 1:2:1.

So, you want to squeeze out every drop of hammers you can. For this, you use a couple of mechanics

1. work more hills than farms (and other improvements)
2. change to Org Rel for more production to buildings
3. build forges
4. use priests or engineer specialists, even if not fully grown yet
5. etc.

Once you are doing all these things, your need diminishes, and you can put more emphasis on other things. Maybe you need more growth as a secundary priority. Like this, you try to balance it out again, and once you achieve such balance, you will have a healthy empire. (and balance means that roughly, a 1:1:1 ratio is ideal and you want to reach that taking into account as much of the circumstancesas possible)

You will see this happening in all games, and it ties in nicely with the specialization discussion. CIV is a system, one with clearly ideal situations (e.g. fast in tech, large, high gpt, etc.). Also, CIV has some lowest, least desirable situations too: you being small, undefensless, losing money, behind in tech, or dead.

As a scientist, you might call that a ground energy, or a maximum cost, or a stable state, or whatever... And the better situations become equilibriums, higher states, lower cost, or any terminology you'd like to use.

In all these terms, one thing is the same, though. In order to get higher in a system, you typically create a large imbalance. Then, you trade in some of that imbalance to something you've neglected, thus creating another stable, but better situation.

Apply this to CIV: Focus on growth early on, and you will lack in other areas, especially commerce (but also military/production). however, by playing a bit further in that less than ideal situation, you might start to turn a bit of that disadvantage in other areas into an advantage as a whole. By using the imbalance, you've leaped ahead farther than you can by simply 'growing' into the new situation.

So extreme example: you start out, and the first 3 builds are settlers. Maybe a worker, if it let's you get to 4 cities faster. You totally lack in commerce, and are probably behind by now. You have no military either, and are each turn fearing to run into some barbs with your unescorted settler. However, get those 4 cities up, and you can quickly turn in a bit of that huge advantage in one area into an overall advantage: by setting all 4 cities to units asap, you can build up a larger army then with just 2 cities. By setting them all to commerce, your tech pace will quickly pick up. You will lose growth by focusing on production or tech, but growth is less important: you're already way ahead in that area.

The trick is, to find the right imbalances, those that can bump you up a couple levels in one mighty jolt. The CS slingshot is a typical example here: you will forego on food and production early on, but by getting an extreme advantage in tech, you can quickly turn part of that advantage into a growth and production advantage too.

Sorry, I still don't see how do you determine when to work a tile, when to hire a specialist, and when to do pop rush. Could you please be more concrete on this? Thanks.

the short answer: it depends

First of all, I will set these goals on an empire level, my cities are only a local reflection of my empire plan. Specialists certainly: you can't run 15 efficient GP pumps, so focus on 2 or 3. In there, specialists are always the highest priority, even if it wouldn't be ideal on a city level (thus I will trade production for specialists quite easily)

Secondly, within your goals, you look at a city level.

- Your empire dictates specialists, which means you've got to run high food? Okay. Work farms.
- No production in such a high food city? Chop or pop rush, whatever is possible. Forests are only important for their health bonus in GP pumps, if taken to the extreme).
- Perhaps you can get some of that lost production from priests or engineers? use them, if it doesn't clash with your empire objectives.
- etc.

Like this, you create a priority based system, and you shouldn't get into trouble picking what to pick. As long as you don't get blinded by the city level: it really is your empire that decides whether a pop rush is warranted. But the moment it helps the empire as a whole, it doesn't matter anymore if it's very, or just normal important: you do it anyway.

Specialists the same, btw. Running a scientist instead of a cottage is simply a matter of circumstances. The moment you decide that science is more important than growth (as you need science now in your empire, and can't afford to invest in the future), any way which gives you most science is best.

A good advice here is to try to play agame, not setting any tiles yourself, just set the governor options. True, you can be better than the governor, but sometimes he will come up with solutions you haven't thought about.

E.g. in Tundra fisherman villages, you might say: "okay, I need to make a commerce city out of this one, but I have no production whatsoever. Let's put the governor to hammers, set a few build options, move a unit in and forget about the city for the next 50 turns."

The governor is actually quite good in optimizing for certain uses: in this case, it would most likely be using a citizen (you know, the non-specialist slacker you normally try to avoid at all costs), giving you 1 hpt. It steals a 2/3/0 coast for that one hammer, but it is the right thing to do. You determined at a higher level the city has a production problem, you made your choices, and now you see your buildings coming double as fast... but your city is growing very slowly. That's okay, you can catch this up later by setting the governor to food once you get some infrastructure up.

About slavery and caste system. Some time ago I thought caste is great, but now I think it's not so great. First, it's medium upkeep, higher than slavery's low upkeep. Second, I don't need that many specialists anyway, most citizens will just work tiles. Third, losing the flexibility of pop rush is ... sad. Therefore, I would like to run slavery at least till emancipation becomes available. How to deal with a high food city? Under slavery you can pop rush some buildings (lib, forge, market, etc.) to hire specialists, which looks sweet to me.

The main reason why I don't like slavery that late is simple: I don't have excess population. It's a matter of playstyle, and as said above, my next game will be a slavery game, just to get some better insight in it.

DeepO

[DeepO]

Originally posted by Velociryx
If everyone is getting 8 bpt, and you're gaining 9bpt, you will reach a tech first. Being first is all that counts.

With regards to tech, I would say this is true only in a few key situations.

I don't have to be "first" to research Priesthood, for example, to get the Oracle. I just have to be first to finish the Oracle....which takes production, not commerce. I don't even have to be the first IN production on the f9 screen to get it, if I'm willing to use pop and chop to get me across the line.

I can pop and chop too, but can do so earlier if I get to the tech faster.

For tech, it's true for over half the techs, and the others generally lead to one of the 'asap' techs, so they basically can't wait either.

It's not just traditional key techs, but stuff like getting a worker tech asap so your worker can start on it immediately, or writing so you can start asap on that lib that your strategy relies on. Or units: upgrade money, and a 5-turn advantage over a civ is all you need.

9bpt is a 12% increase over 8bpt... for the first 50 turns, this gives me a turn advantage of 6.67 turns. That's a whole tech, in that short period of time. At the cost of e.g. 3 warriors (assuming 1 less production for 1 more commerce). If you can get as much advantage out of those 6.67 turns, you should be able to offset your 3-warriors disadvantage easily, still remaining ahead in tech (but not by that much anymore).

DeepO

[DeepO]

Originally posted by Riotamus


Actually, are they roundings or truncations (i.e. is 1.6 treated 2 or 1)? In the former case there should be no net gains or losses regardless of the distribution scheme.
Or maybe the game actually carries out the calculations at full precision, and then distributes the results over the cities (rounding upward or downward individually in each city)?

They're truncations. Sadly, this is one of the areas where MM still can give quite a bonus if you want to go for the tedium: by MMing so you have at least truncations as possible, you will be the most efficient possible.

This is one of the few areas where the current game is worse than Civ 3: in there, whatever you would lose by e.g. the slider was distributed over the other coices. (production was different). Not so in CIV. But perhaps we'll get there one day

DeepO