Crippling the AI's research capability

[Wislem]

Has anyone else made the discovery that my friend has made? The internal value that the AI has assigned to technology with respect to trade is so high that you can cripple that Civ's economy by asking for crazy amounts of gold per turn.

The game plays out like this:
1. Build lots of cities, just grab as much land as you can while keeping your military entirely defensive
2. Try to meet every civ you can and trade of tech that you don't have. When you encounter a civ that wants a tech you have and they have no tech to offer ask for as much gold per turn as you can.
3. Everytime you research a new tech trade it to every civ, they'll trade it amongst themselves if you give it to one so give it to all of them at the same time. Always ask for per turn gold, nothing else. The caveat is of course not to trade something that could threaten you, e.g. allow them to build a wonder before you or give them a defensive/offensive unit you can't counter. Also if they get a tech that you want trade for it and immediately turn around and trade it to all the other civs, for an exhorbitant price of course.
4. There will come a time when their research is basically stagnant and they devote all of their economy to paying you off every turn.
5. When you've jumped 5 to 6 techs ahead of everyone, be selective about which technologies you release to "the unwashed masses", e.g. give them lots of dead-end technologies with wonders you've already built. Because of the way CIV 3 is structured there are a fair chunk of technologies that give no benefit except to open up other areas of research, giving those away is pretty harmless but gives big benefits.
6. Win the game any way you want. The positive side effects of this strategy are:

- The other CIVs love you, every successful trade pulls your reputation up another notch and makes them that much more receptive. Playing them off on each other through military alliances becomes easier because of this diplomatic advantage. Also the UN victory becomes that much easier
- You have all the tech and essentially control what tech the other CIVs can and cannot have. When you want to go to war simply withhold a few tech advances and watch your modern armor crush their piddly musketmen.
- Your economy will be strong enough that you can devote all your income to research and luxuries. In essence the other civs will be paying your maintenance costs and providing you enough income to steal any technologies you need, hurry any production you want and upgrade all units in a single turn.


It seems that the AI is willing to make big sacrifices in order to gain a technology advance, the sacrifices can actually be big enough to cripple their economic and scientific growth for 20 turns. There is a limit to how much the AI will pay you, but this is the absolute limit based on how much all of their colonies are actually capable of producing turn by turn.

After playing it out a few times, I have found no circumstances, given a reasonable starting position and accessibility to all the other civs, that this strategy will not ensure victory. Question is, is this an exploit? If so obviously Firaxis should take steps to correct it. If not, does it need to be adjusted in some way? It seems a little too easy and a little too powerful, but well within the game rules, so I'm not sure.

 

[Wislem]

Comments please. I can see you guys reading it by the "number of times viewed" so I know you're there!

 

[LKendter]

This post has been around < 1/2 hour and you asked where the respones are?

I also try to keep the computer broke.

There is another important aspect here. A broke computer player can NOT upgrade his units. I will also take lum sums of gold. 200 gold during the late middle ages / early industrial is very valueable. I can start upgrading to knights / calvary and Musketman / Rifleman.

This exploits a weakness of the AI. It cost the same amount to maintance a Warrior as it does a Modern Armour. The computer counts units, not quality!

 

[eyrei]

I am not sure what I think of that strategy. I am sure it works, but it seems kind of cheap and unrealistic. It sounds like a good strategy to implement (if you can) on Emperor or Deity level, since it is necessary to take advantage of every hole in the AI to counteract the huge bonuses your competitors get.

 

[tetley]

No, no comments. It's good. The only thing I disagree with is keeping your military entirely defensive. Well-calculated offensives pay off.

 

[Wislem]

I have noticed the same thing about unit values. It's one thing for the power graph to show a higher power because of 10 warriors to 1 mechanized infantry, but imho the AI shouldn't actually BELIEVE that it's better!

 

[Wislem]

I agree tetley, but I have found that in the early going, that, though squashing a new civ is very profitable, it's just too much of a headache since victory in a fight at the beginning is too random. I am always extremely put out when they build a single spearman and manage to kill 3 swordsmen in a row. It's just not worth the frustration.

