The game so far: As planned, I settled Delhi in place, worker first, Agriculture, then Fishing, exploration starting SE.
Screenshot on t16 (3680 BC):
I initially explored the S a little less than I would've liked in order to escape from a Lion. Going NE from there, I came across Ragnar's borders on t12. Random Personalities, so, no sweat. He is pretty close, just 12 moves between our capitals, for such a sparsely populated map, but not necessarily close enough for a Warrior rush, and Delhi is badly suited for that, lacking any 3-production tile. Between the desert, jungle, bay and Ragnar, it seems that good city sites will be somewhat scarce. Settling in the flood plains right in between our capitals would secure a strong, contested site (Financial trait should make Ragnar extra keen on it) and would probably block Ragnar from any sites to the S of Delhi. Not likely to give me early access to a strategic resource though.
My worker finished on t30 and improved the Wheat. Work Boat next, and Hunting after Fishing, then BW. (Forgot Hunting at first but remembered in time to switch from BW.) After running some numbers, I determined that it was pretty clearly best to grow to size 3 (not to size 2 as I had initially intended) while producing the Work Boat.
While exploring the northern rim of the desert, I spotted a Lion and Wolf. The Wolf ended up attacking me on a Desert hill, dealing 0 damage. The desert almost directly transitions to jungle in the north, and, to keep my Warrior safe, I didn't explore any of that jungle. In the W and SW, I found less to explore than anticipated; the coast is pretty close. I finished exploring along the eastern shore of the bay then.
On t43 (3140 BC), Saladin's Warrior approached my borders from the SE. Based on the timing, I expect that his capital is some distance directly S of Ragnar's.
On t55 (2900 BC), my worker ran out of work with 11 turns to go until BW. Decided to build a (non-River) Plains Hill Mine, which should give the city some more flexibility at some point in the future – and makes the worker available almost as soon as BW finishes.
On t59 (2820 BC), the Work Boat finished, and I started a settler, taking a projected 25 turns by working the three improved tiles (Wheat, Fish, Ivory). Will be able to chop once, bringing the production time down to 21 turns.
On t64 (2720 BC), BW has finished, and, to choose my next tech, it's necessary to plan ahead for some 50-100 turns. Screenshot:
My default plan on a map with a lot of space for Barbarians is to place the second city adjacent to Copper or Horse. One Copper source does exist, but it's on Flat Desert (same yield as a Plains Hill Mine), and all adjacent city sites are poor. The least bad option would seem to be SW of the Copper, stealing an Oasis from Delhi and never growing beyond size 2. The four adjacent Forests could power an Axeman rush against Ragnar, but I think I can manage Ragnar better through forward settling. Also a little risky; he may have easy access to Copper, has already finished BW one turn before me and has the Aggressive trait.
There is still time to find a Horse - 17 turns to AH, settler finishes in 16. However, there are fewer than a dozen nearby tiles where a Horse could exist (non-River Flat Plains or Grassland, not adjacent to another resource), and just half of those in places where I would consider placing my 2nd city. So this probably won't pan out.
Claiming the Stone for an early Great Wall seems like a viable play (or also Pyramids) but would likely cause Ragnar to expand to his SW. Also, Desert Quarry takes a lot of worker turns without providing a decent tile yield, and Masonry is a dead end until Construction.
With all the hostile terrain around me, keeping Barbarians in check with just Warriors and Archers (even just Warriors?) might not be so difficult. Let's see:
Given that jungle is unlikely to spawn Barbarians, the W and SW is pretty much secured, the more so once Delhi's borders expand on t105. To the NW, there's a savanna-like patch that ought to be fog-busted. The NE is just desert and jungle, but a lot of it, not feasible to fog-bust, so something will probably arrive from that direction before long. A couple of units near the capital should be able to deal with that. The flood plains to the E is where I'd place my second city. That city will be exposed from the N and, more so, from the S. The southern region should be fog-busted with 3 units, but that'll take a while to set up; so the flood plains city will need to be defended and may struggle to keep its improvements safe.
With Archery, this sounds quite doable, sth. like 4 Warriors and 3 Archers should suffice until the Barbarians get BW. By then, I'll need either Copper or IW. With all the food at the capital and in the flood plains, I'll want Libraries before IW, so I'll probably place my third city at the Copper-Cow-Jungle Gems spot and expand borders through a (chopped) Monument.
So: Settle near the Gold and then spam Warriors and Archers for some time.
Might I have time for founding a religion? The only near-future benefit would be to quickly expand the borders of my second city and to strengthen it against any encroachment from Ragnar. For this to work, the city would have to be founded before the religion. Meditation takes 17 turns. I can beat that if I whip the settler on the turn that the Forest chop finishes (if I've done the math right). Polytheism would take 21 turns. I don't think the AI generally prefers Meditation much over Polytheism, if at all, so Meditation seems like the better bet. Either way, I see a high chance, at least 50%, of being beaten to it by one the 11 other civs. Arabia, for one, starts with Mysticism.
A border pop isn't particularly needed if I found W of the Gold, but it would enable a greedier site: at the coast, claiming both Gold. That city could co-exist with a Stone-Flood Plains city. Ragnar would struggle against my culture in that spot, and a potential Viking city between the Ivory and Cow looks too weak to happen anytime soon. Could also connect that double-Gold city via Sailing, whereas connecting the single-Gold city would require 5 Road segments.
Time-line: I'm in no rush with the third city, so I intend to train some combat units and one or two workers before that; they're just so cheap on Marathon. Should take at least 50 turns until I have the third city and at least 15 more to expand its borders. So Meditation (17 turns), then Archery (13 turns), then Wheel+Pottery (13+14), then AH (probably just 15~16 by then through diffusion). Pottery because I'd rather not farm every Flood Plains. This should give me AH still in time for the Cow, maybe a few turns late.
The main downside of Meditation is that I could otherwise avoid that tech indefinitely, and making that detour will matter eventually, maybe for getting Writing in time or IW. And there's the risk of losing up to 16 turns of research entirely if I'm beaten to it. I think I'll go with it anyway because it's an unusual play (for me at least), one that wouldn't be possible in BtS (AI goes for religions earlier in BtS), and it's not an irresponsible risk.