Vel's Strategy Thread

[Arrian]

Methinks it's tied pretty heavily to difficulty level. Many of us are still learning the game and thus haven't gone up the ladder high enough to really get hit with tough strategic choices yet (like do I pay for this big army, or do I go peacenik and risk getting crushed by the bad man over there?). I plan to go up to Prince for good soon (I played one aborted game there so far), but I don't think I'll run into some of this stuff until I'm up on Monarch (which, far from my CivIII comfort level, sounds vaguely scary in CivIV).

-Arrian

[Footie Mad]

That's the good thing about it, you have to stay on the edge to survive.

It never happened to me before that three of the stronger AI:s decided to attack me, but in my first monarch game Alexander, Montezuma and Napoleon declared war on me on the same turn and charged together at the same border! Don't knock the AI for being to easygoing until you've tried it. They pulverized me and this was not with agressive AI setting...

[Velociryx]

Initial results: At Prince level - 12 units (11 warriors and a worker) = 0 support cost (all units inside cultural borders).

Of those, you begin to pay upkeep costs if you have more than four outside your cultural borders, and you begin paying for upkeep in general on building the 13th unit. (one city).

Interesting thing here....I expected that when I added a second city, my # of free units would dramatically increase. It did not. My second city only gave me +1 unit to that threshold (and each city after that gave me an additional +1...

I added two more cities in rapid succession, at which point, my entire force, now consisting of 15 warriors and 1 settler, cost me 2gpt, and my city upkeep cost me four. Even more interesting...when I had only one city, it was reported on the expense summary screen that I could support up to 4 units, free, outside my border.

Upon founding my fourth city, that value dropped to three units, supported for free, outside my borders. THAT struck me as quite strange.

So...I started disbanding warriors to see what would happen.

Voila!

As soon as I dip beneath the threshold of freely supported units inside my borders, I can once again support 4 units for free, outside my borders.

Next test.....let's keep my force totals the same, and move four units outside....four outside and no maintenance. Check.

Five outside.

Still no maintenance.

Six....1gpt.

So maintenance costs for troops must be +0.5gpt, IF you go over the threshold.

The threshold itself seems to be driven by a default value (two separate conditions...inside borders and outside borders), and increases by +1 for each additional city....

at least, that's my findings so far.

-=Vel=-

[Velociryx]

Hmmm...no. Now I'm more confused than I was before.

I just re-checked my figures above...15 warriors + 1 worker should have been free if my theory was correct...but it cost me 2.

Just went back to that game and looked again. The screen says I get 13 units for free...only two more than when I just had my starting city...so apparently it isn't 1 city = +1 to the threshold.

Curiouser and curiouser....

-=Vel=-

[vee4473]

Originally posted by Velociryx
Initial results: At Prince level - 12 units (11 warriors and a worker) = 0 support cost (all units inside cultural borders).

Of those, you begin to pay upkeep costs if you have more than four outside your cultural borders, and you begin paying for upkeep in general on building the 13th unit. (one city).

Interesting thing here....I expected that when I added a second city, my # of free units would dramatically increase. It did not. My second city only gave me +1 unit to that threshold (and each city after that gave me an additional +1...

I added two more cities in rapid succession, at which point, my entire force, now consisting of 15 warriors and 1 settler, cost me 2gpt, and my city upkeep cost me four. Even more interesting...when I had only one city, it was reported on the expense summary screen that I could support up to 4 units, free, outside my border.

Upon founding my fourth city, that value dropped to three units, supported for free, outside my borders. THAT struck me as quite strange.

So...I started disbanding warriors to see what would happen.

Voila!

As soon as I dip beneath the threshold of freely supported units inside my borders, I can once again support 4 units for free, outside my borders.

Next test.....let's keep my force totals the same, and move four units outside....four outside and no maintenance. Check.

Five outside.

Still no maintenance.

Six....1gpt.

So maintenance costs for troops must be +0.5gpt, IF you go over the threshold.

The threshold itself seems to be driven by a default value (two separate conditions...inside borders and outside borders), and increases by +1 for each additional city....

at least, that's my findings so far.

-=Vel=-

Vel and all,

You know, I find it very interesting to deconstruct civ 4 or any civ game down to the barest mathematical components, but to me, the fun is gone at that level.