 

[Daaraa]

Being the Germans I switched to this strategy late in the middle ages after I destroyed both the English and the French. I traded gold per turn for tech I had the AI wanted, traded with all civs and only ones I didn't want them to have (ie I didn't trade gunpowder) I am bringing in over $500 per turn from other civs and they love me.
I switched because I wanted the continent to myself so I took a militarialistic approach until I got rid of the French and owned my continent.
Then went to the above mentioned strategy as it is similar to the "Pope" strategy mentioned in other threads.
Now I am about to invade Rome and have signed mutual protection with all of Romes allies so they will leave me alone. I know that will be enough because everyone likes me.
I guess I posted this because I agree with both Tetley and Wislem.

 

[Cyrik]

used that one too in a regent game. won with a spaceship in 1812 or something like this.
the only thing is they have FAR to much money they can give me on regant(sp?). i got 300/turn from some civs! and still they had some research going. ok it was good that they reseached cous i just bought the techs and sold it to the others*G*
2 things i noticed:
1. they can always give you more than they say. when you tell them you have fusion and they tell you: 120/turn they surely can pay 145/turn.
2. they get FAR more money out of their normal economy then i do, although i sold them some stupid city´s that have coruption for them and i take them back by culture every 10 years or something like this.
i don´t think it´s a bug that they pay so much. i´d pay 1000gold too for a tach i would need 15 turns to compleat when i can get 300/turn by switching to taxes. what might be a bug, is that they do this as often as you give them a tech, but think about it like this: at the begining you behave like the comp does in the end...

btw someone tested this strat on deity wher the comp gets far more production and research?

 

[stravaig]

This worked pretty well for me, too, although I haven't exploited it quite to the extremes.
What also works for me is monopolizing luxury resources, and selling those for $/turn. It's /great/ to have the AI subsidize all your maintenance!

- Stravaig

 

[DavesWorld]

Methinks far too many people have played some messed up games, when things have gotten to the point where any good situations that occur in a game are automatically assumed to be resulting from exploitive actions.

The "Science Broker" strategy, for lack of a more concise term, has rapidly become my current defacto method of CivIII play. I dislike the time required to play a game in total Conquest mode (as I read about how many players do, always at war making and marching military units to the attack); I'm a builder. I like to develop my cities with all available and necessary improvements, just because. Further, I like to use these powerful cities for my ends.

The best way I've found to do this so far is to get ahead in the science game. Step one is to establish connections; if you have to purchase contact with other civilizations through those you do know, do it. Whatever price they ask, you'll make it back.

Step two is to have a properly tuned Civilization. You need enough population points working the board, with the multiplier buildings in place (i.e., market/bank, lib/uni/rlab, factory/mfcplant/power plant, etc...) to create advantage for yourself. The tricks for how to do that ought to be obvious; make sure the squares being worked are worth being worked (mine or irrigate them, road everything). You should *wince* when you see an undeveloped square working. It should make you feel physical pain, and you should move to recify it as rapidly as possible. Better, plan your growth so you have new squares ready to receive workers around your cities as they grow. *Before* they grow.

Step three is simple. Tune for science. That's not hard, you just need a good amount of commerce coming in, and you need to have science buildings in place. Get ahead in the research race. You need to be one or two steps ahead of the other civs at least, more if you can manage it.

Then you just extract from them payment. This is really the simplest part of the entire "SB" strategy. The AI will discover these advances *anyway*. In my games, even civs that are being hammered extremely hard usually manage to develop their cities with the improvements. They can keep up more often than not. Since they're going to get these advances anyway, they should get them from you.

You can't sell each and every advance to them, if for no other reason than you can't squeeze blood from a rock. They only have so much to give. But the idea is to take as big a precentage of their civilization's efforts as they'll give you.

The best thing to receive is any Luxuries you don't already possess. Second is cash, preferably in per-turn form. You want them pledged to pay you a sum every turn. If necessary take side technologies (things you haven't researched, or that are in one of the other forks in the tech tree the civ has picked up while you've been racing down others).

By taking per-turn payments from the other civs, you are easing the burden on your own finances. Doing this lets you shift a higher percentage of your own civ's efforts into research. The idea when you start SBing is to get all research efforts to 4/turn without having to bleed your treasury dry to do it. Conversely, the other civs have a lower percentage of their efforts that can safely go to research because they're diverting money from their income to your coffers. They either research slower or have less cash.

