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

80% science stupid?

Discussion in 'Civ4 - Strategy & Tips' started by jeremiahrounds, Mar 15, 2006.

  1. Willem

    Willem Deity

    Joined:
    Feb 12, 2002
    Messages:
    7,313
    Location:
    Canada
    You're probably losing in the long run since by having your slider at 100% research, you miss all the percentage bonuses for Marketplace, Banks etc. I discovered that after 80% there was no real value in having those in a city, it was just wasted production.
     
  2. Wodan

    Wodan Deity

    Joined:
    Jun 3, 2005
    Messages:
    4,867
    Location:
    In transit
    Seems to me this strategy is much more valuable early in the game but late in the game it's a losing proposition. Early on, you don't have any commerce (gold) or culture-enhancing effects, from buildings, wonders, civics, etc. Also, I suspect that the loss due to rounding is more pronounced when the number being multipled is smaller (e.g., a city that produces 200 beakers vs a city producing only 7 beakers.) But I'm not sure of the latter.

    Wodan
     
  3. DementedAvenger

    DementedAvenger Prince

    Joined:
    Sep 6, 2004
    Messages:
    435
    This is exactly why this strategy is so useful. If you run 80% science the whole time and hover at 0-10 gold, you get the tech in the same amount of turns as if you shoot up to 200 gold first, and run 100% science, losing some of that 200 gold every turn.

    Wouldn't you rather have 200 gold than 0 gold on any given turn? It's a safety net for unit upgrades or rush-buying. This strategy isn't about exploting rounding errors, it's about having a higher bankroll (on average) with no loss in tech research time.


    And to those who say it loses effectivenes in the later game, it seems to me that it is most powerful then - instead of shooting up to 200 gold, you'll shoot up to 2000 gold, and will be able to massively upgrade/rushbuy should an emergency occur.
     
  4. Zombie69

    Zombie69 Emperor

    Joined:
    Nov 22, 2005
    Messages:
    1,898
    Actually, you can only lose from rounding, never win. This is because in civ, everything is rounded down, never up. For example, a city with 7 base science and a library will produce 7 * (1 + 0.25) = 8.75, rounded down to 8 beakers.

    Wrong again. There is no gold minimum, but there is a minimum of 1 beaker added to research. However, it has no bearing on the calculation since it's always there no matter what science rate you use.

    Wrong again. You'll use the market a lot more if you use them at 100% gold once every 10 turns than if you use them at 10% gold every turn, because you'll be losing less gold to rounding.
     
  5. Willem

    Willem Deity

    Joined:
    Feb 12, 2002
    Messages:
    7,313
    Location:
    Canada
    But you can't use it for unit upgrades since you need to spend that surplus on the 100% research cycle. If you use it for anything else, you'll be losing out on the tech you're after at the time. It seems to me just alot of micromanagement for no real benefit.
     
  6. BlueRenner

    BlueRenner Warlord

    Joined:
    Nov 9, 2001
    Messages:
    121
    Actually, if you plan to run it like this, this strategy is a clear loser.

    You have to pay Inflation on all that gold you keep in your treasury, which reduces the efficiency of the entire system.

    This isn't to say its a bad idea -- flexibility vs efficiency is a valid argument. However, you will be teching at a slightly slower pace than a person who keeps 0 gold in their treasury.

    - Bill
     
  7. Zombie69

    Zombie69 Emperor

    Joined:
    Nov 22, 2005
    Messages:
    1,898
    No, the maximum is 1 beaker per turn per city. With many cities, this really adds up.

    It doesn't work like this. First of all, the rounding down occurs for each individual city before multiplying them each by their respective bonuses, and only after all this adding them up together.

    Also, the one free beaker you get is added after all this is done, and is added no matter what your science rate is, so it makes absolutely no difference whatsoever in the calculation and can be totally ignored from the equation!

    Very wrong. I don't think you unerstand the concept here, and the math behind it. On a tech that takes 30 turns to achieve, using a binary science rate could effectively shave off 1 or 2 turns, sometimes even more.
     
  8. DementedAvenger

    DementedAvenger Prince

    Joined:
    Sep 6, 2004
    Messages:
    435
    In most cases, you won't be using it. The point is that you have the OPTION to use it. I'd rather lose a turn on a tech and keep a city then not have the option at all. The strength of this is FLEXIBILITY.

    Flexibility is huge. If you keep enough money on hand to upgrade 2 archers to longbows, this saves you from having to build 2 longbows per every border city - you can just keep the archers there and upgrade when needed.

    That's a good point about inflation, though (if inflation really is tied to treasury). Do you have any sort of idea as to how much inflation rises based on treasury amount?
     
  9. Zombie69

    Zombie69 Emperor

    Joined:
    Nov 22, 2005
    Messages:
    1,898
    I disagree, it's a very real benefit.

    Let's say that you don't believe in this system and always run the highest science rate possible while always keeping your treasury as close to zero as you can. Let's say that you get attacked and would like to be able to upgrade some units but have no money to do so. Further, let's assume there was a game mechanic through which you could borrow the money for a few turns. Would you use it? I'm sure that in some cases, you would. Upgrading a certain unit might very well save one of your cities, which is well worth the little bit of science lost. However, you can't do it, because there is no such borrowing system in place.

