Announcement

Collapse
No announcement yet.

A real reason why tech race looks like me vs World

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #46
    The New Trade Penalty vs Devaluation

    In prior versions, if you met a civ and neither of you had met anybody else yet, they would trade with you evenly. That is, if you started with Masonry and they started with The Wheel, Masonry is worth slightly more, so they would be happy to trade. Now in 1.17f, this is no longer the case at high difficulty. On top of all their other bonuses, they now treat the player as a pariah, giving him worse deals, while trading amongst one another on even terms.

    THIS is the reason why it now feels like Me vs The World.

    What makes it playable is the devaluation. The current tech devaluation is SO severe, that the new trade penalty all but vanishes, as paying double for 1/8th cost, being the last of 8 civs to get the tech, is still only 1/4th of the primary cost. (My numbers are just an example and don't match the actual penalties and devalution, but you get the point).

    The net result is now that on Emperor and especially Deity, there are two forces pulling at opposite ends: the trade penalty, and the tech devaluation.

    I happen to believe that the emtremity of the devaluation is masking the flaws in the trade penalty. Once you are in contact with everybody, it stops mattering any more. BUT... if you are in contact with just one or two other civs, perhaps playing an archipelago map instead of pangaea, the broken nature of the new trade penalty shows through very clearly. In that situation, there is the inescapable sense of unfairness, as the AI's are able to trade freely but they impose the equivalent of oppressive tariffs on any trade they do with the player.

    This is a problem! It pigeonholes the strategy into "Make Contact At ALL Costs", because the deciding factor in the game is how many other civs you're in contact with, and that applies to the AI's as well as the player.

    I'm afraid the game balance goes out the window. Pangaea becomes simplistic, while anything that prevents you from making contact with lots of civs becomes pointless or frustrating. And yes, I agree with those who are saying this speeds the total pace of the game up too much, at least until Industrial, when wars will plunge most of the civs into the one government sure to sink their economies: Communism.

    (Communism is SO WORTHLESS and broken that Anarchy is sometimes a better alternative! Isn't Communism supposed to be more attractive for sprawling empires than other governments? It's not. Here's proof if you want it, complete with save files if you want to compare Communism with Monarchy, 1820AD vs 1830AD. Click Here The problem with Communism is that corruption contines to climb as you add more cities, and that defeats the purpose of spreading the corruption around. The net result is that distant cities are still worthless, while the home cities are also penalized. Under all the other governments, at least the home cities stay strong! In fact, I'd say that Communism's broken state is more responsible even than Railroads for the pathetic performance of AI's post-industrial -- I'm getting off topic, though).

    The new Trade Penalty for high difficulties is a step backward. I have found that exploiting the devaluation, and riding the coattails of the AI's until I'm strong enough to make a play, I can cope with it and still do well, even on Deity -- all with no exploit play.

    BUT... that just shows up another problem. The devaluation is too strong, also. It renders the tech race nearly moot. Keeping a tech lead is too hard, and recovering from a tech hole is too easy. In 1.07, there was no point bothering with research in the early game, just buy everything. 1.16 fixed that, but now it's back to being broken again, as the best move is to ignore research, wait until you make contact with everyone, then buy yourself into tech parity on the ultracheap. Doing your own research IS A LOSING MOVE and this has to be considered broken.

    If you research while isolated, you may need to do some research to have things to keep building. Need harbors, libraries, courthouses, maybe aqueducts. If you have nothing to do, that's more wasteful than doing research. BUT... short of that, doing your own research is foolish for this reason: tech devaluates, but coin does not. The wise player then saves everything into coin that he possibly can, then uses it to buy into tech parity. Any surplus can be used to rushbuy improvements or upgrade troops. The game is playable this way, but this can't be how it was meant to be!

    I'm of the opinion that tech devaluation itself is a flawed concept, precisely because of that factor that research devalues in the game but coin does not. In REAL life, both devaluate. Obsolete tech becomes cheaper and readily available, but inflation also eats into savings, reducing the value of coin in a similar way. Civ3 doesn't account for this, which is why it has become a tech-whoring game end to end, in various ways. Human players are smart and will find all the loopholes for "storing" their economic gains in the areas where they won't be devalued as much.

