View Full Version : First worker/settler


Xin Yu
Oct 10, 2006, 12:07 PM
This guide discusses different early strategies when building the first worker or settler, from the first city. (We focus on the first one only to avoid the hot issue of whether w-w-s or w-s-w or s-w-w or whatever is the best sequence :) )

Assumptions:
a)standard map,
b)normal speed,
c)the first city has access to at least 1 square with 3 food, and at least another square with 3 productions (food + hammer).
d)bronze working is researched when slavery is needed for certain strategies. (Note: for a worker at turn 15, only possible at lower difficulties and/or with mining as starting tech)
e)anarchy turn for revolution is neglected for simplicity. (Assume the revolution eventually takes place, then in most cases the loss of one turn will happen sooner or later. Major differences are limited to the rare tight situations when you beat/lose to another civ by one turn for a key tech or wonder.)
f)the extra commerce from working a bigger sized city is not discussed.

1. worker

1a) build worker from turn 1:
1a1) takes 15 turns to finish (60 hammers / 4 per turn), zero overflow

1b) build worker after reaching size 2:
1b1) takes 8 turns to size 2, 2 extra food, 8 hammers for a warrior
1b2) start building worker. Takes 6 turns to get 30 hammers (5 per turn), change to slavery and whip
1b3) worker accomplished after 15 turns with 4 overflow
1be) total gain over 1a: 2 food, 12 hammers for a total of 14 productions

Rationale:
Each whip from a size 2 city converts 22 food to 30 shield = gains 8 productions. Subsequent whipping yields less (size 3: 24 food to 30 shield = 6 gains, etc.)
1b) whipped once and stayed in size 2 for 6 turns, hence the total gain was 8+6=14 productions.

2. settler

2a) build settler from turn 1:
2a1) takes 25 turns to finish (100 hammers / 4 per turn), zero overflow

2b) build settler after reaching size 2:
2b1) takes 8 turns to size 2, 2 extra food, 8 hammers for a warrior
2b2) start building settler. Takes 14 turns to get 70 hammers (5 per turn), change to slavery and whip
2b3) settler accomplished after 23 turns with 4 overflow
2be) total gain over 2a: 2 food, 12 hammers for a total of 14 productions (minus loss of production for the halted warrior).

2c) build settler after reaching size 4 (this requires 3 food-abundant squares and a forest):
2c1) takes 8 turns to size 2, 2 extra food, 8 hammers for a warrior
2c2) takes 6 turns to size 3, 2 extra food, 14 hammers for a warrior
2c3) takes 5 turns to size 4, 1 extra food, 1 warrior built and 4 towards the second warrior
2c4) start building settler. Takes 6 turns to accumulate 42 hammers, change to slavory and whip
2c5) accomplished after 26 turns with 7 overflow
2ce) total gain over 2a: size 2, 1 food, one warrior, 11 hammers for a total of 27 productions, but it takes an extra turn.

Rationale:
Each whip from a size 2 city converts 22 food to 30 shield = gains 8 productions. Subsequent whipping yields less (size 3: 24 food to 30 shield = 6 gains, size 4: 26 food for 30 shield = 4 gains, etc.)
2b) whipped once and stayed in size 2 for 14 turns, hence the total gain was 8+14=22 productions (8 of which translated to finishing two turns earlier).
2c) whipped once at size 4 and stayed in size 2 to 4 for various number of turns, hence the total gain was (6+4)+6+2*5+6*3=44 productions. An extra turn gives 5, makes a total of 49 -- 22 of which translates to an extra population.

Conclusion: If you have at least 1 food abundant square, then building a worker/settler from turn 1 is not the best strategy -- unless you cannot get bronze working in time.

Beamup
Oct 10, 2006, 01:42 PM
1c3) the next turn warrior finished with 29 overflow. City back to size 2 with 3 extra food. Start building worker. Unfortunately even if you work a forest squre last turn to get to 30 overflow you still cannot whip now (penalty for zero shield built).
...
2c3) assume you have worked the forest square, then the next turn warrior finished with 30 overflow. City back to size 2 with 2 extra food. Start building settler.

