1. We have added a Gift Upgrades feature that allows you to gift an account upgrade to another member, just in time for the holiday season. You can see the gift option when going to the Account Upgrades screen, or on any user profile screen.
    Dismiss Notice

Expansive caste system GP boom

Discussion in 'Civ4 - Strategy & Tips' started by vicawoo, Apr 8, 2011.

  1. vicawoo

    vicawoo Chieftain

    Joined:
    Feb 12, 2007
    Messages:
    3,226
    I've been putting together a variation of the standard midgame strategy.

    The principle: The most powerful beaker generation is great people post classical.
    you need to things two generate great people: a lot of accumulated food (enough population) and the ability to generate great people (usually a library).

    Usually this requires a decent amount of infrastructure investment: 90 hammers for a library, 60 hammers for a granary. With whipping, don't expect to your city to be ready for 20 to 30 turns, which is too late for a lot of high level/isolation builds.

    However, caste system allows you to skip libraries. And expansive granaries can be chopped almost one chop. Due to this, expansive civs can grow large much faster than a normal civ.

    It may not sound revolutionary. Certainly it's not the first time someone with an expansive civ used caste system. But the real work will be in details, adjustments, refinements. But from the numbers it should be surprisingly effective.

    Calculations:
    Pig city
    Spoiler :

    On 4th turn, pasture improved, chop the granary (9 food).
    It will take 13+12+14=39 food to grow to size 4 from there. So with no other food surplus and 2 food tiles, that will take 7 turns.
    So 11 turns to set up a 3 scientist GP farm


    5 food city
    Spoiler :

    4 to 5 turns to improve
    About 35 to 40 ~ 7 to 8 turns to grow to size 4.
    So 11 to 13 turns to grow to size 4, 3 scientist great person farm.


    1 six food, 1 five food, target size 8
    Spoiler :

    On 4th turn, pasture improved, chop the granary (9 food).
    2 to 3 turns to size 2.
    12 + 13 + 14 + 15 + 16 + 17= 14 x 5 = 87 food from size 2 to 8, with a surplus of 9 food. So about 10 turns.
    Around 16 to 17 turns to set up a 6 scientist GP farm.


    2 five food, target size 7
    Spoiler :

    4 to 5 turns to improve
    3 turns to size 2.
    12+13+14+15+16=70 food to grow from size 2 to 7. Surplus of 8 food. So about 9 turns to grow.
    So about 16 turns to set up a 5 scientist GP farm.


    So: get lots of city, get caste system, and generate an absurd number of great people.

    Some refinements:
    Get your cities at their target size before you hit code of laws.

    Golden age with your first great person. If you time your great person for when you get code of laws, you can golden age for a caste system switch, and your cities can start immediately generating great people with a 100% bonus.
    This way you "waste" minimal turns generating your second great person.

    Two workers when you set up a new city. First improves the food resource, like normal, the second chops a forest. With mathematics (or 10 extra hammer), you should have your granary finished halfway through size 1.
    Then send your chopping worker to your next city.

    The stall trick growth trick, see my granary thread.

    You can possibly skip slavery since you're chopping/not whipping most of your stuff. Or you could use cities already set up.

    Testing:
    I recommend a non-spiritual, non philosophical expansive civ. Mao is a good one to avoid too much influence from secondary traits.
     
  2. vranasm

    vranasm Deity

    Joined:
    Feb 2, 2002
    Messages:
    6,437
    Location:
    Czech Rep.
    well from what I saw players like Kossin like to combine it with Pacifism, so if you have access to religion it's almost always better to wait a bit.

    From my observations (since I try to do it almost every game) you need city with 1-2 high food specials (5-6 food) and around size 10-11. If you have 3 such cities you can sync them to get 3 GP's in 8T golden age, but it's pretty rare for me to get such ideal conditions.

    Mostly I can generate for 1 GP 2 GP's. 3 is really rare for me.

    Actually it seems best to do it after CS to switch to bureau in the GA.
     
  3. vicawoo

    vicawoo Chieftain

    Joined:
    Feb 12, 2007
    Messages:
    3,226
    I'm going to start posting test runs with Mao, immortal. Out of a small amount of pride, let me state this is not a help me play standard/fix obvious mistakes thread, immortal games are a cakewalk if I play standard.
    This is about tweaking if I get a 3rd settler before a library sort of deal.




    Last save I tech code of laws, have my great person, and have a lot of settlers in position.

    Whether I save my golden age for when those cities are grown is a good question.

