Settler Space Race Theoryfight

[Tristan_C]

This type of game caught my interest for the sheer entertainment value. I want to see how early I can get the launch date on settler, where as you know, AI competition is trivial and research costs are reduced to their barest limits. I was wondering if anyone could advise me on what type of map to play, who to choose, and how to proceed.


Speed - Some questions answer themselves. Marathon, almost definitely. Can't think of a reason not to go marathon.

Huts & Events - Off, I think. I don't want to save and reload until I get settlers from all the luckboxes, nor do I want the launch date to be based on how many free cities I popped in 3000BC. I suspect a lot of those HOF entries on this type of game have this stuff enabled, and take a civ that starts with hunting. Probably gives them a hundred-turn edge.

Map script - Thinking of rainforest, pangaea, or great plains. Inclined towards a crowded duel rainforest, with a lot of capitals. As an alternative to creating large numbers of settlers, we can harness marathon's cheap axemen or chariots. Settler AI don't have archery or hunting. Our units can travel in small packs of 1-2 units capturing AIs (capturing an undefended AI before it reaches size 2 razes the site).
Wondering if there is a case for great plains or pangaea which I am missing.


The inclination of this civ player is to skip the religion tech tree including, if at all possible, Monarchy... whereby will flow from Settler's generous handicap (6 "enjoy life"), forges, Confucianism, and calendar resources. Backfill as needed, ideally with AI trades. What do you think? And as for size, I don't really know how many cities to aim for. I suppose the number is capped hard by the domination limit. I guess it's sort of lame to remove the domination VC purely on account of this. And who knows, if I happen to see settler Gandhi run away with the game and press Dom with his stack of 85 elephants 2 turns before I can launch, I can post this stunning piece of news here.

Anyway, peoples' favorite subject is traits & wonders: how they would be applied to the type of map you believe is best for early space.

• FIN - Always strong, and especially so with water and grassland, but will it save as many turns in the beginning as some of the others?
• CRE - one of the options that could save a ton of early turns getting cities running.
• EXP - same; granaries whip @2, and on a high-food map like rainforest, will quickly become valuable.
• PHI - This definitely does not help us get stuff running. But I am guessing that building loads of academies pronto is an important secondary objective of this match?
• IND - there will be lots of forges, but not so many wonders, no failgold, and of course the game very obviously lacks an IND/PHI leader. So, perfectionist GP farming falls behind getting cities large quickly.
• SPI - a strong option on marathon, but even zigzagging OR/Pac or slavery/CS I think there are better options.

Settler starting techs: agriculture, mining, and the wheel. Any civ that comes equipped with fishing and myst has an edge here.
Wonders:

• Mids - I am reluctant to approve of a SE. With stone, I could imagine running a SE that evolves into state property with someone like Peter. I don't know if this would be as fast as FIN cottage spam. The question is if the strong early/mid ages negate the rather weaker modern age.
• Oracle - Available later than normal. Education, possibly? Too early, and the conversion isn't as good. Too late, and it doesn't save many turns. I always find education to be one of the longest researches...
• TGL - not a good operational window on settler, but still to be considered.
• MoM + Taj - naturally, and a tight case for spending three gp on GA's.
• HG - probably an LOL 15 Popspasm here.
• Kremlin - might save time by getting all the buildings finalized. Might not.
• Cristo - late-game gambit on microing civics for max effect. Resides along the space tech line.

Late-game issues: Mining is normally a strong corp. On a rockier map script than rainforest, we could probably draw atrocious production from Mining by controlling half the map and trading with the rest. But in Rainforest, state property is likely to exceed it by date, ease of setup, and overall returns. Cereal & CJ are out, I suspect, because the requisite techs are not strictly required. And no matter how much rice there is, Sushi without seafood doesn't elicit rave reviews.

Well, that's my spiel.

 

[VoiceOfUnreason]

Wondering if there is a case for great plains or pangaea which I am missing.

I wouldn't expect GP to be a good choice at all -- too much variation in starting locations? I suppose it depends on how much start rerolling you are going to do.

It's very much going to be a research game, the AI isn't going to be able to keep close enough to be a threat, and the production at the end is going to be pretty incidental -- especially if the map is large enough for a number of production cities.