Strategy doesn't just lie in the tedious minutiae that can be the calculations of civ. It is in recognizing the overall situation and developing a plan accordingly. Which I think the game does a good job (better than all predecessors) of doing. It does give you the overall situation nicely.

If cold, hard calcs are your thing then cool beans, but I have found after many hours of civ 4 playing that the game itself has been quite complicated to the point that it is impossible (regardless of what many of us would like to think) to be computers calculating the best mathematical approach to what to build and when with how many cities under which civic and when following this religion and allied with whomever depending on the terrain etc...too many variables to say that any course of action trumps all others.


I think that the game has done a decent job at balancing all of these aspects to the point where sheer calculations are not needed (except maybe on Deity).

I love just playing the game (with a general path, not a hard, cold calculated strategy) , absorbing what is going on, and making decisions accordingly.

Having said that, I must admit paradoxically, Vel, that your analysis and others will keep me riveted nonetheless!!

I love this s***!!

[Velociryx]

Believe me...I'm on that same page with you!

I'm definitely not a bean counter (never had the patience for it), but I do like to wrestle some of the basic mechanics to the ground (for instance, the whole reason behind my curiosity about what, precisely, is the cutoff point for units, is that I've got it in my head to create a more organized approach to the whole "defending the empire" thing, and I'm interested to know if my idea will break the bank, or if the empire will be able to support it.

True, I could just build up and see what happens, but I often find it more enlightening to get under the hood, as it were, and find out why it does, or does not work.

For example....knowing what I now know about the whole troop limit thing, I must confess to being quite confused about my maintenance costs.

Last game, I had 8 cities, two garrisons in each, AND an army of some 20 units invading another civ, and I didn't even FEEL the maintenance costs. Based on the numbers from my little experiment above, my I should have been able to support something like 18 freebie units, and then ~0.5 gpt for each one after that. So...my maintenance costs for the troops should have been about 11gpt.

What I find curious about that is that during the war, I was running 90% science and still making money....which should have been impossible, if I'd had troop maintenance (I keep a close watch on my gold, and even at 8-11gpt maint. for troops, I shoulda been in the negative).

On the one hand, I guess the argument could be made that it doesn't really matter how it works, only that it does, and that's true enough. But if I want to replicate that feat again later.....

-=Vel=-

[Arrian]

Jumped up to Prince last night and played an Ice Age map. It gave me something very much like a CivIII 'pelago (stringy islands/continents) on a cold, 3 billion world.

I stuck w/Ghandi - I really like the traits & fast workers. I out-expanded the AI, in part b/c I went for sailing early and used galleys to speed my expansion and in part because I was lucky enough to start on one of the larger & more food-rich landmasses. Probably more the latter, huh? Oh well, I'm not complaining. I feel I leveraged it well. DeepO would be proud - Great Lighthouse, harbors in every city...

For what feels like the umpteenth time, Catherine of Russia just flat *hates* me. Every game I play that she's in, she hates me. Frequent demands, refusal to trade/open borders, and the inevitable war. She slaps me a lot. I showed her, though. St. Petersburg is now mine (if I can hold onto it - the culture situation is rather difficult). I think I've got the game won. Since I'm still in the mid-game, though, we'll see if that holds up. Anyway, I didn't feel troop costs in this game (despite a short offensive/pillaging campaign while running pacifism), but I sure did feel city upkeep costs early on. I expanded pretty aggressively, and the result was, IIRC, a 30% science rate at one point.

Incidently, it occurred to me as I saved & quit for the night that a financial civ would ROCK that map. +1 gold per tile that already gives 2, right? Well, that would be all those coastal tiles I'm working. If I were a financial civ, I'd be raking in quite a bit more commerce (say 25% extra). edit: hmm, maybe that's why China is doing so well despite being so small...

-Arrian

[]LoL[Harm]

Started another Monarch game last night. Again I cannot for the life of me figure out what I'm doing wrong. I wonder how heavy the bonuses the AI gets are at this difficulty level (I have yet to use the world builder to verify anything).

Huge World, Continents, Temperate (random water level), Monarch, Normal were the settings. I picked Peter/Russia and was going to go for a Great Person build. Got my 2nd city built 27 or 28 turns into the game using Settler First. Tech wise I beelined for the Parthenon (+50% GP production) and then back filled with Archery and worker techs so I could improve my land. Got the Parth build and in the beginning I somewhat held my ground against the AI. But as I reached Construction they were getting Liberalism.

