No announcement yet.

Early Conquest Guide

This is a sticky topic.
  • Filter
  • Time
  • Show
Clear All
new posts

  • Early Conquest Guide

    Inspired by Solo's ELG and DaveV's Guide to ICS, I have written a draft of an Early Conquest Guide. DaveV kindly reviewed a short early version, and offered some advice. Now it is a bigger better guide.

    Chs 1-5, at least, seem fairly helpful and accurate. But they focus mainly on a single strategy intended for Deity on normal maps. The guide could be much better with more in Chs 6-8 about alternatives, and better advice about how to choose one.

    This might take a long time for just 1-2 people. If a few strong players are willing to help out, I am happy to make this a team effort. Maybe we can write something comparable to the ELG. I can send you an advance copy of the guide, but may ask for help with play-testing, collecting ideas, or (re)-writing some of the later sections. For now, here is a table of contents:

    CIV2 Guide to Early Conquest:

    1) Introduction to Early Conquest (EC)

    a) Welcome to EC
    b) Playing Levels and Styles
    c) The Scope of This Guide
    d) A Typical EC Game
    e) What's Par for the Course
    f) What to Learn, and How

    2) Notes on the Opening

    a) Cities, Shields and Growth
    b) Making and Using Units
    c) Tech and Wonders
    d) Values
    e) Preparing to Attack

    3) Attacking the Enemy

    a) Unit to Unit Combat
    b) Attacks on Cities: Case Studies
    c) Attacks on Island Civs: Case Studies
    d) Attacks on Other Civs by land: Case Studies
    e) Movement
    f) Effects of terrain, units, etc
    g) Summary: Planning a single campaign

    4) Planning Multiple Campaigns

    a) The "Farthest One First" Rule
    b) Production and Your Late Swarm
    c) Timing and Combining Campaigns.

    5) Middle and Endgame Tips

    a) Diplomacy; M.P.E, maps, tech and tribute
    b) Production, Growth and Happiness
    c) Finishing the Game
    d) Game-stoppers and Heart-breakers

    6) Other Playing Levels and Settings

    a) Emperor
    b) King and Below
    c) Restarts ON
    d) Barbarians
    e) Huts

    7) Other maps

    a) Very poor land (swamps/deserts/etc)
    b) Islands and Continents
    c) Large and small maps

    8) Other others

    a) Trade
    b) An SSC
    c) Other Wonders
    d) Other Governments
    e) Other Strategies
    f) Conquest after 1000AD
    g) Abbreviations
    h) References, Links, Credits

  • #2
    Not much recent experience for me on EC games, so as a fellow author of guides, I'll offer you some encouragement. It will be a lot of work to be as comprehensive as you propose, but there is a gap here that your proposed strategy guide will fill very nicely.


    • #3
      Thanks for the kind words, solo!

      I have put a lot of very enjoyable work into it. I believe Chs 1-5 already present a reliable system for conquest by about 500AD on normal maps (+/- a few hundred years depending on player skills).

      As far as I can tell from forum-haunting, there is not much agreement on what the `best' EC system is (It probably depends a lot on the map). Actually, most forum advice seems to be written by EL players, or players-for-score, and doesn't apply well to EC. Here are some tentative opinions on EC that need testing or discussion:

      1) All you need is Monarchy (among governments).
      2) You don't need trade routes, an SSC, or alliances.
      3) The only attackers you need are mounted units and dips.
      4) Shields beat arrows.
      5) Strategies based mainly on huts are risky, but work well on 1-continent maps.
      6) Island games tend to be shorter, partly because the AI's are weaker.
      7) Games at lower levels have similar ending dates to those at Deity.
      8) You don't need tech after monotheism, except maybe navigation.
      9) Barbarian settings do not affect overall strategy much.
      10) Monarchy 8 turns early is worth less than one extra city.
      11) Starting techs do not affect overall strategy much.

      Also, I am still undecided about the ideal city size for EC. The Ch1-5 system involves about 15-20 small ones.


