Everything I will say here applies only to a game where you intend to double-bulb Astronomy
Only settle cities that will either...
a) Highly increase your overall commerce output. I'm not talking about distant, single desert gold + 1FP spots, but about cities that have at least 30-40BPT potential by 1AD, you really need to keep maintenance as low as possible. Remember that settling a new city increases the "number of cities" cost in every city? Therefore, the single desert gold + 1FP (let's say 12
) spot doesen't pay for itself when you have 6 grown(ish) cities around 1AD. Cities that only provide hammers should be avoided, unless you're going Alpha anyways, then a really strong production spot can be a good settle even on isolation.
or...
b) Pump out GS for your bulbs. Here is the more tricky one. These cities don't always have to be high-food cities. Here there are different cases :
-
1) You have marble and a decent GL + NE spot. Then you only need 1 GP farm (maybe 2 if there's another really nice spot), and you should rather focus on a), as GS production isn't gonna be a problem, but getting Aesthetics + Litt + all the needed techs to free up the Machinery + Optics bulbs requires a lot of commerce.
-
2) You have spots with very high food (2 5+ food tiles or more) Then you should generally settle the high food spots and tech CoL, Caste System will get you the necessary GS easily enough. In this case it's usually better to get 4 GS total (Academy, Optics + 2xAstro), but you still have to tech Alpha and Maths
- 3) You are on a low-food island. Then you can't justify going for CoL, as you won't be able to run more than 2-3 scientists in most cities, even in those that you want to generate GPP with. Thus you have to generate 3 GS (Academy + 2xAstro) and self-tech everything. If you have a lot of high-commerce spots you might even skip Academy, which makes GPP production easier (see
a)). To generate 3 GS by 600AD (this date varies a lot, but 600AD is usually a good Astro date), you have to run 2 scientists for 50 turns as the third GS costs 300GPP. This means that you're forced to have 3 different "GP farms" -otherwise 600AD Astro is off the table - set up by turn 90, i.e.
3 cities with a library at a high enough size to run 2 scientists by 625BC. However these cities don't need to have high food, so they can simply be the best non-commerce spots you find on your island. For example a copper + horses + dry wheat spot is better than a clam + dry rice spot, although the latter has more food. The main reason for this is that the better (lower food) spot will be more useful as soon as it is done pumping out GS. Now for the GS production order, the best you can do is to get the first GS in your best city (among those dedicated to GPP production ofc), the second in your second best city and the last in your crappiest city. I'm saying that because even if your crappiest "GP farm" happens to be your second city, you're usually better off getting the last GS from here and the first two from elsewhere, as you will be wasting less "potential" by putting less specialist pressure on your better spots. Here's an example (in spoiler cause it's taken from the current game) :
You can always catch up on expansion after Astro anyways, no one is gonna settle your island
My second point is that when playing isolation, the most important thing (pre-Astro) isn't to maximize your commerce or GPP, but to
find an equal balance between commerce and GPP
Don't get me wrong, in 99% of the cases you will need to build an Academy somewhere, however,
you don't always have to build an Academy before Astro
For example, in the Boudica Game, I could have stored my GS and gotten 500AD Optics instead of 375AD, but I would also have had 600ish AD Astro, and 800AD Academy, instead of 100BC Academy and likely 800AD Astro. Ideally, you should finish teching Optics / getting the last Astro prereqs at the exact turn where your second Astro-GS gets born. As Astro gives such an enormous boost in isolation games, this kind of timing is very important, and you sometimes have to make sacrifices for it.
I'll take the Boudica game as an example again. I decided to settle cities #3 and #4 as soon as possible, at the expense of capital infrastructure and cottaging (though I made a mistake not building one more worker early on
). But it still was a good choice, because this is a relatively high-commerce island, but it's very low on food and had a slow start, so getting the necessary GS to bulb Astro was the main issue. On the other hand, if you have a high-food-low-commerce island, you should often priorize increasing commerce (build cottages etc.) at the expanse of GS production.
Thus, I think we can all agree on a general guideline that applies to 99% of isolation games :
if you have to work coast or unimproved tiles to generate the necessary commerce to free up the Astro bulbs, OR if you find yourself waiting 10+ turns for the last Astro GS to come, you are probably doing something wrong
Hope this helps