When you Broker Science, you do it with all civs in the world at the same time. Extract whatever they can pay you; some will pay more than others. I have never gotten an identical deal, ever. So just take the best deal you can get from them. As you roll forward, four turns later, eight turns later, you'll often find you can't get a deal out of some of the AIs; you've bled them to their limit. They will only give you so much, and refuse to go past this invisible line. But until that point, you can stack up multiple per-turn payment agreements, even from the same civ.

You learn to space out your Brokering; what point in Brokering out, for the fourth advance in a row, your research if they're going to pay you a fraction of what they would otherwise. Basically they get every two or three advances out of five from you, paying for them. The others they research on their own while turns click off and payments go to you. After twenty turns, the per-turn payment deal expires and those funds are 'freed up' and available for you to tap into again. Make another round of deals to soak that money back to yourself again.

My best example of a game conducted in this fashion resulted in a Diplomatic Victory in the late 1700s. I got 'the hammer down' as I term it around 500BC~, seriously developing my cities. I started SBing with the Medieval age technologies. Throughout most of the game after I put the hammer down I had anywhere from as little as 1/5th to as much as 1/3rd of the income I was taking from my cities as tribute from the AI civs (was playing against five on a Large normal world that ended up having two major continents and perhaps three cities worth of island space). Through the second half of Medieval and most of Industrial I had enough tribute income I could (and did for part of that period, as was necessary to maintain 4 turn advances) have my sliders at 0% cash and yet still was turning positive cash flow into my treasury. I kept a close eye on my score, as the French were the other world power and were nipping at my heels tightly score-wise until we got to the very end of Modern, and by the time I hit Modern I had Luxury at 30% merely for score reasons (to elevate the number of happy and content people in my civilization). In Modern I pushed it up to 40% Luxury and had Science at 40-50% most of Modern. I must admit that finally, finally in Modern, I ended up with about a third of the research taking five or six turns rather than four. I was unwilling to drop my luxury rates to squeeze the extra turn or two out. As a result towards the very end France managed to catch up with me technologically ... almost. I ended up researching FutureTech1 when they had three or four Modern advances left to get (Stealth, the last two in the space fork, and I think one other).

The constant trading rendered all civs happy with me, half were Gracious to me most of the game, and the others were only less than Polite *right* after we'd made contact (before I started working on them). I was beaten to the UN by one effin' turn (I mean, crap, it was only a 9 turn Wonder when I started building it a few turns after I received word the AI was working on it, but I was still beaten to it), but there were three rounds of voting over the next 120~ years of game time. The first I received three votes, France two, and one abstention. The second round of voting split it 2-2-2 between myself, France, and Greece. The third round I won 4-2, which ended the game. My score was something in the mid 2700s, on Regent level as Egypt (Religious/Industrial).

It is not an exploit, it's smart play. I never rebuffed the AI on any of its requests (give us maps! is the common one, and later I had a few demands to enter in MPAs). I consider requests like that shakedowns, and happily comply; any AI that asked for things I didn't want to give received a small cash payment on the spot as a thanks-but-no-thanks. I had one war ever, and it was so ludicurious that I'm still curious why the AI did it.

They were happy with me, even agreed to take their column of 10~ knights and Archers out of my lands the turn before they changed their minds and attacked me. I'd watched the column move from their borders and cross mine, aiming right at one of my Northern border cities. It was like watching a bus slide down an icy hill, amusing. I was already building some Tanks at that point, when they attacked, and they were one of the other two civs sharing my huge huge huge continent (the game had two major continents, each with three civs on them; there was a good amount of room for us all to have competitive populations); my response was to immediately secure a Military Alliance and Trade Embargo against them with every other civilization, then I simply rolled their conscript MechInfantry and Infantry with my Tanks. I bombarded the first two cities I took with a 10+ stack of Artillery, then simply just started rolling them directly with the tanks without pausing to soften them or take population points off the cities to lower the defensive bonuses. The Babylonians went down *fast* with five civs pounding on them; I attacked from the South and took the entire Eastern half of their civ purely for score reasons (so I could have more pop points under my flag on the board); the other civilizations razed or captured the Western half cities. It was the silliest move I've seen the AI make in CivIII yet, but was rather funny to watch the world grind them in less than ten turns for their shocking aggression. Literally, ten turns after they surprise attacked me (didn't even wound the units defending my city they attacked), they perished at the hands of the 'Freedom Coalition' I organised against them.

