Imperialistic settler first: mythbustered?

[vicawoo]

Settler first with imperialistic (not starting on a 2 hammer city tile) working a plains hill can make a settler in 17 turns, while a worker takes 15 turns. A new city produces 4 yield per turn, pretty clever eh?

Well, the problem is that unless you can worker steal, both your cities probably end up making workers for the next 15 turns, and you end up behind in worker turns. And unless you're Joao, your capital's mighty 6 hammer per turn settler production drops to a more modest 4 per turn.

So are delayed worker turns worth the fast second city?

Best (worst?) case comparison:
To compete with settler first speed, we want to get a settler as soon as possible after a worker.
Assume we can get bronze working in time for chopping, yes, that's kind of an iffy assumption.

Results:
With bronzeworking first (requires mining), worker first with 4 chops beat settler first (by 53 hammers)


The analysis
Bronze working first
Worker first finishes in turn 15.
In 8 turns it can chop 2 forests, for 40 hammers, 60 imperialistic.
In 8 turns, the capital can put 6 x 8 = 48 hammers into a settler. So we get our settler out at turn 23.
Therefore we lose 23-17=6 turns going worker-settler. In the settler case example, the capital would have dumped 6 x 4 = 24 hammers into a worker.

So our worker has to generate 24 additional food/hammers by turn 32 to tie settler first. Actually, our settler will produce 8 overflow (though this might get reduced to 6), so we only have to generate 18 food/hammers. This is actually pretty easy, we have time to improve a tile, worst case a mine, for +1 yield per turn starting turn 27, then chop a forest, and we will generate 25 additional food/hammers.

Caveat: to be truly accurate, we shouldn't compare getting 2 workers and a settler at turn 32, but 32+time it takes to move the settler to the second city. In the end it should work out, you just assume everything is a few turns later with your second city.

Improved 6 yield tile first:
If we improve a really good tile first, will we make our second settler about as fast? In the long run, improved tiles are better, so hopefully. Will that still be enough to beat out a settler first build?
Turn 15-worker built
Turn 19-6 yield tile improved. We probably won't have animal husbandry and bronze working, so that leaves irrigated wheat/corn or if we're really lucky a plains hill/copper mine. For now we assume a 6 yield food tile, no whipping. Settler update: 6x4 = 24 hammers
Turn 24, settler has 5 turns of 7 hammers (irrigated corn) or 8 hammers (wheat) or 10 hammers (plains copper mine), so 35 to 40 hammers normal case, 50 hammers special case. Add a chop, for 30 more hammers. Total: 24+30+35=89 or 24+30+40=94. So our settler will finish at turn 25 with irrigated wheat, so we have to make up 8 additional hammers by turn 32.
Turn 32:
Irrigated corn: settler finishes turn 26. 6 turns of 7 yield = 42. Second city has 6 turns of 4 yield = 24. 3 overflow from the settler. So we're automatically 5 hammers ahead of settler first. We have time for 2 chops or 2 tile improvements as well for +45 net yield
Irrigated wheat: 7 turns since turn 25, so 7 turns of 7 yield = 49. Second city produces 28. Add in up to 2 additional chops and 1 overflow from the settler, and we easily get there (78+2 chops = 118, second worker only costs 60).
So we end up with +58 net yield over settler first

Finally, a case with worker first, improve food first, grow and whip.
Turn 15: We lose 1 turn to anarchy
Turn 19: corn/wheat improved, we grew for 3 turns on a 3 food tile (guaranteed for corn, not so much for wheat) for 9 food. 3 turns to grow
Turn 22: size 2, settler production is now 6 + 4 = 10 with a plains/hill
Turn 23: chop finished, 10 hammers in settler, 30 incoming from chop
Turn 25: 60 hammers in settler, we can whip now. Will have 7 to 8 overflow + 5 (might get rounded to 3) from production, but our settler is 3 turns later than we hoped. So slavery's not faster in getting the settler out, but it should get the worker out faster.
Turn 26: So assume 10 overflow from the whip, and +7 production into a worker. We have 5 food in our bin at our capital if we had an irrigated corn, 2 with wheat. So it will take 3 turns to grow with irrigated corn, 4 turns with wheat
Turn 27: chop finished, but technically we have to stall growth to put it in a worker
Turn 29-30 (30 to 31 if we dump a chop into a worker): Grow to size 2.
Turn 31: 3rd chop. Worker has 40+8=48 with irrigated corn, we can whip.
Turn 32: 25 overflow. We also receive 6 turns of use from our second city if we can settle immediately, so +24, for a net +49 over settler first.

