Tech Slaves of Marco Polo

[Stegyre]

Well, I have to say that I'm very impressed:
MPE has always been one of my favorite wonders, and an essential early wonder. I've also been aware of the "slaving" trick - "gifting" a tech that the AI is working on to get it to switch to something you don't already have - but I had used it only in very small doses. This weekend, I was playing a D+2 game, where I tried to use it at every opportunity, at least during the b.c. era (i.e., before I got lazy and couldn't be bothered):

WOW!

We RIPPED through the techs. I set my own research rate at 10%, 'cause I just didn't need to "discover" techs, myself. As long as I was the middleman, everyone was happy to trade with me. Not only that, but all of my "gifts" made me a favorite with everyone (except the Carts - but they apparently had a bad set up, were getting beat up by everybody, [punch] and had set their own research rate to "0").

It helps me appreciate how some of those early landing dates are achieved.

 

[Old n Slow]

Did you by any chance keep some sort of record? (comment #2 -- maybe we should see if this qualifies as experience for Apprenticing?)

 

[Stegyre]

Yes and no. I generally keep a fairly detailed log - at least of my early game - but then I also "write over" it each time I play. In an effort to improve my results (and to try a slightly different small island), I already started another game, and I decided not to keep track of how much I was trading.

Of course, in the interest of science, I am always ready and willing to start again . . .
(As my wife would rightly accuse me: any excuse . . . )

 

[ElephantU]

Welcome to the "Marco Polo Manipulation" Club...

A couple issues to beware of. When you lower your Science rate the AI tends to follow the human's lead. If you want to rely on the AI research, keep the rate at least 50% or they will be devoting too little arrows to research. Also, don't give them too much. Key things to give them are Republic, Writing, CerBur and Myst, and Currency. This will tend to pacify them, allow them to build Temples, Libraries and Marketplaces, and use Lux to grow their cities larger. You want to avoid giving them a lot of techs, because their beaker cost goes up just like yours with more techs. Often I try to switch their research with a gift, but they ask for something different several times in a row. I try not to give them more than 2 other things, unless I am deliberately trying to slow their research on something I want first (like Philosophy...). Also pay close attention to whether the AI civ is Civilized or Militaristic: there is a different prioritization of desired trade techs and research choices (it's in RULES.TXT, but someone published a sorted list a while back).

Now you know why Marco Polo is the first Wonder of the Early Landers...

 

[Stegyre]

I admit those are good points to keep in mind, but I wonder if some experimentation is in order to verify them. As I mentioned in my own game, I was sliding by at 10%. Despite knowing that the AI would likewise reduce its own research rate, the techs still came much faster than my usual game. It would be worthwhile to see if that improves still more when I keep my research high.

My own thought is that I should reconsider my tech path to devote my own research to a single pathway, and using my "slaves" to fill in the other valuable techs on sidepaths, I just haven't decided what that path will be. (I've given a lot of thought to my tech paths and have a nice list, I just haven't sat down to think this through. Usually, by the time MPE is built, I have an early Republic, Trade and Philosophy, and I'm working my way toward Astronomy. I always start with some techs - I virtually insist upon it.)

As for limiting gifts, the essential is that I must gift them the tech they are currently working on, unless the only way to get there is by giving them one of two techs that I don't want them to have (such as Astronomy, when I am still building my observatory). Sometimes, that does mean overloading a civ with techs. ("Tricky b*****rds! I would have thought you wanted CB a little more than all the other stuff you've milked from me, considering that its what your own scientists are researching!") Overall, I don't know if its more costly to lose all the AI beakers spent in the short-term to research an already-known tech, or in the long term to increase research costs with a lot of extra techs. At least the extra techs are not wasted: the improve attitude, and I seem to cruise through at least the bc and early ad era with virtually everyone "enthusiastic" about our relationship, and multiple alliances.

I have the AI preferences. (Basically, I prepared a multi-page spreadsheet of Civ II info, and I just cut and pasted from the rules.txt.) I'm familiar with the concept of their preferences, it's just that so far my pathological micromanaging has not yet descended to that level.

 

[ElephantU]

The "Pathological MicroManagers Club" is preparing the secret initiation rites...

It is possible to calculate a rough estimate of an AI civ's current research cost. Add one to the current number of techs they have and multiply by 12 if they have less than 15, 25 if more (this roughly balances out the irregularity in the tech costs between #10 and #21). The big jump in cost is with #20, but I was just shooting for a "rule of thumb". Anyway, from Marco Polo and the Trade commodity screens you should have an idea of how many cities they have and what size they currently are. Add together all their citizens and figure 1 beaker per citizen. If they are past 15 or so techs and have Writing I usually up that to 1.5 for the occasional Libraries they sometimes build. Anyway, that should give you an estimate of their research rate, which you can adjust if you log the tech discovery announcements you get between turns (when it says "Greeks DISCOVERED xxx", not "Greeks ACQUIRED xxx").

There is another technique worth knowing with Marco Polo: "tech bombing". When they have shifted to a tech you really dont want them to finish soon (like Philos or Astron if you are working on it, or Communism or NucFission later on), dump a PILE of techs on them, even if it might give them a short-term military boost. Driving their research rate way up will give you extra time to "take care of them"...

When tech gifting, the AI often desires the techs that you have most recently discovered, rather than the one that is most urgent to them - like the one they are working on. In general, they seem to follow the AI Tech Preferences list for their own Civ/Mil attitude. I dont check it every time, but I do keep the top ones in mind (Gunpowder!).

 

[Archer 007]

Very interesting.