I've been playing a few games of Civilization 5 in attempts to make the fastest possible tourism victory. My previous game pulled it off with Brazil on turn 136 in the year 540 AD. That's fast, but I can see how to go even faster.
I'm intending to share this report with a wider audience than my usual folks on Realms Beyond, so I'll go over some basic principles first. Regular readers can skip to the report.
Although I don't care to play directly for the Hall of Fame at CFC, I do follow their eligibility rules for game setup. This sets a reasonable baseline for game settings, putting out of bounds some abuses like cramming in an enormous number of city-states or playing without barbarians. Within those guidelines, here are my preferences.
I favor Emperor difficulty, level 6 out of 8. You could go faster (especially for tourism which depends on AI culture) on the lowest difficulty levels, but that becomes a silly solitaire exercise. Emperor is the sweet spot where the AIs pose some challenge, but don't run into carpet-of-units degeneracy. Emperor AIs also have a decent amount of money to pay you for resource sales early on.
I always use the Inland Sea map when playing for competitive fastest-finish stunts. This map has many advantages. It most consistently produces large swaths of rivers, for flood plains, see below about desert for religion. It has a predictable macroscopic layout, where it's easy to keep scouts running at full speed for a long time around the edges of the sea or map, you always know which general directions to go without any dead ends to get stuck. It's easy to find all the civs and city-states. It allows these scouts to find a lot of ancient ruins freebies, as the AIs often go a long time without getting out to the edges of the map. It has all good terrain, no dead zones of tundra or ice or islands. And Inland Sea gives you only two neighboring AI civs to worry about.
More map details: Large size works out best, more ruins to find, more room to expand, and more city-states for quests. Huge is too much, it takes too long to find your neighbors. Hot climate is desirable to generate a lot of desert, see below about religion. (Rainfall does not affect desert; only temperature does.) 3 billion year age generates more hills, which we also want. It's set for the standard number of civilizations and city-states, as per CFC HOF rules. I prefer Normal game speed, even though the slower ones are more advantageous, but they just take too long in real time.
Next, religion. All my fastest games involve a particular setup for religion. This is the Desert Folklore pantheon, +1 faith from all desert tiles. This produces enormous amounts of faith, more than any other pantheon. It's typical to get the religion by around turn 60, far faster than any other approach. Desert is of course weak without flood plains, which is why Inland Sea is the choice, because it both creates large swaths of contiguous terrain and lots of rivers to flood the desert. And this is why we want the young world setting for more hills, since cities full of just flood plains can't get anything built.
So yes, I start-scum. I will roll many maps until finding one I like with the requisite desert setup, often 20 to 40 tries or so. If you don't like this or consider it cheating, you don't have to read this.
For a culture game, that desert setup and start-scumming feeds into a particular plan. It is the Reformation belief Sacred Sites, which makes all buildings bought with faith produce 2 tourism each. With the right religious setup, you can put out dozens of faith buildings, for enough tourism to overcome the AIs' culture production by the renaissance era.
Now back to Brazil. In my previous game, I saw how precipitously Carnival can accelerate the end date. I'm sure now that Brazil is undisputably the king of the fastest possible culture victory. There are two big optimizations from the previous game to do.
The first is through the social policies. The right track is Plan C from my previous analysis:
The key point is getting to the golden age at Representation in good time, with minimal investment into Liberty rather than all of it. The downside of this plan is missing all the good Liberty policies for wide ICS. But this is fine. The Sacred Sites plan requires ICS, but doesn't require it quickly. You just have to steadily build a settler every few turns to keep ahead of the available slots for mosques and pagodas.
The second optimization is in the Golden Ages, or Carnival as they are renamed for Brazil. I kind of blew off the Carnival from happiness last time, just letting it come early. But no, this is an important tool to get this right. I can easily make the happiness GA come slower this time, since we won't have any of Aristocracy or Monarchy or Meritocracy for happy. But it still has to come just fast enough. The tricky part is that the happiness GA has to come first before any other, because any other GA escalates the happiness cost counter so you'll never reach it.