    Soren, if you ever want to tech whoring to stop, you will have to discard or at least dramatically change the tech devaluation system. It seems to me that immediate devaluation is a problem. In real life, obsolete tech only devaluates when the tech leader is several advances ahead. That is to say, a software company (for example) has no problem releasing source code to a program from five years ago, but NO WAY would they release it instantly just because a few rivals have developed similar programs. Do you follow that analogy in pertaining to the flaws in the instant devalution of Civ III techs?

    I know it's not that simple, as the tech devaluation leads players to earn less from brokering techs to everybody. But there is also a flaw in the 100% willingness of the AI's to make trades. The AI will never turn down a trade if the buyer can pay their asking price. (Well, except if at war). They will never refuse to buy from the player, no matter how slimy his reputation gets. They never ever withold techs when in the tech lead, and it would actually be unwise for them to do so. The very fact that tech devaluates means that it would be foolish to hold on to it (economically). If you're going to sell it, sell it while the price is still high! If you wait, you lose the chance. This is the fault of devaluation. If not for the devaluation, the AI's could afford to turn down deals, try to be stingy with their tech now and then.

    My jaw would probably FALL OFF if I saw an AI refusing to trade Music Theory to anybody while they sit and build on their Bach's Cathedral. If players are tech whores, they learned it from the AI, as the AI's will sell anything to anybody at any time, IF the price is right.

    Whatever else you may decide, please repeal the Trade Penalty. The game was fun when civs I met (on high difficulty) were willing to deal with me fairly, or maybe with a slight bias. It's enough that they start with free units, have been given discounts on food, shields, and trade, and that the player has some handicaps too. But adding on top of that that the AI's shun the player and rip him off, accepting only 3-for-1 kind of deals in their favor while the Good Ole Boys AI Network trade with each other... is pointless, definitely NOT fun, and drives the player to have to exploit some trick or other, like shutting down all research to avoid any devaluation, poprushing, or city-trading exploits.


    - Sirian

    Comment


    • #47
      Sirian,

      The Trade Penalty of which you speak... I assume it is responsible for what I see every time I start a new game (monarch level):

      I'm Egypt. 10 turns or so into the game, I meet another civ. I have Masonry (a valuable wonder tech) and Cem. Burial available for trade. They have 2 other techs. I offer Masonry and Cem. Burial and ask what they will pay, and the response is 10 gold. If I ask for a tech (say, warrior code or bronze working) in exchange for my TWO, they will apparently be insulted.

      That is frustrating. It has always been there, but I think 1.17 has made it more pronounced (I think 1.16 would have resulted in a deal for my two techs in exchange for 1 tech + 10 gold, which still indicates a penalty, but not as harsh).

      -Arrian
      grog want tank...Grog Want Tank... GROG WANT TANK!

      The trick isn't to break some eggs to make an omelette, it's convincing the eggs to break themselves in order to aspire to omelettehood.

      Comment


      • #48
        ok i wanted to know more about how the AI values trades, so i did a little bit of testing

        first i gave both the romans and the egyptians 10,000 gold, then i gave the egyptians masonary and i made the romans expansionistic and gave them a 50 movement treat all terrain as roads scout

        i changed the tech cost of masonary to 10 and the world size tech cost to 100

        so here is what the egyptians charged me for masonary

        70
        130
        130
        130
        140
        140

        that's chieftan through diety

        now i switched so that the romans had masonary and here is what they offered me

        20
        50
        40
        40
        40
        40

        again that is cheiftan through diety

        i'll will do a few more tests and report back

        Comment


        • #49
          korn,

          I assume that, for the purposes of this example, you are playing either the Egyptians or Romans, right? And that contact is made in 4000bc due to the 150 moves your scout can make?

          -Arrian
          grog want tank...Grog Want Tank... GROG WANT TANK!

          The trick isn't to break some eggs to make an omelette, it's convincing the eggs to break themselves in order to aspire to omelettehood.

          Comment


          • #50
            I find myself agreeing with Sirian's excellent post and also lament the fact that players have to resort to narrow or exploit-oriented strategies at higher levels in order to have successful games.