I went about this game playing with these things in mind, I was going to Castle and use Great People as mostly Specialist and I was not going to chop any of my trees down in order to see if not using any chops would still allow me to keep pace with the AI. I had one culture city, one gold city (it was still maturing) a production city that was pumping out around 18-20 hammers a turn (pre repl-parts/indust), and my captial was a combo gold/research/GP producer. (my fifth city wasn't specialized yet)

I kept my science rate at 90% throughout most of the game. I finally dropped to 80% when founding my fifth and final city.

I had at least 6 or more units in each city, my border cities probably had around 10 of mixed variety including catapults. Walls and Castles in all but my newest city.

I was a few techs (I think two) away from gunpowder when Louis/France declared war and landed with Infantry. At this point in the game I was around 500 points, the leader I think was over 2000 (I think it was Ghandi), France was upper mid with around 1200.

For the life of me I can't seem to get a builder/castler strat to work on Monarch level. (Though on Prince I can Castle like a mad man). Given I didn't chop in this game which probably hindered me alot, but I wanted to see if the builder/castler strat was viable with Environ/late game production as a goal.

Has anyone been able to succeed with a builder/castler strat on Monarch level?

I only played Monarch in Civ3, Emperor when I wanted an extreme challenge, it's starting to feel like the jump from Prince to Monarch is just too much for me in Civ4 using a non-agressive strat.

[Arrian]

Just a thought... 5 cities on a huge map? Seems rather small... and thus your inability to keep up in the tech race, maybe?

I haven't gotten to Monarch yet, and I've heard it's the first "big step up" in difficulty (AI starts with a worker, plus normal production/science advantages).

-Arrian

[Cort Haus]

The unit support is interesting. Paying maintenance for troops outside your borders means early rushing or mass exploration with a ton of warriors or equivs will be expensive, but heavy defending within the civ borders is not.

In C3C I always felt that the extra upgrade costs to nerf the mass upgrade protocol penalised defensive play alongside the aggressive play it was supposed to blunten. The other guy found it harder to upgrade his ton of horsies to knights, but the defender found it even harder to upgrade his Pikes to Muskets. Yeah, 2 Pikes were better defenders than 1 Musket, but not economical under a C3C Republic.

So in C4 although the upgrade costs are the same, support costs are reduced for defending armies.

Anyway, I fancied trying a Quecha rush last night - probably from a single city (Barracks, Quecha, Quecha, Quecha etc) but found myself alone on the continent. Tiny continents maps seem to give 1 civ per continent not 2 as in Civ 3. The Malis from the north have arrived to take advantage of my non-rexing and plonked down a city, with their hyper-archer UU.

I can probably remove that city with my archer-busting Quechas, and build loads more (got 2 barracks cities now and a third ramping up), but ulimately they have copper and I have no metals for making pointy things. I guess they only have to drop in a Spear + Axe or something and I'm in trouble. I have horses (not hooked up yet) - will they be any use or am I forked without Iron/Copper?

This (Noble level) is only my second start. The other was a builder-only game with Wome on Warlord). I liked the synergy of Romes's traits. A cheap granary early on grew the city to 6/7 in no time, spitting settlers every few turns with fast research (rivers are even more important than in Civ3), then cheap Lighthouses, Harbours and the Great Lighthouse were a commercial rocket.

Generally, I'm with those who like to play with the abilities of their Civ, the mood of their playstyle and adjusting to the environment rather than seek 'bean-counter' derived generic formulae.

[Arrian]

Hmm, interesting thoughts on Rome. Haven't played them yet. Must try...

-Arrian

[vmxa1]

vee4473 I think the idea is that people want to make a good decision. This requires data.

They want to know if is better at this point to make a worker or a settler or neither. What to research and when and all the other things.

You can go along and play a ton of games and get a feel eventually or not. If instead you can learn what the impact of this or that is, you can make an informed choice.

So you will benefit from understand as precisely as you can how the mechanics work.

[]LoL[Harm]

"Just a thought... 5 cities on a huge map? Seems rather small"

This was my thought as well, 8 cities was my goal, I even had a settler built (for my forth city) and idle for about 10 turns while I ramped up some gold production as to not cut my research below 90%.