If done right and precisely micromanaged, I expect I can get that happiness GA just about as Sacred Sites arrives. Then we chain perpetual Carnivals through to the end of the game.
How do you get a Great Artist by turn 120? You must prioritize Guilds, do not get distracted by Theology or Civil Service. In fact you can even skip Drama & Poetry, you don't need the Writers Guild. I aim to reach Guilds by turn 100, build the Guild by 110, and that will produce the Great Artist by 120.
End of planning, on to the action.
After somewhere between 10 to 20 map rolls and a couple partial starts, I got this.
Lots of desert and some resources (that's salt behind the city billboard.) Salt calls for Mining as the first research, which makes for a fast start in improving a big tile and selling the luxury quickly, but does delay the shrine and religion compared to Pottery first.
I didn't get any ruins for a while, maybe one for a unit upgrade or something forgettable. On turn 8, a scout popped Mining from a ruin three turns before normal research would have finished, not exactly the most useful timing but I'll take it. Then on turn 9, I found not a ruin but a way to change this entire game:
EL DORADO. Holy crap, the single biggest early game accelerator. I do start-scum, but don't go as far as demanding something like this show up every game... but I'll sure take it here!
What do you do with El Dorado's money? Conventional wisdom is a settler, but I think conventional wisdom is wrong. A settler doesn't do much without a worker backing it up. I thought about this for a good ten minutes, but then decided to buy a worker and a scout. Having those items covered would let Rio build its monument while researching Pottery. You often don't necessarily need the monument this early, but culture was going slowly as I hadn't found any culture ruins yet, so that seemed like the right plan. Among the worker, scout, and monument, it's correct to buy the first two and build the last, because the monument has the worst gold:hammer conversion ratio.
On turn 13, a scout finally found a culture ruin to take the Liberty opener, a bit late (often this happens before turn 5), but good enough. Then t14 another ruin for 60 gold.
The Dorado-bought worker improved the salt. The fastest way to accelerate in the early game is to sell your happiness resources, you don't need them until you expand out to more cities. I sold the salt on turn 15 for almost full value (6/turn, full is 7), which is about as fast as physically possible.
Pottery and the monument both came in on turn 18, and I started the granary. Why not the shrine, isn't that critical for this dedicated religious game plan? Because this.
Here's a wild ballsy gamble. On that same turn 18, I found two more ancient ruins. I deliberately left them for an extra turn to pop on turn 20. Turn 20 is the earliest that the faith outcome is allowed to occur. I decided to stake everything on this gamble, that one of them would pop faith. That would be another enormous head start, getting the pantheon immediately instead of around turn 35, and also skipping the hammer investment into the shrine.
What happens if they don't pop faith? There's still a chance another ruin could do it. But if not, I'd just abandon the map and try again. I do this sometimes, make gambles that fail, more than you ever know since they never get reported. The chance of this working is around 2/6 for 33%; there are 8 possible ruin outcomes, of which any you've gotten recently are excluded so as to leave about 6 possibilities, and I get two bites at the apple.
Anyway, here we go...
Jackpot.
So the build order at my capital was scout - scout - (bought worker and scout) - monument - granary - (bought settler) - warrior - warrior - archer - shrine - settler - (bought settler) - Pyramids.
Tech path was Mining (popped) - Pottery - Masonry (marble) - Archery - Animal Husbandry (trade route) - Bronze (find iron) - up to Philosophy.
Social policies: I got the Liberty opener on turn 13 with the culture ruin. Citizenship (the worker) followed on turn 20. Then Piety, and we'll check in later on that.
One more helpful ruin checked in, on turn 23 popped survivors to take my city to size 4.
25 turns in, here's a look at my galactically fast start. By turn 25, I have two workers, five scouting units, my pantheon Desert Folklore, capital working 4 improved tiles (two salt), with a granary and +8 food surplus. And look at that gold total, which came from the usual city-state meetings, the salt sale, and a bit left over from El Dorado... that's just about enough for a settler on top of all that.