The big bottleneck has got to be research, so I'd guess FIN/PHI would be your best choice (spamming academies), followed by FIN/CRE or FIN/EXP.

I agree SPI doesn't seem a good fit here - you're basically going to drop into a research configuration and park there.

There ought to be a lot of prior art here among the HOF set.

 

[vranasm]

I would advise you to search for the thread "Oxfordless Hatty"

It's on Noble/Normal which skews a bit the comparation, but I think it's a good start.
It's Great Plains map.

I got there my best space time up to now (~1630 victory), but I didn't follow the rules (I built the OU :-D).

To me it seems that ~30 cities is best.

Fin is very strong imo. watermills, windmills have better commerce output.

 

[Habitus]

It's Big n Small with Corp abuse or Highlands with mills n state property as the normal approach iirc

 

[Tristan_C]

Regarding Great Plains, its main attraction seems to be in a case where you have the maximum possible number of AI players and you spawn on the eastern side. Under these conditions, you are positioned on heavily forested green land which you can control without triggering an inadvertent domination victory. From there, though, it appears you will need to set up a concern on the west side that brings in needed materials.

With only 1 AI opponent, domination is ~74% land area. The serious drawback of having 1 opponent is that there are only 2 capitals available. Resources will be relatively scarce, and the game hangs more on the capital. By contrast a Duel map with 18 civs is crammed with resources. But Domination is 51% in this case, requiring that you plan and measure the size of your borders more carefully. With fewer resources and even fewerer cities ready-made to capture, I would say that a 1-opponent game is a great deal slower coming on, and for low returns on the extra land at that.

I've read Oxfordless Hatty before, but was never able to open the save. Thanks for reminding me of it- I never thought it would apply. By the look of things, though, he either turned off domination or just barely scraped under the tile limit.

I'm thinking now of trying this first under some srs gimped settings. Pacal, No city razing and no domination perhaps- basically 16 cities are premade and all you need to do is walk your warrior or Holkan into it. Economy crashes around turn 80, followed by a strict FIN cottage spam regimen. I will see how that shapes up and might change settings in a later game for more "normal" constraints.

 

[Zx Zero Zx]

Darius is the best for this game.

 

[coanda]

Also in wonders: Stonehenge. If you're settling 100 cities, you can save an awful lot of hammers by getting it.

 

[iggymnrr]

Inclined towards a crowded duel rainforest

Likely to be the fastest desiring no huts. May want barbs off and no city razing. Pacal. Pacal. Pacal. Not many wonders will get built. National Park is handy on this size map which can be 6 pop whipped immediately prior to switching out of slavery into ~16 caste specialists. (Forest preserves on jungle tiles and a 3 or 4 food city.) Late bulbing/GAs can be an art which I have not looked to closely at.

Basically, find a good start. Take BW really early. (Oracle education?) Mapfinder if necessary. Map size isn't suitable for Darius and Pacal gets the +2 health without apothecaries. Cheap granaries and workers make for good whipping practice. Lots of workers. ~3+ per city. Mauso, late-ish Handing Gardens (around SP time) and Taj and Notre Dame maybe for wonders.

And who knows, if I happen to see settler Gandhi run away with the game and press Dom with his stack of 85 elephants 2 turns before I can launch, I can post this stunning piece of news here.

Build the internet and get monotheism.

Also in wonders: Stonehenge. If you're settling 100 cities, you can save an awful lot of hammers by getting it.

Use music or religion instead.

Edit: I think Pacal starts with 4 tech on settler. (No hunting/mysticism or hunting/fishing?)

 

[Amao]

I would advise you to search for the thread "Oxfordless Hatty"

It's on Noble/Normal which skews a bit the comparation, but I think it's a good start.
It's Great Plains map.

I got there my best space time up to now (~1630 victory), but I didn't follow the rules (I built the OU :-D).

To me it seems that ~30 cities is best.

Fin is very strong imo. watermills, windmills have better commerce output.

I introduced "Oxfordless Hatty" to a local funsite. And a few there have made pre-1550AD launch by either picking Darius or MM with Egypt. The common sense was to devote yourself into an early rush, and have enough workers. Other variation seems to get an early Feud to have 50% quicker workers to lower the worker need.

