MAP SETTINGS:
Difficulty: Emperor or Demigod?
Map size: Large
Large map = 12 Civs (at least if you fully populate it -- but
not doing that kinda feels like cheating, plus it makes resources "scarcer" relative to land area) = more (potential) trade routes = more processing time required = turnsets take increasing time to play, especially during the mid to late game. So for an SG, a Standard map (8 Civs) might be preferable, to reduce interturn calculations, so people don't get bored playing their set (or waiting for it to come around!)
As for difficulty, don't forget that for 100K (or 130K), it's not just hitting that Culture-target, you also have to have
2x as much Culture as the next-best Civ. At DG, not only will the AIs be able to grow/ build/ research ~30% faster than you, they also start with an extra Settler. So their core (towns) will likely be established, and at the point of starting their Temple-/ Library-building (especially if REL and/or SCI), that much earlier than yours will -- and they will thus
also tend to hit the magic 1000-year Cultural-doubling earlier.
So unless you manage to keep them all isolated/ backwards, or keep them from expanding, winning by Culture at DG is not likely to be at all straightforward. That said, I do have to admit that a 30% Archi-map -- especially if Large --
would help with both of those 'aims' -- plus the AI also generally won't settle closer than CxxxC unless circumstances demand it, so will tend to have fewer towns per land area than your SG-team.
For a first attempt at this VC, though, I
still probably wouldn't dare go higher than Emp.
*Clucks faintly*
Babs (REL + SCI) are an obvious 'easy' choice for a 100K attempt, since they get nearly all their Culture-buildings half-price, but the Bowman UU near guarantees an early (likely still Despotic) GA, without guaranteeing the simultaneous fast conquest to subdue the neighbours and populate their lands (especially if it has to be done overseas). OK, the Temples will go up
super fast, but the GA will probably already be over before Lit is researched, never mind Monotheism/Education.
Agricultural is always a useful trait, and seafaring will be handy here.
On an Archi-Map, SEA does seem like it would be really useful, but wouldn't you want at least
one of the cheap-Culture traits?
AGRI + SEA = Dutch, but obviously they get no cheap Culture -- although lots of food/pop to whip into early Temples, and possibly later Libs, if Philo is used to take Lit instead of Republic? (Heresy, I know!)
SEA + SCI = Byzzies. I've seen them recommended for 20K games (to build the coastal GWonders in Byzantium, IIRC), but how good are they for 100K? I do enjoy flaming land-units with Dromons, though...
SEA + REL = Spain. The Conquistadore is a fairly poor UU for warmongering at the time when it becomes available, but for 100K, most of the conquering would have had to be done (long) before that point anyway, right?
Celts (REL + AGRI) are also a regular goto for 100K, I believe (at least on Continent/Pan maps)? Not SEA, but it does seem like one or the other of those traits would probably be better sacrificed, and AGRI
is generally stronger than SEA, regardless of map-type.
AI civs: currently random
Rather than Random, I'd suggest maybe picking rivals who are neither AGRI, SCI, nor REL, which leaves:
- SEA
- English (+COMM)
- Portuguese (+EXP)
- Carthaginians (+IND)
- Vikings (+MIL)
- MIL
- Romans (+COMM)
- Mongolians (+EXP)
- Zulu (+EXP)
- Chinese (+IND)
- IND
- French (+COMM)
- Americans (+EXP)
- EXP
That list would allow you to fully populate even a Large Archi-map. On a Standard Archi-map, I'd probably consider ditching the 4 SEA-Civs, just to keep the AIs isolated from one another that much longer.