The problem is I know if I hit 8 cities quickly that I'll be stuck at 40%, maybe 50% research at best, and will not have hardly enough defenses to keep the AI at bay. If I build up defenses in these cities as high as I had in my current setup I'd probably be around 20-30% research due to unit upkeep.

On one Monarch map I REX'd heavy and came out with around 12 cities, it got to the point that I had 0% sci output and was still losing 4gp a turn. My empire went on Strike and warriors and other military units began disbanding until I hit 0gp/turn. My nation collapsed and I was soon at war with little ability to resist.

I almost want to think that I have to pump out workers and get as many of the necessary worker improvment techs as quickly as possible in order to suppliment expansion. The problem occurs around 1000BC when the barbarian hordes sweep in heavy and begin razing the countryside.

I'll try a mass worker flood early on to see if it will support heavier expansion tonight.

[Arrian]

This was my thought as well, 8 cities was my goal, I even had a settler built (for my forth city) and idle for about 10 turns while I ramped up some gold production as to not cut my research below 90%.

Woah! I'm thinkin' holding off building a city when you've only got 3 b/c you're worried about dropping the research rate is part of the problem. City #4 should be a "core" city and if you need to go down to 80% for a little bit until it starts bringing in some more commerce, so be it! After all, 80% science out of a total of X (where X = the combined commerce of 4 cities) may be more than 90% of Y (where Y is 3 cities). Maybe not right at first, but fairly quickly.

I was at 30% (!) science in my last game, breaking even, and was still expanding. Normal sized map, ice age. I built 11 cities. It paid off well.

EDIT: the barbs are probably the biggest difference. On the maps I've been playing, they're a relatively minor concern. Up a level on a huge map (way more land), I imagine they may be more difficult to deal with and that would impact expansion... but still, I think there is a happy medium between 5 and 12 cities.

-Arrian

[Dominae]

Is it ok if I jump in here?

Originally posted by Velociryx
Essentially, when you chop, and you use the chops to build workers and settlers, than you're making the tradeoff, gaining "mobile" production (workers/settlers), in preference for "at home" production (chopping slightly later to speed build a wonder there).

There are multiple reasons not to chop early on:

1. Health bonus (+0.4 per Forest within workable radious): this is not inconsequential on some starting locations. I have sometimes kept 5 Forest around for a nice free Aqueduct effect (that's 100 Shields, is it?).

2. Lumbermills (+1 Hammer on Forest only, +1 with Railroad): none of the analyses have taken into account this nifty little improvement. Late-game a Grassland Forest tile with Railroad produces 2 Food and 3 Hammers, which is competitive with all other options.

Originally posted by Krill
If you want to see what happens with vast forest chops, play the game through, and look at what happens with out lumber mills and the health bonus, and consieder that the tile that is chopped has lost that extra shield for a long time...

3. Land (i.e. the capital's workable tiles): one some starting locations, you actually need every +1 Hammer Forests provide!

What's more, often two or more of the above conditions are present in a starting location.

Originally posted by Velociryx
The thing I'm seeing here tho...the trend is that early chopping to get those initial cities out (3-4) is so overwhelming an advantage that all the other branches of the tech tree wither by comparison.

While I'm generally in favor being greedy when it comes to such things (and not really considering the late-game), with respect to Forests I'm doubtful "always chop" should be the take-home message.

Originally posted by Friedrich Psitalon
I typically start with a worker and then begin the forest hackage, going straight for bronze working. I mow down the forests around my capital pretty aggressively; clearcutting is not unheard of. I can do this for two reasons:

MP is a different beast because for all intents and purposes there is not "late game". So I would say that chopping Forests is close to a no-brainer in MP.

As Fried mentions, Settler-first makes almost no sense in MP because of the increased probability of just losing. The advantage of Worker-first is flexibility: every strategy is helped by having a Worker available. With a little scouting you may realise that REX is the right option and starting farming that Corn; with a close neighbor you can Forest chop some last-minute defenses. In this respect, for the 3 or so turns that the Worker is chopping, you have much the same "production carryover" ability that we had in Civ3.

Flexibility is good in SP, too, but less so.