However, personally, I'm not a fun of a lower than Noble diff levels. For all those HoF records, it just falls into Settler/Worker hut hunting, no fun for me.

 

[Amao]

Likely to be the fastest desiring no huts. May want barbs off and no city razing. Pacal. Pacal. Pacal. Not many wonders will get built. National Park is handy on this size map which can be 6 pop whipped immediately prior to switching out of slavery into ~16 caste specialists. (Forest preserves on jungle tiles and a 3 or 4 food city.) Late bulbing/GAs can be an art which I have not looked to closely at.

Basically, find a good start. Take BW really early. (Oracle education?) Mapfinder if necessary. Map size isn't suitable for Darius and Pacal gets the +2 health without apothecaries. Cheap granaries and workers make for good whipping practice. Lots of workers. ~3+ per city. Mauso, late-ish Handing Gardens (around SP time) and Taj and Notre Dame maybe for wonders.

Build the internet and get monotheism.

Use music or religion instead.

Edit: I think Pacal starts with 4 tech on settler. (No hunting/mysticism or hunting/fishing?)

Here comes my idol. Question, what's the use of NP? You do seem to prioritize it. What do you use all these GP for? And how many GP can you get from the NP city? This together with the final parts rush building are the 2 major parts i don't follow in your original strategy.

 

[iggymnrr]

Here comes my idol. Question, what's the use of NP? You do seem to prioritize it. What do you use all these GP for? And how many GP can you get from the NP city? This together with the final parts rush building are the 2 major parts i don't follow in your original strategy.

I'm not the National Park specialist. (I can tell you who is.) It is used mainly for getting a 4th golden age in time and some late bulbing. IMO it gets stronger as the map size gets smaller and is well suited for a jungle map with lots of food, rainforest. There is the assumption state property will be used. The basic idea is to lib in biology and hard tech communism. If timing is good, when communism is in, whip and switch. (Slavery->caste. Don't forget SP!) Research at marathon settler isn't the real problem. Research and building at the end is tricky. Late golden ages are obviously helpful. But when a person doesn't have enough cities to build all parts simultaneously, research drops and bulbing is again very useful.

That being said, achieving really fast times excludes a fourth golden age. The NP is no longer useful.

 

[Amao]

@iggy: that's very nice explanation.

 

[Tristan_C]

Pacal. Pacal. Pacal.

I came into this wisdom, too, after thinking for a while that Elizabeth would be my choice. Maya will utilize all aspects of EXP on the rainforest script, which may well result in earlier GS's than the PHI leader— not to mention earlier-everything-else. Then there is the Holkan: OK on archers, usually fatal on warriors, and needs no hookup. Possibly useful here.

The National Park may not even be possible in my proposed crowded game. It is going to depend on whether the engine loads one or more (preferably more) spawns with healthy forested tiles.


---------------


Here's an early foray into this challenge under the gimped settings I described earlier. What happened soon was a bit of a laugh. One note: I don't pretend to be a fraction as good a civ player as most of the people who have posted so far, so just take this narrative as entertainment, not instruction.

I rolled a custom game with the following:
Difficulty: Settler
Speed: Marathon
Map: Rainforest
Size: Duel
Shape: Flat
Me: Pacal, Maya
Rivals: 17 Random
Toggles:
- No Domination
- No City Razing
- No Tribal Villages
- No Random Events

Settler Bonus Techs: Agriculture, Mining, and The Wheel

lol... Sid is on a role. Why not just do an advanced start?​

Maya Bonus Techs: Mysticism and Mining

Our Maya techs are of no use here, except for the significant boon that they do not include hunting.​

Mayan UU: Holkan - Replaces Spearman, Immune to FS, No Resource Req., 70

This is a modest UU that could be harnessed through iterations of this challenge.​

Mayan UB: Ball Court - Replaces Colosseum, +3 , 240

Not a common base building in my games. If we need to throw hammers at to dodge Monarchy and create something permanent, it is an option... I suppose. My impression is that the hammer outlay is massive for what you get. We are not CRE.​

My only wish on a starting location is not to be absolutely in a corner. Too many rerolls, and my attention wanders away. After three tries where I was, indeed, absolutely smack-dab in a corner, the RNG pulled this one out of the hat, and I took it!