      • #4
        If your only adviser is dave, this is going to be an ICS guide , and such a vile provocation cannot be tolerated .
        I am willing to contribute (at least as interested reader, and perhaps with a few ideas ).
        Aux bords mystérieux du monde occidental


        • #5
          @ La Fayette - That's great !! Actually, I wanted to ask you especially, because your style is so different from mine and Dave's. I will send you what I've got.

          Any other players with crazy ideas about arrows and size 3 cities are also welcome.

          @La Fayette - How can I contact you?
          Last edited by Peaster; October 17, 2004, 18:20.


          • #6
            Don't forget the Size 1 Settler strategy, especially for levels lower than Deity.


            • #7
              Originally posted by Peaster

              @La Fayette - How can I contact you?
              My PM box should be empty now.
              Aux bords mystérieux du monde occidental


              • #8
                I'm not a strong player, but I am a good writer, and I would love to participate in your process if only to become a stronger player.


                • #9
                  @grigor: Good point. I will put that into Ch6.

                  I know you have a lot of experience with GOTM's, so I could use your help in Chs 6-7, if you are interested.

                  @LaFayette: I don't know how to attach a file to a PM, so I tried sending you a link. It says your box is full (again). You must be a popular guy!

                  @debeest: Your help is gladly accepted! I will send you the link, and you can edit the parts that are not clear or that run on too long.

                  You'll get a pdf file that's easy to print [if you have Acrobat] but may be hard to edit. I can email you a text file that you can open with notepad, if you prefer.

                  There may be some playtesting still to do, and you'd be welcome to help with that, too.
                  Last edited by Peaster; October 18, 2004, 22:17.


                  • #10
                    There are exactly 2 messages in my PM box
                    Last edited by La Fayette; October 20, 2004, 03:58.
                    Aux bords mystérieux du monde occidental


                    • #11
                      Thanks especially to LaFayette, solo and zenon for help (in other threads). Here are some more ideas that have come up. Opinions or PlayTesters are welcomed.

                      1) Play with only 12 cities until you can capture HG.

                      2) Instead of MPE, explore quickly (This might work on smaller maps, with restarts off).

                      3) When is it good to build ST ? (in general, wonders seem to make more sense on larger maps with decent terrrain).

                      4) Does a surrounded AI city tend to produce more defensive units ?

                      5) Does the AI tend to build walls more often in wartime or when threatened?

                      6) When an AI starts building the Great Wall, do you try to do it first?

                      7) Are diplomats a better choice against city walls than crusaders? (My wild guess is each dip has about a 40 per cent chance of taking down walls in the pre-500AD era, which is a good bet if the city is medium/large).


                      • #12
                        The latest version of Ch 3 (on attacking) should be attached as a pdf. Criticism welcomed. Maybe I'll start posting some of the guide directly into the thread, but chapter 3 includes a couple of tables (can you paste HTML tables into these threads...?)
                        Attached Files


                        • #13
                          The pdf apparently will not download, so here's Ch3 again. HTML tables work, but for some reason, large blank spaces are created above them (ideas???).

                          CH 3: Attacking the Enemy

                          First, a word about units - plan to attack mainly with small groups of 3 to 6 crusaders - veterans, if possible. Non-vet elephants are OK until you get monotheism, but they will have a hard time if they meet city walls. Vet ellies might replace crusaders entirely. I will call these various mounted units `cavalry' (the true Civ2 Cavalry Unit is not part of EC strategy). Diplomats are very useful for bribing cities after the capitol is down. A settler can make a colony, perhaps 5 squares from the enemy, from which you can rush-build attacking units. You don't need other units, except some militia.

                          For true EC, you won't have time for musketeers, cannons and so on. Delaying your attack for such units probably doesn't help anyway, because you give the AI more time to build up. If you attack before 500AD, you'll face relatively weak phalanx/legion type units, with fewer walls, and your crusaders will shine.

                          The earlier you can get send out a group of cavalry the better. With enough practice, you might aim to do this before 500BC. You probably can't do much with slower units like catapults or weaker ones like horsemen. If you get some archers, for example, from your first huts, maybe you can take a nearby city, but you can't plan on it.

                          In general, attack the capitol of each civ first, if possible.
                          Then your dips can bribe many of the other cities. The first city often seems hardest to crack, so build up some extra force for that.

