My other computer is almost working again (well, it's working enough so I can compose posts in something other than a browser and save them), so it's time for me to write out some of the thoughts I've had over the past couple of days.
Grangerm, I think you made some good points, and in fact I'd had some of those same thoughts even before you wrote your post.
The limiting factor in the victory this game is going to be tech. We should save an engineer (or two!) to rush the UN, and even without, if we assume Moscow has a base production of about 19, a forge, and Bureaucracy, it only takes about 45 turns to finish it, or 40 with Organized Religion. That sounds like a lot, but it will probably take over three hundred turns to get the necessary techs to build it; meanwhile, that figure for the UN doesn't count engineers or switching to Universal Suffrage to get production from all Moscow's cottages and cash-rush. I don't think we're going to have any trouble at all getting to the necessary population for backdoor domination: the AIs are weak at growing their populations on archipelagos anyways, and we're going to be running a specialist economy and should emphasize growth by sparing the whip more than usual, etc. (We may also want to put some emphasis on diplomacy, to try to get some good friends who will vote for us and possibly shave a few turns off the victory by sparing us the need to go quite so conquest-happy.) Provided we can win the vote without much difficulty, the slowest step in this process will the long road to Mass Media so we can construct the UN.
Thus, we must get to Mass Media as fast as possible, and we need to start thinking about how to do that soon after the Pyramids are finished. I think one of our goals must be to avoid slowdowns in our tech rate as much as possible: we'll be running a specialist economy, so a lot of our science will be coming from specialists, but we should still aim to keep our commerce-based science rate as high as possible. With no city-razing, in particular, this means that we should be careful about fighting wars that will tank our science rate. I still think that grabbing Mao's and Alex's capitals might be worth considering in the opening, but even that may be ambitious: we'll have to estimate how soon they'll pay back the opportunity cost of the military units to capture them (which is hard to quantify, but I think we'll want to look at it nonetheless) and the maintenance (which will be easier to calculate). Of course, we have to balance this against the need to eat civs for the backdoor domination victory, but I think that as long we conquer at a moderate pace, we should be able to manage.
(I don't feel like I can make any more general statements about how to conquer to increase one's tech rate--anyone have anything else to add? Good rules of thumb for deciding when we should declare war, etc.? I know that conquest in general can help your tech rate if done right, but I also know it can slow it down.)
A different issue is the tech path we use to hit Mass Media. I'd like to point out that the number of techs you need to beeline there is . . . surprisingly small, and avoids some techs that practically everyone gets normally. Let me diagram:
The end of the line:
This is all linear.
Astronomy + Printing Press -> Scientific Method -> Physics -> Electricity -> Radio -> Mass Media
Astronomy prerequisites:
The Astronomy line is linear given the techs we already have. (I skipped some techs we already have here.)
Iron Working + Sailing -> Compass + Machinery -> Optics + Calendar -> Astronomy
Pottery + Bronze Working -> Metal Casting -> Machinery
Writing -> Mathematics -> Calendar
Printing Press prerequisites:
Unlike Astronomy, there are lots of different ways to get through this line because Paper has lots of alternative prerequisites in its tree.
Paper + Machinery -> Printing Press
Meditation or Polytheism -> Priesthood -> Code of Laws -> Civil Service -> Paper
Mathematics -> Currency -> Code of Laws -> Civil Service -> Paper
The first of these two paths is cheaper, the second has more overlap with the other techs in a straight beeline and provides better benefits.
Meditation or Polytheism -> Priesthood -> Monarchy + Writing -> Feudalism -> Civil Service -> Paper
I don't like this line at all: it's expensive and gives us little or nothing.
Polytheism + Masonry -> Monotheism + Writing -> Theology -> Paper
The Theology path costs much less than the Civil Service path, but it provides almost nothing we want.
Other Techs We Want:
Those are our beelines. What other techs might we want? I see compelling arguments for the cossacks techs, Liberalism, and Banking. If we get them soon enough, cossacks can more or less run over enemy civilizations all by themselves. Liberalism requires two major extra techs in addition to itself, Education and Philosophy, but we may be able to recoup some of those beakers by using it for an expensive tech on our Mass Media line; Education is also an incredibly powerful tech for our civilization, because we're Philosophical (for cheap universities) and we have stone (for a cheap Oxford). Banking opens Mercantilism, which synergizes with Representation in sprawling empires. Let's look at these lines in more detail.