You can't get more overflow than the current build costs. Which means half these hammers are lost in both cases, and 1c and 2c come out well behind 1b and 2b.

Also, assuming Bronze Working after 13 turns is pretty iffy as a general principle. You can't just toss this off so casually.

Other than that, pretty good analysis.

Xin Yu
Oct 10, 2006, 01:59 PM
Thanks for poining out the overflow thing. That makes the double whipping strategy doesn't work. Come back to fix it now.
At lower levels and starting with mining you should be able to get bronzeworking quicker. But yes there are situations when you need 14-15 turns.

Beamup
Oct 10, 2006, 02:44 PM
At lower levels and starting with mining you should be able to get bronzeworking quicker. But yes there are situations when you need 14-15 turns.
My point is that's an awfully restricted set of conditions. And in many cases, it'll be rather more than 14-15 even.

Basically, I'm suggesting that you also consider the case where you DON'T have BW at the critical time. See how that changes things...

Dr Elmer Jiggle
Oct 10, 2006, 03:58 PM
c)the first city has access to at least 1 square with 3 food each,

You also seem to be assuming that there is a second tile with at least 3 food and hammers combined (ex. a forested grassland). I think that's a very reasonable assumption, but you should add that to your list.

d)bronze working is researched when slavery is needed for certain strategies.

Beamup has already addressed this. I think he's right. This is a pretty big assumption. You're basically limiting your analysis to civilizations that start with Mining, which means you're ignoring well over 50% of the possibilities.

e)anarchy turn for revolution is neglected for simplicity.

This is also a huge assumption. So now you've narrowed your focus to only Spiritual civilizations that start with Mining. As an example, if you include the 1 turn of anarchy, then 1b has a loss of 1 food, a gain of 11 hammers, and a loss of 1 worker turn (or something like that). That doesn't mean it's a bad choice, but it changes the equation significantly. When that 1a worker starts finishing his improvements 1 turn sooner, you start gaining back some of the hammer deficit.

I think Epic / Marathon speed also changes things by a few turns here or there. I haven't worked out the details though, so I'm not certain. You mention that in your assumptions. I'm just saying I think it's important.

VoiceOfUnreason
Oct 10, 2006, 04:03 PM
Basically, I'm suggesting that you also consider the case where you DON'T have BW at the critical time. See how that changes things...

Let's see - Mining gives a 20% discount on Bronze Working. So you've got (under those starting conditions) about 143 beakers of research to get done. Palace, city, free beaker means 10 beakers/turn, therefore 15 turns of research, so you aren't beating worker first unless you are getting extra commerce from one of your tiles. Edit: to clarify - the game probably shows 171 beakers needed for the tech, and the domestic advisor shows 9 research per turn, but when you look at your progress on the second turn it is listed as 12/171. Not the most intuitive of interfaces. I normalized the other way for my own sanity, but I don't promise that it always works.

If your worked tiles have one extra commerce (a floodplains tile, for instance), it's 13 turns of research, so you just barely make it. With two extra commerce (Oasis) you can finish in 12 turns. Growing into a second commerce doesn't save any time, though. Edit: hmm, I think I blew the math here. In real life 10 commerce in the city is 11 research, +20% is 13. 8*13 = 104, leaving 67 research to go, which is 6 turns at 13per, but only 5 turns at 14 per. Got to be careful with the rounding and scaling

Assuming that you start with Mining. Without mining, you've got another 7 or 8 turns of research to finish.

So in your initial city this works for India and Mali. Of course, the worker is going to be chopping or mining, since the other tech doesn't help.

For the other civs (China, England, Germany, Russia, Rome - wow, a lot of leaders start with mining), you are gaining 14 production at the cost of a one turn delay and a happy point that doesn't matter very much. (Of course, you are likely to want to take that Anarchy hit at some point; I'm not sure how to properly judge the cost of this).

England and Rome run into the same problem as before - Fishing gives the worker nothing useful to do.

Germany and Russia start with Hunting, which has smaller scope as a worker tech. In addition, the utility of scouts falls with time; I'm not sure I'll often want to wait 17 turns to get my first one out.