                          This rest of Chapter 3 focuses mainly on how to use your cavalry to defeat the enemy, and especially how to take cities. Then we will turn to long-range planning.

                          a) Unit to Unit Combat, Terrain and Movement.

                          The simple rule when you attack the enemy in the field with a crusader is - you win. A fortified pikeman on a mountain would be an obvious exception. You might bribe him, or back off and hope he wanders away.

                          You will lose units mainly when the AI attacks you first. For example, you have to slog through a forest to get to those ripe un-walled cities - and a hidden chariot takes out three of your units in a blink.

                          You have several options to avoid this. Use rivers, roads or triremes if possible. A crusader can step off a boat onto grass and attack in the same turn. If you have the time and the units, send a pikeman, explorer or a warrior ahead of your cavalry. Or, stop your cavalry unit before using all it's movement points, which lessens the chance of an ambush. Avoid stacking units, to minimize the risk. It is usually worth delaying your troops in these ways for safety's sake.

                          b) Attacks on Cities.

                          This is what conquest is all about. Usually, an enemy tribe starts to crumble after you take its first city, especially if it is the capitol. Let's imagine that you have chosen your target city. How do you attack it?

                          Build up a strong force before you attack (and we'll talk numbers in a just a moment). You will be curious about how many defenders are inside, but don't waste a diplomat on investigating a city unless your decision is really tough. If in doubt, bring in more troops. The troops that are waiting can fortify about 3 squares away, preferably not stacked (unless you find an empty fort), preferably on a road/river. Non-vet units might look around for victims - chances to become vets.

                          Now, how to approach the city? If it is a coastal city, try to attack it by unloading from boats onto a nearby grass/plains square and attacking the same turn. Or, if there is a road or river into the city, you can amass troops on it, at a safe distance, before attacking. Without these benefits, you must plan more. On non-roaded grass, place your troops 2 squares away, preferably on one of the "diamond tips" [eg due north] where the AI cannot see you. Again, try not to stack your units too much (AI bribery is possible, but rare). You might use a spare warrior/phalanx to fortify a key square or to block counterattacks.

                          It is unlikely that the city will be completely surrounded by forests, hills or mountains. But, if so, amass extra troops 2 squares away and move them up to the city all on the same turn. If you suspect the AI has elephants or catapults inside such a city, try to fortify a pikeman or two near it first. If you don't care about reputation, this kind of city might be a good place to lose yours (make peace and sneak attack).

                          If the city is not walled, you can expect most of your attacks to succeed. Typically such a city will have 2 defenders, such as archers. With luck, you can take the city with 2 crusaders, plus 1 horseman (or whatever) to enter. With roads, or against a small city, you might even succeed without the third unit.

                          At your own risk, you can even send a lone vet crusader out against small unwalled towns (c3 or less). If the civ is too poor to rush-build, you may be able to remove the defender the first turn, and enter the town on the next. If it has two defenders, you probably lose.

                          Walled cities tend to be bigger, with maybe an extra defender, depending on whether you have worn the enemy down in the fields. Plan to use at least 6 vet.cru's on a size 5 walled city, which I call a c5+. Remember that a city on a river will be tougher, so add a couple of attack units for that.

                          More rules of thumb -
                          Attacking with non-vet ellies against c4 (no walls), plan to use 4 units. For size 5, plan to use 5 ellies, and so on. Don't attack walled cities with non-vet units. A crusader is worth about 25 per cent more than an ellie. Attack with less only rarely, only for a clear reason. In practice, the question is often not "What should I attack with?" "What can I attack with?". If you have other units only 2-3 turns away, it is usually wise bring them in before the attack, to raise your odds.

                          Solo suggests these rules of thumb: "For cities without walls, 4 non-vet crusaders are plenty for taking out cities with 2 defenders and 6 will be enough for bigger cities with 3 defenders. For walled cities, about 12 provide enough cushion to assure victory."

                          Here are some statistics you can use to form your rules.