    Analysis:
    I tend to have a huge surplus of hammers, especially when libraries/fast academy isn't a priority. It's easy to make a ton of workers/settlers, but then I tend to expand way too much, and the added maintenance slows code of laws too much.
    This is by far the biggest problem. There are a few ways around this:
    Oracle. Converting hammers to an expensive tech. Very nice.
    Fail gold. Requires polytheism or masonry, but skipping alphabet is cool.
    Retiming my investments:
    My food cities are drains on my economy, of course. But the strength of expansive is I can time them to be as late as possible. I can get a size 4 off a pig in about 10 turns.
    If it requires a monument, that should take about 14 more turns.
    So perhaps it would be best to settle commerce cities, get one city with a library, then time my settlers so they settle 10 to 25 turns before I expect code of laws to finish.

    Going alphabet before code of laws helps to deal with the huge hammer surplus.

    I had to branch off too many techs, including fishing, hunting (happiness), and bronze working.
     
  4. VoiceOfUnreason

    VoiceOfUnreason Deity

    Joined:
    Dec 5, 2005
    Messages:
    3,663
  5. vicawoo

    vicawoo Chieftain

    Joined:
    Feb 12, 2007
    Messages:
    3,226
    Yeah, this is sort of a natural corollary to the ideas in the thread. The faster you can micromanage your growth, the more cities you can fit in/more you can delay a lot of the food cities.
     
  6. BornInCantaloup

    BornInCantaloup Agent of Chaos

    Joined:
    Feb 9, 2010
    Messages:
    4,194
    Location:
    Cantaloupe Island
    Maybe I read you wrong but I understand this sentence as "get lots of city, then get caste system, and then generate an absurd number of great people."
    Why REx and where does the necessity to REx before teching come from ? That's precisely what gets you a "huge hammer surplus" and makes you research Alphabet (2nd post)

    There are a lot of implicit thoughts in your OP so... It's hard to get where exactly you want to go.
    I think that knowing what to do in what situation is >>> to knowing how to perform it. Maybe it isn't your intention but it seems that, for the sake of quantifying/testing, you narrow down the play options tremendously. To a point that may render your results not so relevant.

    It would help a lot if you mentionned what the bulbs can be used for and whether it is desirable to sustain bulbs for as long as possible or whether there are transitions that can be achieved and aimed for.


    Why no mention of Pacifism ?
    Is it clear cut that GA > Pacifism via Philo bulb ?
    Triggering a very early GA will make it hard to trigger subsequent ones.
    State religion can be relatively well spread.
    Pacifism's effect doesn't fade.
    It doesn't hurt to switch to Pacifism before each city has grown enough to be a GP farm.
    If you're going to bulb with Scientists, then you may as well bulb Philo early.
    Even: can't we consider a switch to Pacifism via the Shwedagon Paya ? I'm sure it's the best option under certain circumstances :lol:


    Also, why so little mention of the Oracle ?
    Isn't it the prime way to get CoL ? Deity aside, a mere Pottery beeline followed by a Priesthood beeline will secure the Oracle very often. I realize this is linked to the REx focus....


    Now, here's my contribution :

    The more GP farms an Empire has, the bigger those GP farms need to be if they want to effectively produce GPs. Put otherwise : GP farms' sizes offset growing GP costs.
    Ratio-ing that relationship seems to be a part of what you intend to do.
    If an Empire has 3 cities, what minimal size the cities should be to offset the growing GP costs ? How long will it last ?
    Same for a 4 cities Empire, and a 5 cities Empire, and a 6 cities Empire, etc.
    Then, knowing what room you have to expand and what resources you can access, you can estimate an "ideal" size for your Empire when you'll switch to Caste.
    Here comes knocking at the door the question of what bulbing strategy is implemented since it will define the number of GPs you want to produce.

    In short : how many cities of what size will get how many GPs in how many turns ? That should be quantifiable rather easily and give usable hints when playing a map.
     
  7. VoiceOfUnreason

    VoiceOfUnreason Deity

    Joined:
    Dec 5, 2005
    Messages:
    3,663
    I'm still having trouble understanding the goal (beyond experimenting with this idea to see where it leads).

    I think part of the confusion comes from looking at the save, which appears to be more geared toward expansion than a great people harvest.

    Also, it's not clear to me that "burning" a GP at the start of the exercise is going to improve your throughput. You are pulling the last GP in by 8 turns (assuming that you can sustain your maximum GPP rate), by pushing all of them out by 6 to 10 turns. Certainly, it makes the engine go faster, but it's not clear that it brings in a meaningful milestone.

    Which ties back to my first confusion: what are you optimizing for?

    Edit: XPost with BIC.
     
  8. shyuhe

    shyuhe Deity

    Joined:
    Nov 30, 2006
    Messages:
    6,062
    Location:
    Gone fishing for the summer
    Opened the save and I laughed at the 4 parked settlers :) I have to agree with VoU - I think you focused so much on expansion with this save that it's probably stronger to settle the cities and spam caste merchants for GMs to power your economy.
     