Liberalism prerequisites:
Education is just a straight shot from Paper, which I've already diagrammed above. Philosophy is a little more complicated.
Writing -> Alphabet -> Meditation + Drama -> Philosophy
Writing -> Mathematics -> Currency -> Meditation + Code of Laws -> Philosophy
Meditation -> Priesthood -> Code of Laws -> Philosophy
Cossack prerequisites:
Cossacks require three separate techs, Military Tradition, Gunpowder, and Horseback Riding. The third is easy, but Military Tradition and Gunpowder require more investment and have more potential paths.
Civil Service + Philosophy -> Nationalism + Music -> Military Tradition
I've already covered all the ways to get to Civil Service and Philosophy. I think Divine Right is a dead waste we should avoid; let the AIs waste beakers on it.
Literature or Drama + Mathematics -> Music
Education -> Gunpowder
Monarchy + Writing -> Feudalism + Machinery -> Guilds -> Gunpowder
Education is the fastest way to Gunpowder, but we don't need speed here.
Banking prerequisities:
Monarchy + Writing -> Feudalism + Machinery -> Guilds + Currency -> Banking
Tech Path Proposals:
Now that I've analyzed what we need and how to get there, it's time to put it together into some concrete suggestions for our tech choices. Before we develop anything else, I think we want these techs, not necessarily in this order: Iron Working, Pottery, Sailing, Writing. After that, I think to get the Great Library on time (i.e., right after our engineer appears), we want to research Alphabet, Polytheism, and Literature. Then things get interesting, because I can see three plausible targets for our beeline after Literature.
Education through Theology:
The first is an unorthodox gambit I've never tried: Education beeline through Theology! The cute thing about this path is that we have all the foundation techs for it right after finishing Literature. Monotheism should be cheap to research because it will be spreading among the AIs at that point, or we can trade for it. Then, we research two cheap techs, Theology and Paper, and we can start Education. We'll have to look at the exact timing of the great scientists, but I expect by then we'll have birthed two or three: if the first one appears much before Education, I think we want to build an academy in Moscow or St. Pete, then use the second and the third to speed Education, otherwise we save them all for Education.
The point of this gambit is to get super-early universities, probably in our four original cities along with Athens and Beijing, and then follow it up with Oxford. Depending on the details, I think we should be able to get Education not long after 500 AD, and finish Oxford before 1000 AD! Besides the obvious science bonuses arriving much earlier than normal, don't forget that Oxford would allow us to run more scientists for more research and more GS points. In ancillary benefits, we might found Christianity and might be able to get the Sistine Chapel. (We don't have marble, but the Sistine doesn't seem to be a high priority wonder for the AIs, and we could great-engineer rush it. Since we're not going to be founding many cities ourselves, the Sistine isn't as strong as it is in some games, but it would still be handy for popping borders in conquered cities.) We might also use Paper to help with early circumnavigation, though we'd have to be careful. The downside of this path is that until we've finished Education, we won't have much in the way of other tech and thus economic improvements to build. We can consider trading Theology, but I'm not sure we should trade Paper: we may want to discourage the AIs from researching the Liberalism prerequisites so we'll have more time to sandbag the slingshot. Even trading Theology will probably cut the AIs' time to Paper-Education and thus increase the possibility we'll have to take an earlier and cheaper tech with Liberalism.
(More on the great people and wonders in a later section.)
Philosophy lightbulb:
Another beeline would be to use our first great scientist to lightbulb Philosophy. We'd probably want to take the Drama route to Philosophy, since Code of Laws, while probably more useful, is more expensive; and we'd have to research or trade for Mathematics. The point of this would be to found Taoism and then switch to Pacifism to maximize our GP points for faster lightbulbing of subsequent techs. Running Pacifism might sound like a big disadvantage, but I don't think we'll have the economy to support extensive conquest in the midgame, and provided we're careful with our diplomacy and alert with our naval forces, we can defend ourselves using ships at the choke points instead of land forces. The big ancillary benefits here are having early Philosophy for Liberalism and the possibility of Angkor Wat.