In a situation where your first priority is going to be mowing lawns, then the weakness of the second tech doesn't matter, and this looks like a win.

My overall take is that it's a useful trick to have in the bag, but I wouldn't expect it to be part of my basic strategy.

Xin Yu
Oct 10, 2006, 04:34 PM
You also seem to be assuming that there is a second tile with at least 3 food and hammers combined (ex. a forested grassland). I think that's a very reasonable assumption, but you should add that to your list..

You are absolutely right.



So now you've narrowed your focus to only Spiritual civilizations

Yes you'll lose a turn for revolution if not Spiritual, however you are going to get this revolution later which also costs a turn (assume you don't combine the revolution turn with, say, heredity). By that time the turn difference is gone, except for a turn of unit moving, and maybe slight difference in cultural boarder expansion. Shouldn't be a big deal.



Palace, city, free beaker means 10 beakers/turn, therefore 15 turns of research, so you aren't beating worker first unless you are getting extra commerce from one of your tiles.


Correct me if I'm wrong. I think the beaker count starts on turn one but your city starts growing on turn 2, so you have one turn extra worth of beakers. Therefore 15 turns is enough for revolution (because your city is at turn 14 of its growth).

malekithe
Oct 10, 2006, 05:11 PM
Let's see - Mining gives a 20% discount on Bronze Working. So you've got (under those starting conditions) about 143 beakers of research to get done. Palace, city, free beaker means 10 beakers/turn, therefore 15 turns of research, so you aren't beating worker first unless you are getting extra commerce from one of your tiles.
EDIT: Nothing... I didn't fully read your post...

VoiceOfUnreason
Oct 11, 2006, 12:00 AM
Correct me if I'm wrong. I think the beaker count starts on turn one but your city starts growing on turn 2, so you have one turn extra worth of beakers. Therefore 15 turns is enough for revolution (because your city is at turn 14 of its growth).

I don't think this is right at all.

I rolled a Chinese start with a 3/0/0 tile available. Both Bronzeworking and the Initial worker take 15 turns to build. And sure enough, in the year 3440, BW is at 168/171, and the worker is at 56/60, both listed as one turn to go.

Or I go the other way around - warrior first, then worker. At 3440, I have 168/171 on Bronzeworking, 2/24 in food, 30/60 for the worker, 8/15 for the warrior. So everything is in place, but I'm not ready to pull the trigger yet.

Now, if I was able to grow into an extra commerce (a 1/1/2 tile, for instance), the extra commerce brings in Bronze Working a turn earlier. At 3480, I have 164/171 on Bronzeworking, 2/24 in food, 25/60 for the worker, 8/15 for the warrior. So in this case, the extra commerce that you grow into does make a difference.

Xin Yu
Oct 11, 2006, 12:54 AM
VoiceOfUnreason you are right. I was posting at work and just didn't get it right.

Came back and tested it further. If your city has a 3 food 1 commerce land to work on (riverside flood land), then your research per turn can be as big as 13. So in 14 turns you total 182 beakers. Suppose you have another riverside flood land, then the 6 turns you are at size 2 will give an extra beaker each, making it 188 total. That is enough for emporer level difficulty (187 beakers) but not enough for immortal (195) and deity (202).

On emporor level there is a trick: at turn one manually select BW from tech tree and click on it. If you don't do it, then on turn 2 you only get 24 beakers rather than 26, which makes you one beaker shy of the emporer level requirement.

If playing monarchy, one riverside flood land is all you need. Without it you even cannot pull it off in prince (as in VoiceOfUnreason's example).

VoiceOfUnreason
Oct 11, 2006, 01:04 AM
On emporor level there is a trick: at turn one manually select BW from tech tree and click on it. If you don't do it, then on turn 2 you only get 24 beakers rather than 26, which makes you one beaker shy of the emporer level requirement.

Right - that's the 20% pre-requisite bonus being lost on the first turn.

Micro, micro man... I want to learn to micro man....

The Lardossen
Oct 11, 2006, 03:49 AM
It all depends on wood tiles.

W - W - S is optimum, but with 4 tree squares it just doesn't do the trick, especially on FP.