            Comment


            • #51
              Arrian

              i played as the romans each time, and i made contact in 3950 so both sides would have a city

              a little bit of further testing...

              i changed the tech cost for masonary to 100 and then the egyptians offered 430 gold for it

              i reset masonary to 10 and changed the world size tech cost to 1000 from 100 and they still offered 430 gold for it, so it seems that the uses the multiplied total of tech cost times world size modified by difficulty level and then devaluation to determine the value of the tech it is trading for

              Comment


              • #52
                ok here are some more tests results

                *i set the tech cost on masonry, alphabet, and the wheel to 100, and i then set the world tech cost modifier to 10
                *i set the minimum number of turns to discover a tech 1 and the maximum number 100
                *i then gave masonry to the egyptians and the greeks
                *i then gave alphabet to the egyptians
                *i changed the tile bonus so i could allocate up to 20 gold per turn to science
                *i changed the starting treasury to 10,000
                *i then gave roman uberscouts
                *on 3950 i made contact with both the egyptians and the greeks, and i hadn't accumulated any science yet
                *i played this on diety

                at 2 gold per turn to science
                to research masonry would have taken me 33 turns (66 gold)
                to research alphabet would have taken me 67 turns (134 gold)
                to research the wheel would have taken me 100 turns (200 gold)

                at 10 gold per turn to science
                to research masonry would have taken me 7 turns (70 gold)
                to research alphabet would have taken me 14 turns (140 gold)
                to research the wheel would have taken me 20 turns (200 gold)

                so this means that the techs actually cost 66, 134, 200 gold

                to buy masonry at offered price would have taken me 22 gold
                to buy alphabet at offered price would have taken me 80 gold

                to buy masonry at lowest negociated price would have taken me 20 gold
                to buy alphabet at lowest negociated price would have taken me 73 gold

                conclusion:

                the AI sells tech too cheap, and that the optimal strategy is to always buy devalued tech

                Comment


                • #53
                  conclusion:

                  the AI sells tech too cheap, and that the optimal strategy is to always buy devalued tech
                  On Deity that holds true. On lower levels, even Emperor, its readily possible to play efficiently enough that in most games, you can reach a point where you outproduce the next best civ by SO much, you can take and hold the tech lead. On Emperor, this often does not happen until Industrial age, when they go to war with one another (sometimes takes a little nudge from you, but usually not) and end up in Communism.

                  Communism is so broken, that getting the AI's into it is just as much an Always-Winning move as is building markets and banks and not bothering with research while you're trailing in production power. Except for my very first game ever of Civ3, I have yet to lose a game after surviving to the industrial age. Not only is Communism inherently weak economically, but the AI's poprush WAY too much. Way WAY wayyyyyyyyy too much. And why do they have to do this? Because you come close to one of their cities, which is invariably defended by just a couple of units. You don't even have to DO anything, just move a couple of warriors into range of cities and watch them draft or rush, shooting themselves in the foot. Park a few infantry on a hill and watch the AI civ scramble into panic mode. And my goodness, with the new 40 turn penalties, the AI's are now set to permanently cripple their cities at the first sign of any threat. They are damned either way. I've taken over cities that the AI has whipped 30 times, or 30 pop worth of whippings left on the memory. Something is wrong with that.

                  But... even without Communism, even if they opt to stay in representative governments and you don't conspire to start wars, you can still reach a point of outproducing the AI's anywhere below deity. Once you reach that point, as difficult as it may be, it's a winning move all the way to maintain a tech lead. Even two to four advances ahead can do the trick, as you get your rails online sooner, or you send your cav against muskets, or your tanks against rifles and infantry. At that point you can translate your advantage into more and more advantage.