My plans are to keep SBing in future games; its loads of fun to work.

 

[Machi]

ok here's the dilemma, when you are following this strategy called science broker or the pope startegy you are totally screwed if a civ just decides to attack you and wage war.

the careful equilibrium that you have created is thrown into total chaos as democracy causes war weariness and rioting.

so here's my question how did you end up protecting yourself from idiots like the Aztecs or the Zulus who don't care that you are bigger and have mech infantry?

 

[DavesWorld]

As I indicated in my post, I made sure to have good relations with the other Civs in the game. I gave regular (small) cash gifts, a few times gave something more (a technology a couple of times during a round of sales to one of the civs that flat didn't have anything to pay me with). No one was *ever* less than Polite with me, half were Gracious most of the game. Even the war was odd, the Babylonians were polite with me before they attacked.

As for the rest of if, you seem to assume using diplomacy to ensure peace means I had no military. For shame. What do you think I used to roll the Babylonians with, promises?

Finally, even if the game had been marred by deal breaking, my diplomatic positions were strong enough I don't think I would have had much trouble doing to anyone else what I did to the Babylonians after their treachery; secure Military Alliances and Trade Embargos against any Threat civs.

Rest assured, I had enough to secure a MA and TE against anyone I wanted; my treasury was in five digits most of the game, and I had lots of extra resources and luxuries. When we hit Oil and Rubber I had two and three extra of each.

As for War Weariness, I went from Despotic to Republic to Democratic, where I remained for the game. I had 40% luxuries, Universal Sufferage, and Police Stations in every city. I only missed a few Wonders along the way, and aside from Shake.Theater (which I don't like anyway in CivIII), the only Happiness wonder I missed was JSBach's.

Remember, wars of aggession started *against* you have less WarWeariness than wars you start; you can't easily wage Conquest as a Democracy unless you perfect the short war concept (fighting for maybe 8-10 turns before declaring peace and getting your troops back on your soil to let WW ease before redeclaring I guess). But Democracy is pretty good, if you're set to deal with WW, for waging *defensive* wars.

And the war was quite short. And happened on my continent; no navy or transport was required. My tanks came off the production lines and railroaded instantly to the front to help roll up Bab cities.

 

[ChickinSht]

Sweet Strat. I used that with the Iriquios, and, playing on a huge map, I had to research a whole 2 technologies to get to medieval. From there I sold all the low level crap for money, cranked up the science rate...by the time my neighbors advanced to the next age, I was already building Shakespeares theatre. Unfortunatley, there was another continent not under my diplomatic/economic domination, and another 5 civs to deal with. So, I searched, and found one civ, bought all the contacts to them, sold my world map. Then, I did that for every other civ, selling contacts to the highest bidders. By the end of that turn, I was making over 100/turn from other civs, my treasury increased by over a 1000, I had two new techs, had maps of every civ and advanced into the industrial age. Soon afterwards, I just allied with the two most powerful civs (next to me), and completely neglected my military, which meant even greater income under a democracy. Pretty much, I don't see military as being a problem. Keeping minimal defense, and maybe building heavier defended borders near warlike peoples, while placating your more powerful allies and playing Mr. Nice Guy, you can almost completely neglect your military.

 

[Buecephalus]

I haven't had time to try SBing from the get-go, so I tried it in a game I'm playing currently. Playing the Iriquois, I'm in the early 20th century, way ahead on tech although I have had to drop research rate to 50% to prevent going into debt. So I shopped several dead-end techs around for one turn. At the end I was collecting 275 gold per turn from other civs. I cranked up my research to 80% and everything's ticking along nicely! Thanks SB!