  9. vicawoo

    vicawoo Chieftain

    Joined:
    Feb 12, 2007
    Messages:
    3,226
    My apologies for trying to minimize explanations, it's my impression that long posts make people skim through it anyway.

    I'm not going to directly answer a lot of the queries, but I'll try to answer through an example:
    If you play a lot of isolation games, you will find a lot of potential specialist cities take too long to set up, and will slow down your astronomy date.

    Also the high level analysis of HR cottages dislikes them because they take 30+ turns to break even. Similar arguments for markets over wealth.

    Obviously there are benefits to extra cities/markets/cottages, but not if they take too long. What we are doing is finding a way to shorten the investment/return time (obvious), and hopefully use that to an impressive effect (pretty open ended).

    It's kind of like in Starcraft 2, they found that queueing your workers to manually gather/return increased mining efficiency by 8%-12%. There are two responses: that's amazing! an extra edge I can leverage, or so what, what am I supposed to do with an extra 8% of minerals. The analog would be, what can I do with a potential method to provide extra great people?

    Pacifism: I don't address pacifism because using it is largely a matter of your diplomatic situation rather than your build. Your usage probably won't affect your overall gameplay too much.

    I decided to just post these games rather than test it for a couple of weeks and cherry pick example games because I thought it might be interesting.
     
  10. vicawoo

    vicawoo Chieftain

    Joined:
    Feb 12, 2007
    Messages:
    3,226
    Since there's no highly appropriate way of doing this, I'm going to make this sort of a test game thread. These games are not played very carefully, like in the previous game, I thought, why let the AI take land, so I stole it.

    Here's Oracle->Code of Laws. My reservations for trying such an approach was unreliability at higher levels.
    However, it was a lot easier than the previous attempt. It's very much goes very well with the themes of spending minimal effort setting up farms. It's very ... pure.

    Once you have code of laws and pottery, you can almost spam your GP farms as fast as possible. There is of course maintenance issues which delays your basic techs, but it's surprisingly manageable when you can
    a) grow very fast
    b) use your population to run scientists merchants.
    The faster you get your cities up, the sooner we can try to generate as many great people as we can.

    The Oracle pollution helps us, since we wanted a golden ager anyway.

    Though we have to revolt to caste, there's no need to get slavery when we can chop our granaries with one forest, and we don't need any other buildings.

    We don't need monuments! Especially useful for those stupid fish.

    Our first great person is too soon to golden age, so I went academy.

    Spoiler :


    A good second city.


    Unfortunately our immortal enemies took forever to tech alphabet.











    Some more possible refinements:
    I didn't run merchants, but it's very simple to manage our farms to get a trade mission at almost any time. It might even be desirable to make our first great person a trade mission to power towards civil service. We can easily get a second great person in time for a possible golden age switch. Could also do a 3rd to power through paper/part of education, then get 2 scientists for liberalism/printing press, more for steel, etc...

    Not sure if academy first or trade mission first would be better.

    With code of laws that fast, we can even do the stupid culture bombing two barb cities at once which I've been planning for awhile. Or you could go merchant first, tech aesthetics/drama and bulb music.

    Seafood sucks. Thankfully other cities can build workboat, but those lighthouses are a pain, they might not even be worth building with double seafood.
     
  11. Mylene

    Mylene Deity

    Joined:
    Jan 30, 2011
    Messages:
    4,000
    Location:
    Pangea
    You are reading way too much stuff into this, and i just picked out this example ;)
    I think you wrote you are not PHI and want to use several cities as GP farm...with the help of expansive (?), but you also don't adress pacifism despite it being the strongest help (besides PHI) for this plan.
    Iam not sure what to say, sorry but just forget this whole idea ;)
     
  12. vicawoo

    vicawoo Chieftain

    Joined:
    Feb 12, 2007
    Messages:
    3,226
    Nah, but feel free to ignore the thread and continue playing your usual way.
     
  13. vranasm

    vranasm Deity

    Joined:
    Feb 2, 2002
    Messages:
    6,437
    Location:
    Czech Rep.
    vic I think you misunderstood me a bit with my initial reaction. I am sorry if I offended you.

    What I had on mind is, that usually the usage of several GP farms comes really late for me (like 300 AD) and would like to hear what exactly should bring using it around 500 BC (as I understood your OP), since basically you want to quickly grow cities, ignore slavery and jump directly into Caste
    (I had one such game with gilgamesh and it was interesting to see 100+ beakers with rep+caste around 100 BC)

    What I would be interesting to hear is, if this can be viable approach with any exp leader and if you would rex like mad (to the point of crash) and run great merchants + scientists in every city for future gameplay to dig out. And I have in mind something like settling 8-9 cities asap (before 1 AD).

    How this approach changes with representation (makes it more worthwhile)?

    I kind of am looking here to broaden the repertoir of generating beakers in early game. My style changes all the time and I want to have it as broad as possible.
     