                          TABLE I. Factors that depend on City Size:

                          Size Freq Defs Vets
                          2 4 1.0 0
                          3 9 1.5 0.2
                          4 12 2 0.3
                          5 3 2 0.5
                          6 1 2.5 1
                          7 0 2.5 1
                          8 2 2.5 1

                          This data comes from 31 cities from 4 of my games, all from 400BC to 300AD. As you can see, there were 12 c4's, which had an average of 2 defenders each, including 0.3 veterans (on average). I counted warrior as 0.7 units (but they were not too common). Almost all these units were fortified. Roughly 10 per cent of the cities had walls. The average city size grew from 3 to 4 over a period of about 150BC to 250AD. The average number of defenders per city grew from about 1.6 to 1.8 over that period.

                          TABLE II. Number of attacking units required to destroy a phalanx.

                            2 3 5 6
                          ele 1.8 2.7 3.5 4.0
                          cru 1.5 2.2 2.8 3.0
                          v-ele 1.2 1.8 2.3 2.4
                          v-cru 1.0 1.5 2.2 2.3

                          The numbers across the top are the number of bonuses (+ 50 per cent) the phalanx has. For example, the column under the "2" might represent a phalanx that is 1) fortified and is 2) a veteran. If it is also on a river, look under column "3". Since city walls add 200 per cent, count that as 4 bonuses.

                          I have left out some rarely-needed columns, and do not claim the numbers are terribly accurate. These stats are based on tests of about 10-20 attackers starting out adjacent to a city with about 4 defenders. They do not include losses you might suffer getting into attack position. You will need an extra unit to enter the city, and should include extra attacking units as insurance.

                          How to use the chart? Ex: Let's assume you are attacking a c4+ with vet ellies from a distance of two. Table I suggests the city has 1.7 non-vet defenders and 0.3 vets. They are probably fortified, and the walls add 4, making them types 5) and 6). So, Table II suggests you need 2.3 x 1.7 units for the non-vets, and 2.4 x 0.3 for the vets, which is 4.63 ellies. Use no less than 5 vet ellies (I'd suggest 6) for attacking, plus any available unit for entering. Expect to lose 2.63 ellies (on average) in the process.

                          You might run into a city that defies the odds. It might have 6 units waiting for you. It might have built an early barracks, and have 3 vets inside. It might have pikemen with extra bonuses. So, build up a larger force than suggested, if you have time, and watch for clues that a city is going to be trouble. [You see traffic, the civ has feudalism, etc].

                          You can try using dips to sabotage enemy walls. I haven't studied this option very carefully yet, but estimate that most walled cities will have just one other improvement, and your first dip has about a 40-50 per cent chance of getting the walls down. If you can use more dips on the same turn, your odds go way up. The production cost of the dips you lose this way is probably less than what you'd lose in a direct attack (not sure).

                          After the capitol is down, is it easier and faster to bribe cities than to attack them (but do some of both, to keep costs down). Some typical costs are: c1 = 60g, c7 = 350g, a vacant c2 = 50g. Also, you will probably rush-build the dip, adding about 60g to your cost.

                          After taking a city, sell the improvements (you can keep barracks/walls a while). I usually change the production to dips unless I really need something else. It is rare that the AI will try to retake/bribe the city very soon, but watch out.

                          c) Attacks on Island Civs.

                          These are not too different from continental attacks, so read this section and the next one together.

                          Recommended strategy: After using MPE to get a map of the island, send two boats (four crusaders) to capture a city within one square of the sea. Ideally, it should be about size 3-5, without walls, so that you can win it without razing it. Choose the capitol, if possible. Once you take this outpost, you can heal up and rush-build more units (mainly diplomats if it was the capitol).

                          This works faster if you have back-up units on the way when you take that first city (this is also good insurance against disasters at sea). The plan can fail if the city builds walls during your voyage. But hopefully you can find another victim city, or will start so early that this
                          doesn't happen too often.

                          It is pretty risky to try this with less than four units, or even with 4 virgin ele's. If you haven't made enough ele's yet, you might send a settler along instead, to colonize. It is really annoying when the AI (at peace) asks you to withdraw your troops from an AI city's ZOC, but a nearby colony makes this less of a problem. Instead of a settler, you could send a diplomat to bribe the first enemy city. If the enemy has no good target city on the coast, you can easily modify any of these options, but travel over land is a bit slower and riskier.