    You see, a binary science rate gives you that luxury! Of course, most often you won't use it, but there will be time when you will, and at those times it will be very much worth the loss in science, otherwise you wouldn't have used it now would you?
     
  10. Zombie69

    Zombie69 Emperor

    Joined:
    Nov 22, 2005
    Messages:
    1,898
    As far as i know, inflation has nothing to do with how much gold you have in your treasury, and depends only on how much gold you're making. Over the course of 10 turns (or however many turns it techs to research a tech), you have the same expenses, and therefore the same revenue used to cover those expenses, no matter what your science rate is, especially if after those 10 turns your treasury is back down to 0 either way (thus you had the same total revenue).
     
  11. BlueRenner

    BlueRenner Warlord

    Joined:
    Nov 9, 2001
    Messages:
    121
    Nope, nope. Given that I tested this before I posted, and just now to make sure I'm not crazy, makes me stick to my guns. I am afraid you are the one who is wrong, oh sophmoricly sexual zombie.

    Here is the test.

    Start up a new game on Noble. Settle your first city, crack open the world builder.

    Give yourself a pile of gold, so you can run 100% science for a few turns.

    Give yourself another city. Give both your cities five or six population points. Give both cities a Library and University and Globe Theater. Throw down a few farms, dot the landscape with a variety of evolved cottages. Quit out of the builder.

    Arrange your citizens in your two cities such that they produe an even number of commerce. In my case, it was 24 (+50% = 36 beakers) and 10 (+50% = 15 beakers). This means that, according to that filthy lying Domestic Advisor, I should be producing 51 beakers per turn.

    If only.

    So, pick a tech you want to test on -- one that has zero beakers in it already. Hit 'next turn'. Mouse over the science bar at the top of the screen to find out how many beakers you actually produced -- in my case it was 61.

    Right.

    Now arrange your citizens in your two cities so that they make an ODD number of commerce, one greater than the turn before. In this case it was 25 (+50% = 37.5 beakers) and 11 (+50% = 16.5 beakers).

    Again, that lying pigdog of a Domestic Advisor says I should be making 53 beakers per turn. I should have him shot.

    Hit next turn. Pick a new tech or stay on the one you have. Either way, note you have made +1 beaker from the fractional addition from each city. I made 64, rather than 63.

    In short -- that tiny fraction of a half or quarter-beaker is only rounded down AFTER all your tech from all your cities are summed. This means that all this dancing about with the science slider can only, at very best, net you a single beaker every turn.

    Which elicits a huge yawn from me. Yeah, maybe it can make a difference after 300 turns or so... but then you're only talking about maybe 200 beakers, at a time when techs cost 8000 or more.

    If you can devise a test to show that you are in fact correct, please elaborate. But until then, I stand by the above assertion.

    - Bill

    -- edit -- botched a few transcriptions. Corrected.
     
  12. DementedAvenger

    DementedAvenger Prince

    Joined:
    Sep 6, 2004
    Messages:
    435
    I think there's a communication error - I thought the strength of this strategy was flexibility, not saved beakers. Am I wrong?
     
  13. jeremiahrounds

    jeremiahrounds Prince

    Joined:
    Nov 3, 2003
    Messages:
    404
    I started it as flexibility [with a hint of changing opportunity cost dictating either a pure research or a pure coin advantage] however then someone on the second posted pointed out that a more sophisticated analysis concluded because of round off errors you are much better to do the 0/100.

    So its evolved quite a bit.

    My practical adaptation is to just 100% research until you go broke then 0% research for 1 turn rinse and repeat.

    Im not the best player. My general philosophy though is not to be quite because "if you open your mouth you remove all doubt". But rather to speak cause nobody can ever show you your wrong if you never say what you think. So thats the spirit of my post.

    In the end though we are up to two easy to spot advantages
    1) avoid round off sapping your research
    2) have a more flexible bank account

    with 0 down sides.

    So it sounds like something to do.
     
  14. Willem

    Willem Deity

    Joined:
    Feb 12, 2002
    Messages:
    7,313
    Location:
    Canada
    But if you need to use that surplus during the 100% research phase, you only have the illusion of extra money. If I'm running at 80% research, which is what I aim for, I can still drop to 0% to upgrade my units if I choose. But since I'm accumulating up to 20% each turn towards my treasury, I don't usually need to.

    Plus you still haven't addressed the loss of the commerce bonuses you get from Markets/Banks etc. If you're running at 100% research, you're getting nothing from those buildings so you're losing income each time you're in the research cycle. So over the long term you're going to be losing money.
     
  15. Zombie69

    Zombie69 Emperor

    Joined:
    Nov 22, 2005
    Messages:
    1,898
    If what you say is correct, then that indeed changes everything. However, i think it's a lot more likely that that following happened.

    In your first turn, your research before bonuses was 51. Added to that was a 20% bonus (or thereabout) for having prerequisites for the tech or from knowing civs that already have the tech. This bonus has already been documented to be invisible in the game and so it was "normal" for your advisor not to tell you about it.