While I don't think we want a lot of prophets, one or two may be handy for the Dai Miao or a holy city we capture; the major point of Angkor Wat would be to run two-hammer priests in food-rich cities without much production that will never produce any great people. I can see two places to put Angkor Wat: in our GP factory, where we'd run priests to increase its production to help build essential items (Oxford comes to mind) and GP points until we decided we didn't want to risk any more prophets (keep in mind that, on average, increasing GP points will get us ; it's only the variance that could be a problem), and then hope the small amount of contamination doesn't produce any more; or in another city, where we'd run priests to generate as many prophets as we felt we needed, then stop.
Drama and Philosophy make good trade bait, but again, we may want to be careful about trading Philosophy because of Liberalism. A final ancillary benefit is that if we do develop Literature and Drama, and the AIs are laggard about pursuing the Alphabet line, there's a good chance we can get Music first: we can then burn the free great artist to lightbulb some later tech, recouping some of our beaker investment, and trade Music around. (Notre Dame is all but useless on archipelago maps, and if anything I'd like to encourage the AIs to put hammers into that instead of something dangerous to us. Likewise, Music doesn't open any techs we don't want them to have, since cossacks murder ordinary cavalry: if AIs want to spend the beakers and cash getting cavalry instead of something like grenadiers or rifles, I'm all for it.)
Economic Development with Naval Lightbulbs:
Since it doesn't seem like it's a spoiler for the SGOTM2 people, I'm going to bring in some ideas from CFR's strategy in the last game.
A third approach commits our early great scientists to Compass and Optics, then we research Civil Service, Paper, and use subsequent scientists to lightbulb Education and possibly Astronomy. Compass will be open for the first scientist, but to open Optics for the second, we'll have to have developed Metal Casting and Machinery. Our exact tech path will depend on the timing of the scientists, but I imagine that (after Literature), it will look something like Mathematics, Currency, Calendar, Code of Laws (we probably want to get Priesthood first if we're researching CoL, since it's dirt-cheap), Metal Casting, Machinery, Civil Service, and Paper, not necessarily in that exact order, but in something like it.
The big advantage to this path is that, instead of sinking beakers upfront into techs that don't offer us many immediate benefits, we research techs that improve our economy. Meanwhile, there's not much reason to hold onto our lightbulbed techs until we get to Education, so we can trade them to pick up even more techs. (Compass offers almost no benefit to a monopolist, and the major advantage of holding onto Optics would be to maintain naval supremacy, since we should get circumnavigation long before Optics.)
Combining The Three:
None of these paths are exclusive, and we can mix and match them in various ways. The only one that's vulnerable to AI disruption is the Philosophy lightbulb, because it relies on getting Taoism and, to a lesser extent, Angkor Wat and possibly Music for its benefits. The Theology-Paper-Education line might lose Christianity, but the religion is not essential to the beeline. I'll briefly list some possible variations, starting after we develop Literature.
Mathematics -> Meditation -> Currency or Priesthood -> Code of Laws -> Philosophy (lightbulb) -> Civil Service -> Paper -> Education (lightbulb)
The early path depends on what we can trade for, when. Researching Currency would be risky, since the AI can lightbulb Philosophy too.
Mathematics -> Meditation -> Drama -> Philosophy (lightbulb) -> Theology -> Paper -> Education (lightbulb)
We give up Christianity here for certain, but I don't think we care. This is a long time to spend researching unproductive techs, but this would allow us to develop Liberalism insane early.
Putting It Together: the Liberalism Slingshot
After we have Education and Philosophy, Liberalism opens and we can decide what approach we want to take to the rest of the game. I see two valid ways to approach Liberalism: the first is to use it for Astronomy ASAP, and the second is to try to race up the tech tree so we can use it on more expensive techs. To derive any benefit from the latter choice, we have to make it to Printing Press and Astronomy before the AIs can develop Liberalism so we can use it on Scientific Method, since all the techs earlier in the tree don't cost more than Astronomy itself. The big reason to try for the second is that after Scientific Method, each additional tech is thousands of beakers more than the one before it until we hit Mass Media. The big reason for the first is to open observatories, which not only give us a cheap +25% boost to our science rate, but also raise our scientist cap in each city.