Example incoming

 

[vranasm]

nice I always like your analysis even though they are sometimes a bit too tough to read and not clear where they actually aim.

What I find too disturbing about "settler first" or "settler size 2 soon" are actually 2 things.

1) How do i know where I want to actually place 2nd city?
2) How I will battle the commerce drop tied to too soon city 2, which can lead to <9 commerce too soon (no pottery, writing yet) when I run out of builds due to too big effectivity?

 

[vicawoo]

We start with our unworldbuildered start. I'm going to add a corn north of the lake, so I can test a corn first build and a wheat first build

Spoiler :

As we anticipated, we can't get bronze working right after the farm finishes. Time to restart the test with bronze working first.

Spoiler all chop, all the time :

With bronze working first, we build our settler at turn 23. Math works!


Our turn 32 results:

2 cities, 2 workers, and an extra 32+21=53 hammers. Note that settler first would have 2 cities, 1 worker, and 1 worker 1 turn away from being built in the second city. So really it's +57 hammers.

Ok, so we can't tech agriculture-bronze working in time. So let's improvise and improve 2 six yield tiles, after which bronze working will be ready. Will we still beat settler first?

Spoiler 2 6 yield tiles first :

Of course we stop growth at size 2


Settler out at turn 29, with the help of a chop. Can we make it on time?


Turn 32-33 results

So in the case of settler first, on turn 33 we will have 2 cities, 2 workers, and 4 yield in our first city, and 1 worker turn.
With our 2 improved tile approach, we will have 2 cities, 2 workers, 17 hammers in our capital, and 8 hammers in our second city. So we're up 25 hammers, at the cost of 2 forests.

In the slightly longer term, our capital will be producing 11 yield, which is 7 more than the settler first approach until it can improve the corn (turn 37), so we gain an additional 21 yield. After that it will have 5 more turns at 11 yield vs 7 yield, so that's another 5 x 4 = 20 yield.
So we beat settler first in the short term, and we pull further and further ahead, enough to offset the loss of 2 forests.

When does this work:
This 2 improved resource approach will work if you start with agriculture OR mining, and you can improve 2 improved resources. Bronze working should finish about when your second tile is improved.
Bronze first will work if you start with mining

When won't it work (in the short run):
If you don't have start with mining, have pasturable tiles, and can't get animal husbandry before your worker.
If you don't start with mining and have bad tiles (unirrigated rice), flood plains/sugar/bananas).

When aren't you sure:
you start with fishing/mining and you go fast workboat->whipping out settler.

 

[vicawoo]

nice I always like your analysis even though they are sometimes a bit too tough to read and not clear where they actually aim.

What I find too disturbing about "settler first" or "settler size 2 soon" are actually 2 things.

1) How do i know where I want to actually place 2nd city?
2) How I will battle the commerce drop tied to too soon city 2, which can lead to <9 commerce too soon (no pottery, writing yet) when I run out of builds due to too big effectivity?

1) you just have to hope you find a reasonably good city close by. Change your scouting pattern to a much tighter circle around your capital instead of venturing further away.
2) Your commerce will be lower, so you can't expand anytime soon, so it's simpler to just grow out your 2 cities for now.
With lower commerce, you don't have a lot of time to get economic techs. So if you go animal husbandry, bronze working, and archery before your second city, you're going to crash. If I'm afraid of crashing, I'll only go one "bigger" tech (animal husbandry or bronze working), especially if I have to tech things like mysticism, hunting, mining, and the wheel.
Finally, using mysticism to get fail gold on stonehenge (and later polytheism ->temple of artemis) is a good way to convert excess hammers into beakers, and it solves that excess hammer problem very nicely.