    Adding 2 more total science may have in fact added 2 more science, plus the invisible bonus from those 2. By that i mean that with 51 science, an approximately 20% bonus may have added 10 more. But from 53 science, the same bonus could have added 11 science, since it was applied to a broader base. This bonus is of course applied after all the cities with all their individual bonuses have been added up.
     
  16. Zombie69

    Zombie69 Emperor

    Joined:
    Nov 22, 2005
    Messages:
    1,898
    Not the illusion, the option.

    True, you could do that. But what if you can't afford to wait one turn before upgrading and need to do it NOW because you're under attack. Having the option to do it now rather than one turn from now may well save your day.

    If you're accumulating gold just in case, then you're truly wasting science, while we're not, and you'll get your tech after us. Why? Because on the turn we get our tech, we're (ideally) down to zero treasury. Only on the other turns do we have any money in the treasury.

    No, YOU are losing the commerce bonus! Because the bonus is greater if applied to 100% once every 5 turns than if applied to 20% every turn, for one simple reason : fractions being rounded down.
     
  17. Mutineer

    Mutineer Deity

    Joined:
    Feb 20, 2006
    Messages:
    3,715
    Well, cost of tech is not = amount of beakers you spend.

    Yes, it is cheaper to research tech fi more AI allredy know it.
    From othe rpoint of view, if you first to know it you can get unique benefits, like know-connect-use-sell resources, get special benefits like free GP or Tech, trade tech a few times getting mach more out of beakers you spend.
    So, there are a lots of reason to get research done ASAP.

    So, if Tech take 6 turns on 50% - there are no reason to change.

    However, there are times when I did what you suggested.

    Example: I had a supper sci capital (all cottages and biggest town) and lots of smaller towns. Capital has Academy, all sci building and in future Cambridge.

    I want to go to war. I switch to nationalist and draft my towns, not touching capitals. Research set to 0, as in nationalism money produced by all towns equal.

    Switch back to burocracy. Research switch to 100%, as my capital getting 50% bonus which converted my collected money to beakers more efficiently.
     
  18. Zombie69

    Zombie69 Emperor

    Joined:
    Nov 22, 2005
    Messages:
    1,898
    You can't trade all techs all the time. There is a set limit in the game on how many techs you can trade total before the AIs flat out refuse to trade with you anymore. This has already been discussed in another forum. Therefore, it's best to be last to techs that don't give any advantage to those who get them first, and first to those that do give advantages. The binary science rate is perfect for that.

    But even disregarding the above, a simple version of the binary science rate, where you always time it so that you reach zero treasury on the very turn when you discover a tech, for every tech in the game, will never get you any tech later than you would otherwise, and will sometimes get you those techs sooner, if that's all that you're aiming for.

    The system allows you to target the techs that you want to reach asap. If you define that as all the techs in the game, it allows you to do that too. Adap to your own needs.

    If applied properly, there is no downside whatsoever to using it, but a whole lot of upsides (not losing fractions, getting more bonuses on research, giving less bonuses on science to opponents, having money in the bank for emergencies, and probably more that i'm forgetting at the moment). The only downside is that with a lot of money in your treasury, opponents can ask for more. But if you're like me and like AIs declaring war on you, because it gives you free XP and a city or two, then that's not even a downside either!
     
  19. jeremiahrounds

    jeremiahrounds Prince

    Joined:
    Nov 3, 2003
    Messages:
    404
    On the first point, the illusion of more money in this case is also the reality of more more flexibility.

    On the second point, that just seems wrong to me. The reason is the bonus is percentage based and not additive.

    So lets play with fake empires to make sure we have the correct picture.

    Our empire makes 100 a turn prebonus. It has a building that gives 50% bonuses to tax and a 25% bonus to research. We have bills of 15 coin a turn. We have a ten turn horizon. For our purposes fraction amounts will be accumulated [since were just analyzing whether this bank effect works as i describe and what the illusion of more money is worth]

    One approach is to put the slider at 90% for 10 turns. In this case you make 15 coin a turn. If your starting savings was 0 your average saving for 10 turns will be 0. Your net research will be 1125 over 10 turns.

    The other approach is to put the slider at 0% for one turn. Your average savings then will be (135 + 120 + 105 + ... + 0) / 10 something like having an average coin of 65 a turn onhand. Your net research will be 1125 over 10 turns.

    As you can see there was no "not using banks" penalty. [because with a percentage it doesnt matter in what turn you do the gain]. And your average savings was much higher.

    Whats average savings worth? Its worth the ability to change plans with more agility.
     
  20. Willem

    Willem Deity

    Joined:
    Feb 12, 2002
    Messages:
    7,313
    Location:
    Canada
    I just did the math and you're wrong. With a single city producing 100 commerce with Market, Bank, and Grocer, 1 turn out of 6, which you are suggesting, will give you 200 gold. However at 80% research rate, or 20 commerce per turn + bonuses, over 6 turns will give you 240 gold. So you're losing alot of money with your system.
     

Share This Page