Obviously, if we're not going to wait on Liberalism, it behooves us to get it to ASAP. The best paths to get there would probably be the Philosophy lightbulb via Drama followed by Theology-Paper to Education, or Philosophy lightbulb via Code of Laws followed by Civil Service-Paper to Education.
Banking and Mercantilism:
The AIs algorithms' love Feudalism and the subsequent techs, for some reason. Meanwhile, Monarchy will be useless to us except as a prerequisite, and we shouldn't need longbows or Vassalage, leaving the first useful thing from this line grocers at Guilds. I think we should let the AIs develop this line for us, so that we can either pick up the techs in trade at much-devalued costs, or just develop them ourselves with the discount from so many other civs knowing them.
Cossacks:
As I observed earlier, if we're going for early cossacks, we should get Nationalism with Civil Service and Philosophy. We're not going to have the tech rate to sandbag Liberalism and rush with cossacks, so if we want to maximize cossacks' power, I think we should combine it with a Liberalism slingshot to Astronomy. The tech requirements for this route suggest a tech path with something like a Philosophy lightbulb via Drama leading into Music, then following part of the economic development line to Civil Service while lightbulbing Optics, then lightbulb Education, develop Liberalism, and use it for Astronomy, get Nationalism and Military Tradition, then hit Gunpowder and unleash the cossacks, trading for Guilds if possible to make Gunpowder cheaper.
Great People and Wonders:
Maybe I shouldn't have left this section for last, because it's informed everything that's gone before, but here are my thoughts on how we should acquire and use our great people, and arrange our wonders.
Great People:
I think for this game, where we have a specific objective we want to achieve as fast as possible, we should weight purity of great person pools heavier than in an ordinary game. Knowing we're going to get a specific kind of great person will help with planning. The two dominant kinds of great people we want, I think, are great scientists, because they can lightbulb most of the techs essential for our strategies, and great engineers, for rushing wonders. Here are my suggested uses for the great people we get, with stars beside the ones I consider most important.
Great Prophets:
Build a shrine for a religion if we found/capture one.
Lightbulb Civil Service w/ Theology but w/o Monarchy.
Unfortunately, the great prophet lightbulbs are blocked by Monotheism/Theology/Divine Right/Monarchy, ahead of everything else we would want to use them on (Printing Press, Education, Liberalism).
Great Scientist:
*Build two academies, one in Moscow and one in St. Petersburg.
Lightbulb Physics.
*Lightbulb Education.
Lightbulb Printing Press.
Lightbulb Philosophy.
Lightbulb Optics.
Lightbulb Paper.
Lightbulb Astronomy.
Lightbulb Electricity.
Lightbulb Compass.
I think we want two academies, one in the capital, the other in our GP farm, unless we're doing a lot of lightbulbing. The only lightbulb I see happening no matter what our tech path is Education; the others are all distinct possibilities, though.
Great Engineer:
*Rush the Great Library.
*Rush the UN.
Rush Angkor Wat.
Rush the Taj Mahal.
Rush the Sistine Chapel.
Rush Oxford.
Rush the Forbidden Palace.
Lightbulb Machinery.
The Great Library is obvious. I want to save at least one, and possibly two, engineers for the UN. Rushing Angkor Wat in St. Pete would give us actual production there in a hurry. If we get Nationalism and have an engineer available, I think rushing the Taj Mahal might be worth it, as much for denial purposes as the actual benefits it will provide. Normally, I prefer not to rush national wonders (particularly when I have the resource for them) with engineers, but Oxford is so important in this game it may be an exception; if we need to reduce maintenance costs early because of too much fighting, the Forbidden Palace is also a possibility. Lightbulbing Machinery is something unorthodox we might look into if we want to hurry Optics for some reason.
Great Artist:
Lightbulb Radio.
Lightbulb Mass Media.
Like prophets, anything else we'd want to do with them is blocked by Monarchy/Divine Right.
Wonders:
I think we want to look at the following wonders during the course of the game.
Pyramids (duh!)