                          After landing, your boats can either return home, or disband for shields, or hang out. I usually let 1-2 boats hang out, in case my units conquer the island in time to move on to another target. Don't worry too much about leaving defenders or militia behind.

                          A typical island invasion takes about 5-10 turns and costs about 400g. You'll get some of that gold back.

                          If the civ has the Great Wall, all the cities get city walls (in effect). So, take out the wonder-city ASAP to get the GW yourself. You might even decide to build the GW before an AI does, but that's a lot of shields you could spend on units.

                          d) Attacks by land

                          This applies to civs on your continent, and to nearby island civs which you plan to attack by ferrying troops across a channel. These invasions tend to be a bit slower unless there are good roads, so you may want to treat these like island civs and attack by sea if you can. But you might choose a land attack anyway, so you can deliver more troops, so

                          - you are less dependent on bribes

                          - you can move on sooner to another target.

                          About 10 attack units is typical. The main delay is amassing enough units near the capitol, but then it is usually over quickly. Another option is a blitz attack starting at a boundary city with about 4 vet.cru's (before walls). Send or build more units asap.

                          One useful strategy here is to make outpost cities near these civs, probably during peacetime, about 15-20 turns before you really need to attack. It's a good use for excess settlers, and the inevitable border skirmishes usually work in your favor. Try to settle about 5 squares away from the enemy, on the side nearer your homeland, and prepare to defend.

                          Movement Tips: When building a road or city, let the settlers take their turn first. Use the `pre-rushed settler' trick if you like. Also, if you are undecided about where to send a unit, let it go last. Making cities on hills/swamps etc can speed up a road, especially late in the game.

                          If the enemy is poor, with small towns scattered across grasslands, consider the risky lone-crusader idea from part c).

                          Summary: Planning a single campaign

                          Assuming the civ is about 4 cities and is not walled or spread out, you will need about 6-8 units to take over. The first 4 should arrive together and attack together. Some of the others may be rush-built later, especially in the case of a big distant island civ.

                          The invasion will probably be over within 10 turns after you arrive, or about 4-5 turns after you near the capitol. Reserve about 400 gold if you expect to rush/bribe, but expect to get a lot of that back. As discussed in Ch 4, you may move your best troops on, before victory is complete, especially if it is not urgent.

                          Add 2-5 turns to your estimate if the civ is spread out, has more cities, or has few roads. Try to use your speedy dips in these cases. Add a few units if the civ has feudalism (pikes), walls, or some other scary feature.
                          Last edited by Peaster; November 12, 2004, 11:27.


                          • #14
                            Nice chapter

                            Those results are in accordance with my own experience.
                            Therefore: not much to add, except that I have studied thoroughly the problem of destroying walls with dips and spies (the results have been published in the GL "Info dips and spies", but SlowThinker kept them in the 'to be discussed' part, since he was not completely satisfied).

                            Most detailed results are to be found here:

                            Anyway, here is what can be kept in mind:

                            There are 2 methods, but diplomats can only use method #1 (method #2 = only for spies).

                            1) "Use her judgment" (you are not allowed to choose which improvement you attack).
                            2) "Choose primary target".

                            I'll give some details about method#1 (since spies are beyond the horizon of EC).

                            When the diplomat (or spy) attacks, the first "improvement" destroyed is USUALLY the improvement (or unit) under production, followed by the improvements in accordance with the list in the city screen, walls being last.
                            (I think that SlowThinker disliked the "USUALLY" and that is why he didn't publish those results in the summary, but IMO I can't be held responsible for the slight random factor introduced by the authors ).

                            Anyway, to make it short, if there are n improvements inside the city, you are on the safe side if you attack with (n + 1) diplomats = 100% chance that the walls will tumble down.

                            (but you have also a very slight chance to see the walls down after an attack with one dip )
                            Aux bords mystérieux du monde occidental


                            • #15
                              This is good stuff, Peaster! I look forward to seeing the rest when it is ready.