That was a great start, but can we keep up the momentum? My area here has a decent but not spectacularly large amount of desert for Desert Folklore. Remember that bare desert is useless and even all those salt and dry wheat resources aren't worth working either; what really counts are flood plains and hills. This should be good enough for a competitive pace from here forward; we should get up to about 40 faith/turn from Desert Folklore and then the faith buildings can take over production from there.
The one thing missing, though, is luxuries. My capital has the usual two, but beyond that are only that far-off cocoa and even farther spices. What this requires is keeping an eagle eye out for luxuries available from the AIs. On turn 25, Germany turned up with silver available and I swapped them my second salt for it. A while later I also paid cash for a gold resource import.
Okay, just in case things weren't going well enough, two more boosts happened on turn 41. One was yet another ruin that popped 60 faith towards a Great Prophet. The other was a worker steal from Vilnius there. Yes, I'm blowing up my friendship with the mercantile CS, forfeiting a bit of happiness... but actually I need to slow down my happiness accumulation for Golden Age purposes. The counter is already at 300, but I must not let it reach 500 and start Carnival until we get Sacred Sites.
Yet two more ruins a bit later popped 90 gold (turn 46), and Trapping tech (turn 48)! You can find these amazingly late around the edges of Inland Sea.
The next thing that happened was a Declaration of Friendship from Germany, on turn 42. I sold him my marble for nearly full cash value, and with it bought my third settler. Rio also built a settler the normal way, so here we are with two more cities.
Carefully treading the knife edge on happiness, at just 1 happy headroom. Sao Paulo is built on top of salt and Brasilia is on the cocoa. This is a key move in Civ 5, build directly on luxuries to make them available instantly. If we had to wait for a worker to hook up the cocoa, we'd be unhappy for several turns. Salvador has no luxury and is mostly a space filler; its location is to make sure I exploit all the flood plains in the area, not repeating the mistake in my previous game.
Rio de Janeiro has the Pyramids underway, which would complete as scheduled. This is always worth it with marble, costing just barely more than the two workers, and stacks very nicely with the Citizenship policy.
The other important thing in that picture is my Great Prophet, who now established my religion at the very early date of turn 51 1960 BC. That's what Desert Folklore does. Already at +12 faith/turn.. there just isn't any other way to produce that much this soon. I took Pagodas as the follower belief as planned; my overall plan requires both this and Mosques, but the AI seems to prefer Pagodas more and there's less chance they take Mosques before you can enhance to get it.
Turn 57 Poland also friended me, perfectly timed to buy the horses south of Rio just as I hooked them up.
Always love it when this happens, a scout lucked into an easy barbarian camp clear for a quest. However, I didn't get to return that settler to its rightful owner, because I hadn't contacted them yet.
Continuing the build order at my capital: Pyramids - settler - library - temple - caravan - stone works - Oracle.
Another overview with lots of stuff happening. One item was the spearman in the middle which came from Sidon thanks to that quest, perfectly timed just as I'm having trouble pushing outwards against barbarians. He would clear that barb camp north of Brasilia.
Desert Folklore is up to 26 faith production. I grew the cities well this time. In the previous Brazil game, I got a little carried away with the idea that you don't need to grow the cities for a culture victory, and didn't push faith production as much as I could have. But that went correctly here as I took care to use all the flood plains.
Brasilia has converted to my religion, despite not using a missionary. I held Brasilia at size 1 until this happened. It turns out that it takes exactly 17 turns after founding the religion, which makes sense, it takes exactly 100 pressure to convert one follower and 17 turns * 6 pressure = 102. This is well worth doing, even though it gives up about 20 potential food, it saves 80 faith worth of a missionary and then continues to snowball quickly. The combined pressure from Rio and Brasilia would then convert both Fortaleza and Belo with just small delays to growth, and then all the interlocking pressure would quickly convert everything.
Sao Paulo is building a settler. This is correct. For a game with this short time horizon, all a city needs is the shrine and monument and temple (which wasn't available yet.) Forget the granary or library or aqueduct or anything with such a long-term payoff; there will be no long term here. What's more important is to drop down more cities for more Desert Folklore and more slots for mosques and pagodas. So build settlers.