Originally posted by Jcg316
Worker, Settler without chops is as fast as Settler, Worker without chops (the turns and benefits of each are listed in the first couple pages) plus already you have improvements and are still capable of varying from out of worker/settler to something else. Frankly there are very few situations where settler first is better, mainly if you have no starting tech for any of your starting resources and still want an early religion. Settler/Worker has the benefit, and only benefit, of starting a second city 8 turns sooner. You will, however, be able to upgrade the capitol as quick the other way.

I'm glad to see this consensus reached so early on, with some great testing efforts to support it. I agree that the "X cities in Y turns" is not a very good metric (especially for CIV), but in this case it works to demonstrate that Worker-first is often better than the greedier Settler-first.

I play Emperor and Immortal, and let me tell you, the AI can and will punish you for a frail, expansion-heavy start. The Barbs are no push-overs either. Getting X cities in Y turns does not seem like a strong opener to me when I have to spend the next Z turns making sure I do not get run over.

Just by the way, the fact that tile improvements in CIV are so generous means that that "base city tile" aspect of ICS is now less relevant. Building an extra city will still give in 2 Food and 1 Shield automatically, that's true, but putting that Cow in a Pasture will do the same.

Originally posted by Velociryx
Yep....a *special resource* is worth more, improved. That's true. If we assume an average start (as I have been, from first mention), you may or may not have one (but you could have access to one quicker if you hurry up and build that settler!). If we're talking about general tile improvements (farms/mines on "generic" terrain), then my assessment is dead on. Farm = +1 Food, and you can't rely on having juicy specials laying at your feet....at least not to the extent of designing a strat around it (proviso: not without committing yourself to LOTS of reloads until you get a start that's "optimal")

It seems you have been pretty unlucky with starting locations: on average you should get at least a couple of bonus resources, usually one Food (this includes Flood Plains). If all you have is Grassland, Forest and Hills, then I agree that deforestation is a strong opener.

One thing about Farms is that they take 5 turns to complete which I believe is the highest of the early-turn improvements. And this for a measly +1 Food. Unless I have Flood Plains and know that I want to spam GPs, I avoid Farms as long as possible (not counting those require for grains).

Originally posted by Jcg316
One religion does not impede your abilities with worker first in any significant way.

It certainly depends on what your initial techs are. If you go Worker-first and Meditation-first, unless you start with Agriculture or Mining your Worker will likely be standing around uselessly for a few turns (or building Roads, meh). Sometimes if you really want that early Religion it makes sense therefore to build a Warrior first in your capital, and let it grow to size 2.

One aspect of the early Religion (and the immediate focus on the religious branch) is the pacification of potentially aggressive neighbors. All it takes is an Open Borders and one Missionary to get the another civ to adopt your Religion (assuming it failed to found one itself), which translates into a heavy boost to relations, depending on the leader.

Originally posted by Aeson
Don't ignore "big city" in favor of "quicker cities". Each has it's benefits and drawbacks. CIV is not like past Civilization games where expansion was exponential. Playing catch-up in number of cities is viable, and can have greater long term potential. Locusts are fast expansion, but will be very slow early research, miss most the early religions, miss more wonders, and will be extremely poor at choking a neighbor or fending off an early rush.

Yes! Listen to Aeson!

Here a couple of non-standard openings I have used to some success, both in MP and SP:

1. Worker, Worker, chop Forests for Granary, grow to maximal size while building Warriors, Settlers.

Because I had kept my population at size 1 my Granary sped every single one of my capital's growths. The presence of dual Workers so early on meant that all resources were connected whenever they were needed, and my road network was never lacking.

Obviously this sequence meant I had to keep my Warrior close to home in case some enemies showed up. But note that, with Workers chopping Forests, I was never far away from a flash defence.

2. Worker, Pottery, Cottages on Flood Plains, Writing, Library, Settlers.

The idea here is to maximize research, letting Cottages grow while you build your Settlers. This strat clearly shines when using a Financial civ.

Your Palace/capital produces 8 Commerce per turn right from turn 1. This is far and away better than anything your expansion cities can hope to do until much later. So if what you want is a high research rate, this is one good way of doing it.

3. Scouts until size 2, Worker, Mines, Barracks, Chariots.

There are many variations on the early rush in CIV (try it with Quechua), and they're all quite potent. Generally it's best to build your units to rush before you found your second city (i.e. you build your first Settler while your first wave is en lieue).