Great Lighthouse (I can dream, can't I? If this gets built anywhere near us, I think we ought to consider . . . acquiring it.)
Great Library (duh!)
Hanging Gardens (More engineer points, more health, more population for whipping and specialists, we have the necessary stone--one of my favorite wonders in any game, and well-suited to this one.)
The Colossus (Not a huge priority, particularly if we plan to hit Astronomy with early Liberalism, but it's dirt-cheap with copper.)
Angkor Wat (This wonder raises some complicated questions I'll address in a bit.)
The Sistine Chapel (Also not a huge priority, but it's almost like being Creative for free and combines well with a specialist economy. I wouldn't build if we aren't rushing it with a great engineer, since it's marble-based and not that important.)
The Taj Mahal (Might be worth rushing if we develop Nationalism for cossacks early enough and have an engineer available.)
The UN (duh!)
Forbidden Palace (Timing on this depends a lot. If we take over some other AIs early and then spend a long time avoiding warfare, I'm not sure it will be that useful. If we have to capture cities far away from our capital in the early war, or decide to go adventuring before cossacks, it might be worth looking into.)
Oxford (duh!)
National Epic (duh!)
Heroic Epic (duh!)
Globe (If we can find an appropriate site for it, the Globe is probably best used as a military-whipping center.)
West Point (I don't know if we'll have enough time to build this wonder and justify its cost, but we do have stone for it.)
Wonder Cities:
Here's my proposal for wonder and GP allocation.
I think Moscow should be a pure GE factory as much as we can make it: we should time the Hanging Gardens there to finish after any early wars, then build a forge ASAP and hire an a permanent engineer. This will generate 14 GEP a turn, which is about as good as you can get for a pure great engineer pool before the modern age. The Hagia Sophia offers 2 more GEP, but it requires marble and two techs off our beeline. I don't know how many more engineers that will pop, but we should get at least one more for the UN.
I think the Great Library should go into St. Pete, and then we can follow it with the National Epic and Oxford. Putting Oxford into a city other than the capital is an unorthodox choice, but I think I can support it: if St. Pete runs seven or eight scientists under Representation, it will pump 54 beakers or 60 beakers per turn before buildings or commerce, and 27 or 30 unmodified GSP. Moscow will never be able to run that many scientists, and while it will generate more than 54-60 beakers from commerce at close to 100% science, it's quite possible St. Pete will generate more beakers than Moscow when you average over all the time we're running cash for upgrades, or culture for war weariness, or whatever. Meanwhile, until we build the National Epic, we'll have a pure great scientist pool, ensuring we'll get two or three great scientists first, with no artist contamination.
The biggest issue with making St. Pete a GP farm is that it has no production. With an engineer from a forge, we get 6 hammers (2 from the engineer, 2 from cows, 1 from the gems, 1 from city square). There wasn't really any way around this, though, since there were no hills anywhere near it, and only one forest; but neither the forest nor the river will add more than 1 production per until after the time-frame of our game (Replaceable Parts, which is nowhere on any of the tech paths I see us taking). We have enough food to whip in necessary infrastructure (health buildings, science buildings, and happiness buildings are all it needs); the problem is that it's hard to pop-rush national and world wonders. This is the reason I proposed looking into building Angkor Wat in St. Pete: three 2-hammer priests will double its production, and the extra GPP points will help us out until we've finished Oxford and can run more than two scientist specialists. However, this is far from an obvious play.
I think Novgorod should build the other world wonders whose GPP we don't want, artists and merchants in particular. Hopefully, we'll never get a GP from it.
I think we should put the Heroic Epic in a junk city: something with one or two food resources, some production (hills or resources), and lots of dead tiles. Sparta looks like a good candidate if it has seafood, with all those desert tiles. The idea there is that we can use said city to build/whip military for the entire game at no opportunity cost: it would have no need for other infrastructure aside from a granary and maybe a forge, meaning it can build the lion's share of the military units we'll need until, possibly, we get to cossacks. An alternative possibility is to find a city with a lot of food and nothing else, and combine the Heroic Epic with the Globe for constant military whipping. We could also divide the Epic and the Globe, creating two all-military cities, depending on circumstances.