There is Hill Silver, indicating that we are still at the northern or southern edge, even if not in the corner. Green Forest Deer is a superb unimproved tile, something of an issue here since these settings scream Warrior-first. Teching Hunting first puts us partway to Holkans and discounts AH, without the annoyance of generating a scout. I scooted the screen around and determined we are north, with longitude indeterminable. All this aside, I do not yet see a long-term role for this capital on its resources alone, apart from being the epicenter of some kind of cottage spammering wet dream. But you see, I am ungraceful at this game. Any ideas?


I checked the terms of trial again, just to make sure everything is in order.



And the victory conditions:



...it looks like there are a couple of erroneous settings left over from when I last rolled a game of Civ4: Barbs, Agg AI, and Choose religions (mere flavor). Mostly no concern, but Agg AI might cost us. If the AI beelines military and spams it, our initial rush is halted by the stockpiling of archers, and we can't trade as much religious tech. Those are turns down the drain on this objective. I decided to let it slide, though. Anyway, Agg AI increases the normalized score.



Anyways - one last starting screen. Just to emphasize the fact that we are playing with 17 AIs on a duel map.



Moving the warrior onto the hill revealed what are undoubtedly two rival spawns, South-ish and West-ish, given that there is no jungle. To be honest, I'm surprised not to actually see a settler from this commanding overlook, but there is no question that they are close.

I usually like to name my first warrior. This one is named for the Viscount Althorpe Earl-Spencer, Princess Diana's grandfather, a man of quality and standing. As the Viscount was doubtless aware, VAES can also stand for Value-Adding Elimination System, appropriate for his task of roaming systematically around the edge of the map converting unnecessary AI cities into extremely necessary Tristan cities.

I settled in place, rather than ambling over towards the riverside grassland gems. Two reasons:

1. High delta on a city appearing 3N, 1-2W of the warrior.
2. I am a bad civ player and SIP pretty much regardless.​

Alright, that's enough introductory material. Turn 88 progress and commentary is in the spoiler, and 4000BC is attached in case you want to play this mess. Since it takes me forever to play civ, and forever+1 to post information like this, I cannot make many updates I'll do what I can though.

Spoiler :

Turn 88, 2680 BC, A Year Of No Particular Importance

No City Razing and a starting warrior were... eh... whimsical on this map. Economy was utterly crushed at turn 30 under the combined weight of 12 cities. My strat with the first warrior was to head for the corner and loop the edge, under the assumption that this crammed map would force a fairly uniform distribution of AIs. Warrior 2 was 15T. He started with nearby Persepolis and swung upwards to loop the map going the opposite direction. When a pair of Holkans came out, they bought me Cuzco, a second holy city. But now, the other AIs have archery and two of them have broadcasted Slavery, so no more Holkan rushing. Even without copper, they will whip extra archers if I approach.



Look at all that f'ing river. (and all the AIs that settled off them, which means roadwork )
Tech order was Hunting->BW->Pottery, at which point I had to halt research and stockpile gold. Mutal was warrior-worker-holkan-granary. Every other city was worker first. The economy crashed so hard, it wiped 100 percent of the gold and I would have had my first built worker go on strike had I not started working unimproved commerce tiles, like the capital's Hill Silver, as an emergency. I managed to improve London's Hill Gem and recover from there; the 12 workers started rolling out one-after-another, and I then 1-pop whipped a bunch of EXP granaries as well instead of piling up unimproved tiles. As of turn 88 I am pushing 85 commerce, squeezed AH, and have basically been saving a lot of gold trying to figure out what to hit next. This was an error worth about 45 beakers. Earlier Fishing & sailing will bring maybe 5-8 more and hook certain cities to copper. I will rush out axes with 1-pop whip overflows from workers and granaries (or vice-versa)

Below is the path of conquest using the edge-hugging method I described. Exploration was trivial since capping a city reveals everything 3 tiles out. One needed only move from town to town. Cities were undefended until turn 37. This is because AI's starting unit runs AI_SCRIPT_EXPLORE or whatever it is, and on settler a warrior costs the computer 48 - 11 initial . The EXPLORE script sure is convenient. On a test run with Elizabeth earlier, I even saw a warrior leave a city undefended after I declared war and approached the city. This leads me to make a wild guess that Settler AI is completely obstinate, but higher difficulties might enable the computer player to adjust its units' scripts based on nearby opportunity nodes; e.g., undefended workers and cities, enemies pillaging, unexplored terrain, etc.