                  In that regard, the tech devaluation keeps all the AI's close and prevents the kind of tank vs stone age gap we often saw in Civ1 and Civ2. It prevents the game from becoming a cakewalk once the player gets a production edge. But it also pretty much renders a lot of the game pointless, since the system is SO bent toward maintaining parity and keeping every civ in the game bunched in one clump, techwise, that the course of the gameplay falls into a narrow range. Frankly, it's too predictable. You can glance at a few factors and, within a reasonable margin of error, project the outcomes. The black fog masks some of the flaws in the game. Not knowing where the other civs are, or the exact lay of the land and the resources, is what keeps the game alive for me. If the map were revealed from the start, the ONLY element of the game that introduces significant uncertainty would be gone. There is still a goodly range of things the player can do with the game enforcing such extreme parity onto all the civs, but like the size 7 food preservation or the overriding contentment, the tech devaluation has some flaws that wreck the game balance. So, too, with Communism.

                  It would be an interesting experiment to remove Communism from the game, and lower the draft penalty down to five or ten turns, and see if the AI's would do better. I'm rather certain they would, and dramatically so. If that's the case, what's the fix? I don't know. Several possibilities. Right now, the bigger your civ gets, the LESS attractive Communism becomes. The whole deal with communistic corruption continuing to grow and grow and grow as you accumulate more cities is the main problem for the human player. The AI's it seems to be that plus that they whip and draft too much.

                  The AI got some impressive upgrades in the last patch: they now use artillery offensively (still in the pairing approach, an artillery piece and a defender, but it's MUCH better than not using any at all). Their priority on attacking workers is so much improved, it probably doesn't need any further attention. Overall, things are moving in the right direction, but still a few major loopholes left to plug.

                  Deity as it stands now might be rendered unplayable without the tech-devaluation loophole, but that's a separate problem and should not be the reason for not addressing the loophole presented by tech devaluing improperly relative to coin.


                  - Sirian

                  Comment


                  • #54
                    Hey Soren,

                    Have you had a chance to look at that saved game?

                    -Arrian
                    grog want tank...Grog Want Tank... GROG WANT TANK!

                    The trick isn't to break some eggs to make an omelette, it's convincing the eggs to break themselves in order to aspire to omelettehood.

                    Comment


                    • #55
                      good work guys...

                      Best thread, you folks have it pegged. I ever so much wanted to post here while Ming had me on vacation, but now there is little more to add.

                      I expect Soren is working on fixing this mess.

                      Comment


                      • #56
                        As far as I can see the best solution would be to delay the tech devaluation, ie make the minimum value dependent on the number of turns passed since the new tech was first discovered (or, even, since it was obtained by a civ you are in contact with). Also, devaluation should indeed happen at the end of (your) turn. This way:
                        (a) It doesn't matter to which civ you sell a tech first during the same turn;
                        (b) Techs won't taht easily be (re)sold by the AI to all other civs once one of them has it;
                        (c) Setting the science rate at zero (until the final burst) is still a viable strategy, but it is risky military-wise and you'll have to plan well ahead to get key wonders and resources;
                        (d) As a consequence, at higher levels deals with the AI can be less unfair.

                        I'd say it should take 8 turns before full devaluation can be realized. That is, if you lag behind and want to buy tech as cheaply as in version 1.17, you'll be a couple of techs behind at all times. This will make all the difference.
                        A horse! A horse! Mingapulco for a horse! Someone must give chase to Brave Sir Robin and get those missing flags ...
                        Project Lead of Might and Magic Tribute

                        Comment


                        • #57
                          Just confirming

                          Just to confirm that spending NO % on tech is a very winning strategy. I started a game in Emperor mode, went into 100% gold mode, and never actually spent a single coin on those light-bulbs. Result? Because of all the trading everyone is fairly peaceful towards me (only had one war to boot the roving English out of my continent), I'm now making about 150 gold a turn, I am always 1 or 2 techs behind the tech leader, but no more, as even when they ask for ridiculous price I can buy it and make back some of it with the other civs usually (eg Buy Democracy for 1000 gold or so and sell it back for about 300-500 gold to other civs). The only times when I can't make it back is when everyone already knows it - and when that happens, well the tech is damn cheap so it doesn't really matter.

                          So yeah... 0% research is definitely a winning strategy at the moment.

                          Daniel
                          http://dwdt.xs.mw

                          Comment


                          • #58
                            Forgot to add...