 

[JammerUno]

It's totally logical to assume those forests magically disappear when you go settler first.

There's also the case of non-1 city tiles. If your second site is 2 with a 3 yield tile to work, the worker will take 12 turns.

So a 15 turn settler, followed by a 3 turn track to the site, settle on t19. First worker at t30, second at t31. Second city will usually add to commerce.

I usually aim for a 2 city for my first one, precisely because I want it to produce a worker quickly.

 

[VoiceOfUnreason]

How important is that river that you added?

If you are already constrained to BronzeWorking, and have a 3F tile on the go, it might be worth mathing out the solution that whips the worker. I don't think you can necessarily make the settler faster that way, but you may be able to up the total yield (subject to some additional constraints about how you spend it, if you want to avoid hammer rot).

 

[Habitus]

Well with the Lake there it basically gains him commerce, and once farmed health. So it could affect timing with techs, but map size and difficultly all do that anyway.

 

[vicawoo]

I added the river to see if I could get agriculture-bronze working, but of course it's only like a 10% increase where I needed more like a 33%. I believe without 1 commerce on immortal we got bronze working on 15, so on deity you can still get it by 16. With 1 commerce I think it was 14.

I grew averse to growing then whip is unattractive (like in my last example in the first post), because it requires both agriculture and bronze working. You can't revolt to slavery while improving the corn, your worker can't chop while waiting for the bronze, so in the short run you're pretty close to just a straight up settler first build.

Growing off a 3 pop tile should be:
8 turns to grow
5 yield per turn building a worker, so 6 turns to 30
1 turn anarchy (assuming you can get bronze working in 14 turns, which means 1 commerce on immortal, may be late on deity)
Turn 15 whip
Turn 16 worker, 5 overflow, 2 food in the bin, so we can grow in 7 turns. Though we have to stall every 4 turns for chopping, so it will be 8 turns.
So turn 24, it should be, 36 hammers in the settler, next turn we get +30 hammers from a second chop and 7 hammers from the city.
So turn 25 we have 73 hammers and can whip, which isn't terrible. +45 hammers, +6 production, so there will be 24 hammers overflow.
Turn 26, we have the settler and the overflow. Not sure if the overflow will get reduced due to the imperialistic bonus, so it will be 16 or 24. In 6 turns, we will generate 24 more hammers for the worker and potentially 2 more chops.
So if we assume reduced overflow, that's 64 hammers at turn 32 in the capital, and about 5 turns of worker production in the second city, and 2 chops (for 4 total).
It seems worse than just chopping, since that gets the settler out at turn 23.

 

[JammerUno]

Why grow at all?

t15 worker, start settler with 6/t
t16 worker starts 1st chop, settler at 6
t18 worker chopped 20, settler at 18
t20 worker starts 2nd chop, settler at 60
t22 worker chopped 20, settler at 72, chop will finish it

edit: Habitus makes a good point

 

[Habitus]

With Imperialistic shouldn't it be 30 a chop since it gets the multipler iirc infact t22 is correct at 72, and t20 should be 60 if you are including the mutiplier.

 

[Keilah]

Slight necro 'cause this thread was linked to:

worth pointing out that if you settle close to the cap along a river or coast so that you start with a trade route, your second city will net you +1 commerce instead of costing you -1. Bonus if it can work a 3-yield 1-commerce tile such as a floodplain.

 

[plasmacannon]

It's totally logical to assume those forests magically disappear when you go settler first.

There's also the case of non-1 city tiles. If your second site is 2 with a 3 yield tile to work, the worker will take 12 turns.

So a 15 turn settler, followed by a 3 turn track to the site, settle on t19. First worker at t30, second at t31. Second city will usually add to commerce.

I usually aim for a 2 city for my first one, precisely because I want it to produce a worker quickly.

If, he settled his 2nd city 2N of the corn, this would be the 2 hammer city you are describing.
Also, I see a desert hill on a river with at least 2FP and 2 Incense, another good city site, when, calander gets researched.

Good workthrough vic.