And also here is my second Great Prophet, to enhance the religion. You need to do this ASAP, before any missionaries or faith buildings, to grab both Mosques and Pagodas before the AI can get either.
For the founder belief, my usual choices are Tithe for a long game or Ceremonial Burial for an ICS style game constrained on happy. But here, I tried a new approach, going with Initiation Rites instead. Tithe (2 gold/city/turn) exceeds Initiation Rites (100 gold on city conversion) after 50 turns and is preferable for any long game... but the time-compressed world of this game would reach victory less than 50 turns after most city conversions. Ceremonial Burial is also something I can live without thanks to the mosques and pagodas for happiness. And actually, Initiation Rites can even just be converted to happiness if you need it, worst case is that I spend 5 cities worth of Rites (500 gold) to buy a colosseum that would match 4 cities worth of Burial.
Technology: After the classical techs for the various tile improvements, Philosophy was the target for both temples and the Oracle, reached on turn 66. Then the right move was to beeline to Guilds, for the Artists Guild.
I got a selection of typical city-state quests: two were pleased by the Pyramids, one by finding a natural wonder, one by a science quest and one by a faith quest. I also even spent a trade route for a quest, didn't really need an internal food route anywhere. One maritime CS came into alliance, and then I actually had to pay to keep it, needed both its luxury and the food.
Social policies were advancing through Piety. Turn 35 was the opener, t57 Organized Religion, t74 Mandate of Heaven perfectly timed for the first pagoda, t89 Oracle for Religious Tolerance, perfectly timed for t90 Reformation for Sacred Sites.
After enhancing the religion and converting my cities passively, all my faith production could go into buying the buildings. I had to prioritize pagodas over mosques, unlike the previous Brazil game. Not to keep the empire happy, that was manageable well enough -- but to push enough surplus happiness to make the Golden Age. I had to crunch hard on pagodas and get my surplus happiness up to +10 per turn for a while.
On turn 90, here's another major overview with a lot of stuff happening. The major milestone is the policy for Sacred Sites. And you can see the happiness counter timed very well just about to kick off the Golden Age. I have seven cities with two more settlers on the way.
I also just reached Guilds, to start the Artists Guild. It will finish on turn 97. Carnival also applies a +50% bonus to Great Artist production. With that, the city will produce 8 + 50% = 12 artist GPP per turn, spawning the artist in just 9 turns, on turn 106. Perfect timing, just ahead of where I originally planned.
But the biggest thing going on in that picture is that enormous faith production of 80/turn. My previous Brazil game had only 24 at this time! This came from a mix of all available sources. Desert Folklore steadily added more by growing onto flood plains and desert hills. But what I really did correctly this time was build shrines and temples first in every city. Seriously, faith is everything, don't get distracted by longer term build options or even monuments, just temples and shrines. They not only produce the faith you need for this plan, they also beget more faith by way of the mosques and pagodas.
On turn 94, just about exactly as planned, my accumulated happiness reached the threshold to kick off Carnival. I now had this precisely planned, to reach Representation with the next policy and then to spawn a Great Artist, to chain at least three consecutive Carnivals. What does this do to the projected victory date?
Already down to just 25 turns. Splendid. That 36 tourism is from 9 faith buildings, doubled by Carnival.
But what I didn't foresee was how that would continue to drop even faster. My faith production neared 160, almost a mosque per turn. Within another 10 turns, I had added another 9 faith buildings, doubling the tourism I'd had at the start of Carnival and halving the victory date yet again. With the gold from Initiation Rites, more resource sales, and the golden age, I bought more settlers to get more mosque/pagoda slots.
In fact, remember that Great Artist I so carefully planned and beelined for? Turns out he wasn't even needed. The win horizon came within the time frame of the Representation golden age, so didn't even need to use the artist for another, so did a great work instead for the bit more tourism.
Culture Victory on turn 109. That's 150 BC. Won the game in BC years. Holy crap.
I think I might be done with this little escapade now.