It's never occurred to me that I should ever leave any city ungarrisoned, which would result if one had only 9 military on 10 towns; wouldn't they see my weakness and jump on me rapidly?
A human probably would, if you were playing an online MP game, but the AI won't: the AICivs (and your own military advisor, who uses the same algorithm) don't rate your military strength by 'how many garrisoned troops per city?', they basically total up
all your current units' current A- and D-values (also including minor modifications for fortification bonuses, total HPs, etc.), and compare that number with their own units' totals. If TheirTotal >> YourTotal, i.e. they rate you as 'weak', they are more likely to DoW you (especially e.g. if you provoke them by trying to evict them). For the purposes of the mil-strength calculation, A-values are also counted as being worth ~1.5x their actual value, which is why running a military composed primarily of 'defensive' units (i.e. low-A, high-D) is a 'good' way to invite an attack.
And under Republic you have no MP, and high maintenance costs for excess units. So as an early Republic, so long as you don't sign RoPs with your immediate neighbours, you can quite safely use the 'hard shell/ soft centre' troop-distribution described by Vorlon_Mi, to maximise your ability to repel invaders while minimising your military support costs -- at least until the AI gets M=3 units (e.g. Chinese Riders, or Cavs), at which point you may also need defenders in your 2nd-line cities (especially if you've used 'tight' CxxC city-placement).
I was concerned about the floodplain cities that they'd have megatons of food, but not enough shields to build settlers fast enough to take the pressure off.
Under Republic, your increased income can (if necessary) buy the needed shields when the city gets close to rioting due to overpopulation; although spending gold like this this is not advisable if you can avoid it, Floodplain-cities
are a special case.
Micromanagement will almost always be a better option though: if your current LUX%+Luxes can only sustain a certain pop-level in your Settler-building Floodplain-town(s) (say, Pop6, because at Pop7+ it will riot) then you should aim to reach that level ASAP, then shift tile usage to increase shield-output (and decrease food-harvest) so as to match the remaining Settler build-time to the growth-time. On the interturn, commerce/happiness is calculated first, then food/growth, then shields/production, so the city will grow and then immediately shrink again, without rioting. As soon as the Pop drops, send enough citizens back to the high-food (low-shield) tiles to grow the town back to max. sustainable size in the minimum time (ideally without food-wastage). If you can't match the build- and growth times exactly, it may be better to have the build complete just
before the growth (especially at Pop6-7), so that the food box starts the post-build turn at (nearly) full; that way, you also get some shields into the box over the next growth-IBT.
If you already have too many people to keep happy, and need to stave off a riot, or prevent further growth, just make the unhappy citizen(s) into a Scientist for the first part of the interturn before the Settler gets completed. That way you get some beakers out of them, and the city drops to a more manageable size. If the unhappiness-problem is more widespread, up the LUX%-rate a notch.
So you are saying that I'm coming out okay enough to move up a level in difficulty in spite of my systemic missteps?
If you can 'beat' Regent already (1st in most/all demographics says you can), then yes, up your game. A Monarch-level AICiv's not significantly different to a Regent-level AICiv, so you should find Emp winnable (if not every game!
). An Emp-level AICiv's
build/research advantage is now noticeable (20% discount -- so e.g. an Archer costs them 16s vs. your 20s), but still not yet significant in the early game, due to the AI's inability to manage wastage/ overruns on smaller builds (e.g. it will happily build that 16s-Archer in 3T at 7SPT, wasting 5s and negating its discount). Emp-level AICivs also do not get large numbers of starting units (1 offensive, 4 defensive, 2 Workers) and still only get 1 Settler (they get 2 Settlers at DG, which makes an increasingly large difference to the early game at smaller map-sizes, i.e. where a single town takes up a greater proportion of the total settle-able space).
In that scenario, I'd guess that emphasis on research will be a lot more important because of the increase in costs
If you habitually play on Huge maps, that will be a far bigger factor for your turns-to-tech than the increased difficulty/ AI's research discount (especially under an early Republic, before you have many cities down). In the early game, commerce/research is limited by the fact that (regardless of mapsize), it takes a more-or-less 'fixed' amount of time to build X Settlers. On a Huge map, you need a bigger value of X to cover 'your' territory, and hence it takes that much longer to get your cities down and developed to the point where your total beaker-output per turn becomes comparable to the increased tech costs on Huge maps.
You will almost certainly find that you need to do (a lot) more tech-trading at Emp, especially on Huge maps: At Regent, even an early Human Republic can still research quicker than the AIs, so you find yourself going through the Middle-Age tree well ahead of them. At Emp, they get teched-up 20% faster -- but still not so fast that the Republic-slingshot isn't viable any more, so if you get it, you can usually pick up almost everything else in trade for the CoL/ Philo/ Rep triumvirate.
it will be more important early on to raise the luxury rate
Yup. But ideally, under Republic, use your free-unit limit to cover your military (50 cities at Pop7+ = 150 free units!) and foreign income rather than TAX% to maintain your buildings (which should be limited to game-aims, at least at first). Aim also to obtain Luxes ASAP -- 3-4 Luxes should be enough to keep an Emp-level Pop7 city happy under Republic at LUX%=10-20%, with the rest going into SCI%. Be careful of making per-turn trades though, e.g. only sell excess Luxes to your neighbours for GPT if you can be sure of maintaining the trade route(s) for the requisite 20T.
So if I am doing that, building settlers as fast as I can until I've platted out all my land, and end up with a bunch of ungarrisoned small towns without road connections for a while, that's okay
Yes. While those outlying towns are still small/ undeveloped/ (nearly) ungarrisoned, losing them to AI-aggression is much less wasteful of your output than if they fall after you've built/rushed (expensive) units/ improvements in them. And if the loss happens before they've reached Pop2 and/or popped their border, they'll get auto-razed, so the attacking AICiv will gain nothing (while Barbs are annoying, they don't capture cities). Culture-flips
are a pain though, because the AICiv then gets to keep the town.
As for the garrison and roads, in my games, a Warrior and/or a Worker are frequently the first 2 projects for a new town: the Worker's first job will then be to road the shortest-but most-immediately-useful (flatland-)route back towards the nearest older town which is already connected to my trade-network (hence supplying Resources and Luxes, and possibly also reducing corruption). The Warrior (if built) will then be sent to nearest Rax-town for upgrading, while a better unit(s) moves to garrison the new town.