Discussion in 'Civ4 - Strategy & Tips' started by CivIVMonger, Jul 11, 2010.
Many people will discover that doing this will cause you to lose that spot in the first place.
I would argue this: it seems to violate the principle of maximizing the number of turns you are working improved tiles.
If that's true, then it probably falls behind the alternative lines assuming otherwise normal play.
It might be equal or superior with additional changes in play - without additional details there's really no evidence to hang an opinion on.
It's hard for me to believe that you can afford the dedicate the capital to settlers, but must distribute the training of workers (none of your cities can train 60 workers at the rate your capital trains 100 settlers?).
A bit easier to believe is that there could be a big timing problem - assuming a start like worker - warrior... - settler*, city two may not develop quickly enough to support its brothers, and therefore cities 3 and 4 have to fend for themselves.
You're also going to get very different results if worker #1's job is to stay at the capital and chop, instead of helping new cities develop.
That's definitely true, but that falls under the "unless if you're rexxing" category. If you can afford to wait, you should. Heck, a lot of times, it's going to be to your benefit to send the settler out and just space them out until the worker gets there (assuming it's not too far behind). Then you avoid the maintenance costs, but you can still plop down the city if someone else shows up.
Ouch, I almost always build worker first in all cities, since I read it was good for the capital. thought it would be good for the rest too. Well has been working fine for me. Will stop doing it tho since it seems all agree it's not a good option.
It does seem less than optimal, but it also depends on the situation. If you don't need to do a monument for the border pop in a new city, and for whatever reason won't have a worker from elsewhere available for quite a while, then it might be worthwhile. That said, I rarely do it.
A prime example of a city that can build a worker first is a gold city . You need to work the gold soon to be able pay for itself/research more. Build setter, build city on gold, mine gold, build worker there on to the next spot hehe
What's the alternative? Granary? If you suggest building a granary and using a fresh capital worker to improve this new city I would say that it would work fine, but you would expand much more slowly.
I know that there are several approaches to this game and many strategies work near equally well according to the map/situation, but I thought expansion was not optional? Getting as much land as possible without screwing yourself economically was the way to go. I used this strategy under the assumption that land is power, and being able to improve that land 15 turns after the city is planted and significantly increasing the rate of which settlers is built, quite phenomenal.
But on the other hand, getting a granary up first and using a capital worker emphasizes development, whereas worker first in all cities emphasizes expansion.
So... Conclusion is, worker first in all cities is okay for a REXing strategy but is not ideal? Of course, as with all things in Civ, it depends on the map/situation and I am only speaking in a general sense.
Clearly settlers often have to go out and claim land before it gets snapped up by somebody else, but in pure efficiency terms it has to be best for workers to be built in cities with no more good tiles to grow onto, and for them to be there ready to improve new cities' tiles as soon as they're founded.
In isolation (i.e. a fairly common situation, where there is little or no competition for sites) this is going to make a big difference to the rate of development of the civ. I'm sure you could demonstrate it easily enough.
Building worker first in new cities (except capital) seems to be a textbook example of "not enough workers" syndrome, which we all agree is bad, right?
Monument if not creative (and if any special tiles are outside first ring), followed by warriors for fogbusting/escorting settlers/exploring, or maybe a workboat if coastal. This is for early cities (#2, #3). Fourth city or later I might start on a granary after the monument, if I have enough units. I think it's important to let the early cities grow to the happy cap ASAP (modulo whipping).
I used to build worker first in cities #2 and #3, but I think my game improved a good bit when I switched to the above.
In all of your cities? definitely not.
Every game is alittle different, depending on the strategy you intend to employ, the leader you choose, and the empire you choose.
If my empire starts with fishing, or I have a lot of sea resourses near my capitol, I pump out work boats and grab those fish/clam/crab to pop up quickly.
If I am playing a good early attack empire (Incan, Nat. Am. or Celt) I build units and go get someone else's workers and save time on having to build them. Piliging farms, mines and hamlets for gold to keep my tech % high. The AI workers don't run from your units that are 2 spaces away through forest or hills, because, the game thinks its worker has time to do more work. A Gallic Warrior with Hills2/Woods2 can move fast enough to capture these workers and protect them to bring home.
If my starting empire has nothing to use a worker for (like starting with hunting and mysticism and researching Hinduism/masonry/Judaism to grab 2 religons quickly) then you are better off building something else, such as the work boat above, or your barracks, then a settler.
If you are playing an Expansive leader or India, then, getting workers out is key to your game. With unrestricive leaders, you can choose Bismark of India. Expansive and has Fast Worker. Is Industrious to build Stonehenge and Great Wall faster. And starts with Mysticism to research Hinduism quickly.
The alternative is to grow.
Buildings do make a nice hammer sink, because the decay rate is slower, but you can probably put hammers into a warrior without risk. Even with a "mere" 5 food surplus, you have 15 turns (normal speed) before you hit size 4.
Please note that the alternative offered here is not "construct a granary, and then", but "grow, and then..."; the buildings come off the queue unfinished once the city has matured....
Although the "construct a granary" approach might be better or worse than either.