W - S is the thing I always do, even without trees (haven't had that problem yet, but three trees is a possibility).

Sometimes I get as much as 12 tree squares, this means W - W - S and SH within reasonable time at any difficulty level. And all this nice grassland will be cottaged or irrigated after the chopping is done :)

cabert
Oct 11, 2006, 07:45 AM
I didn't see your article and just made the same analysis :(
I do the whipping thing even without mining as starting tech if I'm spiritual, and do the whipping thing even without spiritual if i have mining as starting tech.
It may not be optimal, but valuing the worker turns and not the warrior turns bugs me! I need exploration!

Dr Elmer Jiggle
Oct 11, 2006, 01:52 PM
It may not be optimal, but valuing the worker turns and not the warrior turns bugs me! I need exploration!

That's an excellent point. There's an opportunity cost associated with any delayed build, not just workers.

Earlier exploration leads to better city planning, earlier and better trade relations, better opportunities to scout an opponent's capital before his borders expand, etc.

There's more than just exploration too. If you get your warrior out sooner, that means he can get started earlier on earning experience. Earlier promotions means earlier Woodsman II (if you're lucky) which then feeds back into better exploration. Alternatively it might just mean more opportunities for experience and thus more experience for an extra promotion that he otherwise wouldn't quite reach.

Maybe he gets to an enemy with better timing for a worker steal.

And so on ...

drkodos
Oct 11, 2006, 04:06 PM
That's an excellent point. There's an opportunity cost associated with any delayed build, not just workers.

Earlier exploration leads to better city planning, earlier and better trade relations, better opportunities to scout an opponent's capital before his borders expand, etc.

There's more than just exploration too. If you get your warrior out sooner, that means he can get started earlier on earning experience.....

....Maybe he gets to an enemy with better timing for a worker steal.

And so on ...

Exploration returns are both more speculative (you could go bad directions) and have greater risk (get eaten by a bear). Improving with a worker is a solid, known commodity that can be predetermined and met with sufficiently higher success rate, IMO.



I also reserve the right to do just the opposite of what I seem to preach when it it justified to do so. :p

cabert
Oct 12, 2006, 02:56 AM
right drkodos, it's not a "play it safe" thing to go on exploration. But you can't turtle up the whole game, can you?

I don't say a warrior's turn is worth as much as a worker turn. Far from it, but it's not worth 0.

VoiceOfUnreason
Oct 12, 2006, 04:02 AM
But you can't turtle up the whole game, can you?

OCC, why not?

As always, its a question of leverage. How much value does the marginal exploration you do with another warrior give you at that point in the game? Woo hoo - you sprinted across the continent to meet another civ. So what? You don't have writing yet, so you can't sign open borders, and even if you could there aren't any trade benis until you get the capitals connected. You aren't going to settle way out there very often. So what's the rush?

Note that the value of a second warrior is going to depend in part on how quickly you get a settler out the door. If you beeline Bronze Working to pop the worker, in most cases he has little to do more significant than chopping, so you are going to have the settler out in a hurry, which escalates the importance of finding your second city, and having a garrison ready.

Under those circumstances, it may make sense to apply a bit of micromanagement, swapping to a 1/0/2 tile for one turn during the initial growth phase, so that you have 14 hammers invested in the warrior during the turn the worker is produced, so the warrior is ready on the very next turn.

If you grow the worker, you aren't constrained to beeline BW, so you potentially have more options.

cabert
Oct 12, 2006, 04:27 AM
OCC, why not?

As always, its a question of leverage. How much value does the marginal exploration you do with another warrior give you at that point in the game? Woo hoo - you sprinted across the continent to meet another civ. So what? You don't have writing yet, so you can't sign open borders, and even if you could there aren't any trade benis until you get the capitals connected. You aren't going to settle way out there very often. So what's the rush?