                            Btw, I'd just like to say that in my opinion the whole tech system is flawed from the start. I would remove tech devaluation completely and introduce tech leaking, which is much more realistic. Tech leaking should mean that every turn, you have a small chance of gaining each of the technologies which the people you are in contact with own (and they have a chance of gaining yours). How so? Simply through people travelling around, passing the word, etc. The chance would increase as the tech gets older (and it should become somewhat cheaper to trade for, too, but still not that cheap). But actually buying the tech should still cost bucketloads - so if you want to do no research, you're gonna be either poor or very patient.

                            How is this better? Well, it's more realistic for a start. That's the main way technologies (at least until various agencies appeared to keep secret techs secret - perhaps introducing a small wonder that allows you to keep X of your techs totally secret would be a good idea). How did the printing press spread? No one went and bought it from Gutemberg's country. People heard of the idea, travelled, then other people tried to do it too and some succeeded and some failed. The same goes for most of the techs in the tree. And academic subjects like Atomic Theory tend to be developed by researchers who publish/share everything the minute they discover it. And it's that way today still, for the majority of research (exception made for government/army sponsored secret research, which should definitely be allowed but not the only way of doing research).

                            The only time when leaking doesn't happen that much is:
                            1) when you're at war - but even then some techs leak as you capture/examine/reproduce your opponents' artifacts.
                            2) once copyrights and patents get invented and implemented in a super-restrictive way. This could be the object of a great wonder, which would make leaking of technologies to everyone instantaneous but make everyone have to pay you for the privilege of using tech you've developed - unless they're at war with you. Maybe it could work that way for everyone, but give more money to the civ with the Patent Agency
                            3) When researched and kept in a secret lab (small wonder?).

                            That's all. Hope something like this gets implemented sometime... if not in civ3 then in the next civ. But seriously, game mechanics are important, and must be balanced, but the tech devaluation trick is cheap and very unrealistic.

                            Daniel
                            http://dwdt.xs.mw

                            Comment


                            • #59
                              I´m pretty sure everyone would hate your system, KDan. People are already complaining over unlucky combat results - how do you think people would react if they notice that the neighboring civ gets a vital advancement leaked to it 20 turns before the human player? Almost everyone wants a game with a clear, fixed, intuitive set of rules. Your idea is mostly the opposite.

                              Comment


                              • #60
                                Vital advancement?

                                As I said, I've been playing this game with no tech development, emperor diff. I get all my tech after most of the AI (cause that's when it's cheap). Apart from wonder-building, the only time when you could feel bad enough about a tech to call it 'vital' is when you're at war and you need the tech edge, or when you want to build a wonder. If you wanna fight people and/or build wonders, tough luck. Do your own research. Plus, assuming my idea of a 'top secret labs' small wonder, if you really want to you can keep some of your research secret in the later eras when it really matters, and so can the AI's, if they consider that it really matters. So there will only be 'vital' tech leaking in the ancient and medieval ages, mostly.

                                By the way, now that you mention this vital tech thing, I'd like to point out something which has always bothered me about the civ games. Namely the way that getting into a war actually slows down your economy and your tech research because you have to divert resources to build/upgrade units. That is the most unrealistic aspect of Civ (all versions). In reality, war is extremely good for both business and research. And that's not counting stuff gleaned from the enemy towns (which always seems to be pathetic now, since there's no techs in cities in civ3 and who gives a damn about receiving two gold after capturing another civ's capital!).

                                War is good for business. It massages the economy like nothing else does. Why do you think the US economy is picking up again? Because Bush got the army budget increased by a nice lump. And war is good for research. Research was buzzing like mad everywhere during WW2. Some even say that it was research that won the war rather than plain troops. And that was while all these countries were building fleets, tanks, rifles, training soldiers, spending all kinds of money that could probably not have been better spent elsewhere if they wanted mroe research to go on...

                                So second idea - I reckon that war should boost your commerce by some margin - like about 10%. Or maybe not war - mobilization. It should boost not only your production, but also your trade so that you research stuff more quickly. Why? Because that's the way it is out there...

                                Daniel
                                http://dwdt.xs.mw

                                Comment

                                Working...
                                X