 

[Baron23]

OKay, firstly the strategy is not an exploit but that's obvious.
Secondly, it works in higher difficulty levels however maintaining balance is a bigger issue. I know many of you read that long post and thought, damn this guy was only playing with 3-4 other civs on a huge on easy but I was playing with 10 on a higher difficulty and it works even better.
The trick to keeping balance is whenever one of those bastard civs like the zulus or whatever are itching for a fight by demanding something crazy - just make sure to say no but... give them 5 bucks. Its crazy but they are thankful for the fiver. I have never upped it - just use Abe as tribute and let them say thanks.
If they are outright demanding - well you can't counter and you will proabably end up in war cause I don't know how many of you will just give up gunpowder to stay out of war. And if you teach these civs that you are and can be their ***** - they keep making demands.
So, utilize the AI's hunger for techs and remembe that if you don't sell it to all of the civs at once, the civ you do sell it to will sell it to everyone else. This doesn't necesserily cripple them in the highger levels though. They make so much money in the higher levels but it defintely slows them down. I don't know - if others respect this method but I hate military wins and find them boring. I like utilizing culture and economics to make my stake and this is the way for those who agree. That's just my opinion, I could be wrong.

 

[Wislem]

But IS it so obviously not an exploit? Realize that an exploit isn't a cheat or a bug necessarily, it's simply something that gives an unfair advantage to a player. Something that unbalances the game. A couple of points that argue that fact:

1) The computer obviously sacrifices too much for tech that you offer. Would YOU ever sacrifice that much money per turn for a single technology? This is obviously a weakness in the AI, playing into it is exploiting
2) This strategy seems to never fail. If you can get it started and maintain it properly, it's just a matter of time before you win.

Now don't get me wrong, I'm not saying this is cheating and I'm not saying this is bad. I'm just pointing out that it is a deficiency in the AI that allows this to be more powerful than it should be. I agree that trading tech advances is a key to the game and should be allowed, I just wonder whether this stuff makes the game too easy.

And for those of you who get offended by the merest suggestion of the word exploit, get a life, the mere fact that you feel offended means that you feel guilty about it. Let it go, it's just a game, even if they "fix" it (which I don't think is truly necessarily) you can just not patch it and you'll be fine. We play the game to have fun, if we use an exploit and still have fun it then doesn't matter. It's just that it may be unbalancing and if Firaxis wants you want to be challenged to your utmost maybe this needs to be tweaked.

 

[shennessy]

On your exploit number 1 ... yes it does seem that the computer does allocate to much money to techs. Maybe the patch will adjust some numbers?

But one point number 2
The balancing act that is created so you dont get into a war is pretty tricky and sometimes just does not work. If your are in the early game and the AI goes to war with you, you can kiss that easy win goodbye. You might not have enough money/contacts/techs to get other civs to help you out.

Is it an exploit, maybe but I havnt seen where the computer is so crippled that they cant wage an effective war on you in the early game. Maybe this changes latter in the game after you have a thousand years of manipulation over them. (Sounds like an X files episode).

 

[DavesWorld]

As I indicated in my post I'm playing on the even difficulty level (no handicap, no bonus, for me or AI). NOT easy, thanks very much. My next game will probably be one up, minor advantage for AI. In CivII I normally played on King (on under Emperor).

 

[DavesWorld]

Also, as to the post claiming this strategy "always works", I would disagree to a point. It is predicated on your ability to A) Get Ahead Scientifically, and B) have makeable deals. If you can't get A), B) doesn't matter. If you can't get B), A) didn't matter much either.

I don't think the AI necessarily overvalues the advances either. What happens when I sell science to another civ is they get to skip ahead and not spend turns researching. I agree the prices paid are a bit overpriced *IF* you calculate their value based on 4 turn/research times. If you're not getting research every four turns, that research should be more valuable to you. Remember, we now have a all-but-hard limit on Research time in CivIII; it is exceedingly difficult to produce sub 4-turn research times; in CivII one could create situations where every turn saw you getting a new advance. That made it easier to simply keep everything for yourself and plow massively ahead; with the mininum advance cap you have a harder time pulling clearly away from the AI scientifically.

Finally, I would argue the *best* reason for being ahead scientifically is so you can apply the improvements and build the wonders before the computer has the chance to; particularly with wonders. I really enjoy completing Universal Sufferage before the AI even researches Industrialization. I enjoy that a lot.