You won't be able to scout after the AIs settled all the land, so it's better to scout early. (disclaimer : i know you can have open borders, but you don't know where tokugawa is ;))

Note that the value of a second warrior is going to depend in part on how quickly you get a settler out the door. If you beeline Bronze Working to pop the worker, in most cases he has little to do more significant than chopping, so you are going to have the settler out in a hurry, which escalates the importance of finding your second city, and having a garrison ready.

right, and having BW and having explored a lot makes it easy to settle next to the copper :)

Under those circumstances, it may make sense to apply a bit of micromanagement, swapping to a 1/0/2 tile for one turn during the initial growth phase, so that you have 14 hammers invested in the warrior during the turn the worker is produced, so the warrior is ready on the very next turn.
? I want exploration to be done before my first settler comes out, then the warriors should come back fogbusting the way for the settler to a good city spot (which often, but not always, is the copper spot).
That's why I value each early warrior turn.
For me each early warrior turn is worth 3 tiles exploration. How often do you see copper? let's say it's 1 out of 100 tiles. So each warrior turn is worth 3% chance finding copper for the first settler. I know there is luck involved, but with 2 warriors you cover twice as much land, so you have twice as much chance to be able to settle next to copper, so you have twice as much chance to axe rush a neighbour. It's still a gamble, but IMHO it's a highly rewarding one with really low costs.

If you grow the worker, you aren't constrained to beeline BW, so you potentially have more options.
True!
If I start research with agri or AH (for those having hunting as starting tech), I usually need a worker first.

VoiceOfUnreason
Oct 12, 2006, 05:30 AM
? I want exploration to be done before my first settler comes out, then the warriors should come back fogbusting the way for the settler to a good city spot (which often, but not always, is the copper spot).


Well, lets see. If you run
7 x 3/0/0
1 x 1/0/2
1 x 1/0/2 + 1/0/2

That gets you a warrior out after 9 turns, at the cost of delaying the worker by a full turn. That gives you 6 or 7 more turns to explore (depending on anarchy) before the worker (now running one turn late) is ready. OK, I can see 6 or 7 warrior turns versus one turn of chopping and 4 food?

I probably wouldn't run a 0/0/3 tile out of the gate to get a warrior finished early, though. Workboat? Sure. Scout? Maybe. Warrior? not my taste.

Let's see - what's the fastest time to produce a settler after the worker gets out? Lets assume you aren't India. So after 8 turns, you have 40 hammers from chops, 32 hammers from the citizen, and 4 hammers of overflow. Four more working turns is another 16, which brings you to 92, and the third chop finishes it.

If you grow to size two, you've got your 8 hammers tucked into some other build, 40 hammers from chop and swap, so you still need the third chop. That's 60 hammers from chopping, 3 turns of working at 5 per, and then a pop rush for 30 hammers, and a turn working at 4 per. More than you need, so again you can trade food for hammers to get to 10 hammers during the growth phase, then spend one more on the warrior exclusively, finishing him in 9 turns without delaying the settler at all.

So what are we at this point: 28 turns into the game, having built two warriors, a worker, a settler, and have a size one capital with no buildings and no improvements? We've probably got a worker tech by now, though. Unless somebody was in a hurry to build Axemen, in which case we have the wheel :smoke:

cabert
Oct 12, 2006, 05:35 AM
i'd go for the wheel very soon!
This way my worker(s) never lack something to do = my way to avoid unemployment:lol:
+ it's needed for pottery and granaries are the best whipping tool (imagine your leader hurling a granary at you when you get lazy :crazyeye:)

VoiceOfUnreason
Oct 12, 2006, 05:51 AM
This way my worker(s) never lack something to do = my way to avoid unemployment

Presumably they can keep mowing lawns until some useful activity appears. I normally delay Wheel as long as I can, given my other needs - pottery, or to hook up happies, or whatever.

ps: is anybody else experiencing whip lash switching between these two threads (http://forums.civfanatics.com/showthread.php?p=4633015#post4633015)?

Xin Yu
Oct 12, 2006, 12:19 PM
I think I'll add some contents to the original post. I'm thinking about building a worker/settler with a worker improving/chopping. Number of trees to chop is a factor to consider, so there are 7 scenarios (0, 1, 2 trees for a worker or 0, 1, 2, 3 trees for a settler). Strategies as always are start building at size 1 vs. growing to size 2 then whip.

Need to do some testing.