I definitely could have microed everything across the board better; cities started getting highly involved while I was oblivious and just merrily capping more of them. I might even replay the same moves to this point, this time with proper micro.

All in all, this is a somewhat absurd match. No idea what date of launch it will bring. The settings are, as I worded it, gimped, but I'm sure it will continue to be an entertaining, if not exactly realistic, settler space race.

 

[iggymnrr]

Classic settings and definitely entertaining. The only downside is that the overcrowding seems to have cleared all the jungle. Where is the rainforest? I won't complain about not having to clear jungle though. Kind of looks like corps might be faster. Cities are overlapping too much and exec travel time will be fast. Either way a fast launch looks easy. I haven't dabbled on duel for a while, but 750 AD or faster?

 

[Tristan_C]

Hah, downside? The capitals clear a diamond-shaped space of jungle but the green remains, as do those capital-grade resource clusters. I know this pretty much preempts a NP, but on the other hand, a preponderance of capital-grade resources means we may be able to make a move on GPP from a wide swath of cities. I'm glad you thought of corps, because I definitely wasn't, and Mining draws 4 on each resource playing duel. With 15 such resources visible so far, that one will be rich. I would be glad of your opinion on Cereal Mills, which requires a 1-tech diversion and a GM, and also what you think of Wall Street. Though... perhaps Wall Street is the only way to pay for a Mining like that one.

 

[iggymnrr]

I would be glad of your opinion on Cereal Mills, which requires a 1-tech diversion and a GM

If Mining Inc. is that good, the rice for sushi must be pretty good. With no domination limit to worry about, the culture from sushi won't matter. Do you plan on building supermarkets?

what you think of Wall Street

Ha! You're asking this of someone who's considered not building Oxford. Now you want me to build Wall Street too? I will let someone else answer this. My general instinct is to keep finding ways to build workers, but they will quickly run out of things to do here.

 

[Tristan_C]

Ha! You're asking this of someone who's considered not building Oxford. Now you want me to build Wall Street too? I will let someone else answer this. My general instinct is to keep finding ways to build workers, but they will quickly run out of things to do here.

I know— you're precisely the person I wanted to ask

 

[iggymnrr]

I toyed with this some and decided on something extremely aggressive and patient.

Spoiler :

There are usually two turns a settler AI will produce it's first warrior, 18 and 37. So I decided to go for the center first killing everything I could while the 18 turn window was open. I wanted to get some of the far away AIs ASAP. So I settled in place and started on a warrior. I didn't know how hard the economy would crash so I thought sailing would be nice to get early. Fishing->sailing.

What I was looking for was the dangerous AIs first if possible. AIs on a brown hill or working dangerous tiles were priority. A little house can be seen on a worked tile. Both Spain and Persia were dangerous but the warrior from Mutal can hit Persia in time. So no backtracking. Onwards! What could be ignored and left for backup warriors were the safe targets. AIs working food tiles not on a brown hill. The Celts were safe (verified by line of sight from hill,) a backup warrior can take it so ignore for the time being. AIs near me were ignored since I figured there was time for a 3-4 warrior rush if they couldn't be taken safely.



After the 18 turn window I never attacked anyone unless I could verify them to be safe. By turn 36 I had 14 cities, one of them holy. Tech path Hunting->pottery. Then began the long patient stretch spamming workers. Every city went worker first (after warrior if they built one) followed by granary->grow to size 2->another worker. Somewhere late I had enough money for AH->BW. BW in at turn 90 and here I paused. 17 workers out the door with 5 more coming soon. Avoided strike entirely.

 

[MarigoldRan]

Darius is the best for this game.

Yes. Immortals massacre settler-AI. Apothecaries are actually useful because of fast factories. FIN+ ORG? Lolz.

EDIT: Wait. It's settler. Apothecaries are useless then.

 

[iggymnrr]

EDIT: Wait. It's settler. Apothecaries are useless then.

6 to 1 says I build grocers everywhere. Pacal's +2 health + grocers = apothecaries. With Darius on a standard size or larger map supermarkets will get built.