  14. TheMeInTeam

    TheMeInTeam Top Logic

    Joined:
    Jan 26, 2008
    Messages:
    25,747
    I've played with this idea in the past, and not just with expansive.

    The problem is how to transition out of a very low infra low production setup into something decent later. In other words, how are the great people USED to get a devastating tech position? With bulbs + trades you seem to keep up, but you do that with cottages too. Even under ideal conditions isn't it a little hard to get more than 5-6 great scientists before the 500 AD mark without PHI? That's a nice chunk o beaker but it applies to a very specific path (or one of 2-3) and necessitates trades. What do you do with those gsci bulb techs if your raw research rate once you're in the lib timeframe isn't so hot...or how do you improve it? I don't see any way out of this opening other than rep or workshops.

    IMO this thread is more about the power of expansive in general though; turns of city setup not a particularly obvious cost in-game but its effects can be pretty drastic, especially given the other perks from expansive.
     
  15. VoiceOfUnreason

    VoiceOfUnreason Deity

    Joined:
    Dec 5, 2005
    Messages:
    3,663
    My feeling is that, if you want others to contribute, you need to provide enough explanation so that the reader can make sure he's in the same context that you are.

    "This is a work in progress". Awesome.

    The main plot line seems to be
    1) Peaceful expansion in the early game, in particular prioritizing land claim over infrastructure (horizontal over vertical)
    2) Transition
    3) Broad midgame GP farming via Caste

    Where the end of phase one is essentially that you have all of your cities/settlers in place, and the end of phase two (which is relatively short) is that everything is ready and in place to begin phase 3.

    And in the examples you've shown so far, the final save is essentially at the beginning of phase three.


    On top of that, we're particularly looking at how an Expansive leader can gain extra advantage here. But we're not yet at the point that we have a hard goal that we're trying to optimize.
     
  16. ahcos

    ahcos King

    Joined:
    Mar 19, 2010
    Messages:
    919
    Huh? I thought that exactly those two traits are the ones to heavily support SEs. I'd rather test with Peter/Isabella tbh, those two can really profit from this approach (not only those two, but especially them).

    I really appreciate your work on this, although it seems to me that the whole point here is to showcase people that there is the possibility to do such moves rather than showing how this approach would pay off more than another, more common approach.

    Good job for the work you put into this, but i think it's important to embed this into a game where you really start to show off the advantage. When i'm playing Maya with tons of rivers+grassland, i won't forgo cottages. When i have marble, one dedicaded GP might lift almost the same amount of GPPs/GPs. And on and on and on ...

    I'm almost certain that alot of players don't use CS to it's full potential (including me, sadly), so i'd love to see some kind of "Leveraging Caste System", when you know what i'm refering too (and i'm certain that you do). That said, i'm not too sure about the qualitiy of information you provided here, but there might be alot of potential in it.
     
  17. Mylene

    Mylene Deity

    Joined:
    Jan 30, 2011
    Messages:
    4,000
    Location:
    Pangea
    Look, if you don't want feedback on that what you write is weird...alright ^^
    You are saying things that make no sense at all, like you don't recommend PHI for this idea.
    What the heck? :D
     
  18. VoiceOfUnreason

    VoiceOfUnreason Deity

    Joined:
    Dec 5, 2005
    Messages:
    3,663
    Unfair - vicawoo clearly states he doesn't recommend PHI for testing the idea, for exactly the right reason: PHI is too directly applicable to this type of strat to be a useful measure of the strat's general effectiveness.

    PRO in particular is a good choice for a "stay out of the way" trait, and the Chinese techs are reasonable for "I want to get out of the opening and focus on the early game".


    Clearly, the ideas aren't finished until we have explored how the secondary trait/starting techs influence the main line, but equally clearly it is correct to eliminate those influences while the main line is under review.
     
  19. Mylene

    Mylene Deity

    Joined:
    Jan 30, 2011
    Messages:
    4,000
    Location:
    Pangea
    So he is testing this strategy, with taking out most things that make it useful. Got it ;)
    If it floats your boat :)
     
  20. TheMeInTeam

    TheMeInTeam Top Logic

    Joined:
    Jan 26, 2008
    Messages:
    25,747
    That's not an accurate assessment. GPP is very strong regardless of traits. However, in this case he's specifically looking at the turn advantage from expansive; in doing so PHI adds noise and could bring one to a conclusion that the tactic in general is more powerful than it really is.

    The probable best way to evaluate a tactic is to look at how it functions without any traits/extra factors, and only after that consider how extra factors might influence the tradeoffs.

    If his goal is to see the power of the EXP turn advantage then using mao is appropriate. If the actual goal is just to look at cutting infra in favor of caste GPP ASAP, he should probably be using toku or sid :p.
     

Share This Page