Show me the evidence. How quickly are you getting your new cities founded? How much more slowly does the "normal" approach expand? What metric are you using to compare the two?
What I'm inviting you away from is an argument "it's better, is not is too is not..." but an exchange of data. Because my feeling is that the difference between your approach and a more standard approach is smaller than you think - small enough that the other compensations make the overall position stronger.
But I certainly don't claim to know that this is so.
At the risk of repeating myself, the problem with worker first (and, again, this is something we all have to do sometimes) is that for fifteen turns a new city will be nothing more than a drain on the economy. If you have another city somewhere that is already profitable and can pump out workers in 5-6 turns, then it should be feasible to time your new settlements so that they can get online faster with the help of a worker when they are settled (or very soon after).
So, I guess the answer is - worker first in a new city is optimal when it is the fastest possible way to make that city productive and profitable. In most cases there are better options. As always, blocker cities and other critical sites take precedence over all rules of thumb like this. When it is feasible to do so, delaying settling a city until a worker is available is usually going to be much more efficient, economically speaking.
Building a worker first in a city should be useless, because you've already transferred a worker with the settler in the first place.
"But then my capital/other cities don't have a worker!" Not a problem, they can build another one, and do it while working improved tiles, whereas your new city will not. Furthermore, your capital should have around 3-4 improved tiles when your second settler comes out, so there's no need for another worker as long as you aren't wasting time chopping/roading.
For example, let's say my first worker farms a corn, pastures a cow, then mines 2 hills while producing a settler at size 3. He goes with the new settler to the second city, where he pastures a sheep then a horse. By then, your capital has produced another settler, and your worker leaves your second city to join the 3rd settler. All this with one worker.
"But my second city won't be able to work an improved tile at size 3!" Ok, NOW build a worker there.
So in summary, a better rule of thumb: transfer workers to new settlers. Replace them if necessary by building a new worker at the city from which the worker was transferred.
now that is interresting a lot! I will try it for sure.
The thing is my usual routine at cap is worker, grow, worker, settler, worker, worker, settler.
But what you wrote has a lot of sense, I would definitely pay less for workforce ;-)
I want a swarm of workers and building a worker in the new city is much better than taking my capital offline to build workers - the capital can pump out settlers then get a library going once I have claimed the land I want. At least for cities 2-4 anyway. Later cities won't go worker first because I should have spare workers or have secondary cities with good production that can build workers instead of my capital.
But the worker must be chopped out - otherwise too many turns are spent on unimproved tiles. Chopping a second worker for the city and then using that worker to develop tiles while the first worker continues chopping out a monument and granary works pretty well for me. Then once I get the city developed a bit one of the workers can move to the next city under development.
I do it occasionally for say a third or fourth city if my capitol has to pump out some extra military units to deal with extra barbs I wasn't expecting. But I try to avoid it by having the worker built in another city and arriving within a turn or two of the city founding. I just feel like it really slows down my city getting online when it takes 12 or 15 turns working on a worker before it does anything else.
Hmmm... Let's see - wet corn and grass cows.
15 turns to train the worker.
6 turns to improve the corn, 4 turns to improve the cows, 5 turns each to improve the hills.
City is 18/22 when the corn is finished, 23/24 when the cows are finished (T25). That's 10 hammers. You could start cranking out a settler now - corn + cows + city tile is 11 hammers per turn (3 + 4 + 4). If my micro calculations are correct, there's just enough time to train a warrior before the settler (you hit 16/22 with 2 extra hammers, and then 20/24 to get 3 more).
Alternatively, you aim to grow one more turn, to 7/26. Here you have 11+2 hammers without micro. Now you are producing settlers at 12 hammers per turn, and 13 once the mine comes on line. That should give you a settler in 8 turns - same speed as before.
So at T34 you have 2 warriors, a settler, a worker finishing a mine, and a capital that spits out settlers every 8 turns (13*8 = 104).
If you decide "woo-hoo, settler pump!", then after some walking City #2 is founded next to warrior #1, 8 turns later City #3 is founded next to warrior #2, 8 turns after that City #4 is founded ... with no support near by. You have an extra worker at City #2, finally, and a new Settler ready to go, but city defense is starting to look a little thin?
Second observation: a worker is 60 hammers. At 12 hammers / turn, that's 5 turns. At 20 hammers a chop, that's 3 chops. At 4 turns per chop, that's 12 turns.
So if we go with the Settler pump, we get two settlers by turn 16. If we train a second worker to chop, we get a worker and two settlers by turn 17. Obviously, this costs some trees, and a bit of lag in the first Settler
Farm is 5 turns without movement, so you end up the second resource one turn earlier, often 4 hammers. You also get access to the resource the turn it finishes. 6 food is too fast to a warrior before size 3 without a ridiculous hammer resource.
16: 0 (1 turn)
17: 3 (2 turn)
20: 12 (5th turn)
21: 18, is this what you mean, or are you saying
23: 18 (slow farm finishes)
If you're going to compare a size 2 settler, you stop growth immediately on reaching size 2 and start building the settler, and suffer a small hammer decay in your warrior.
In this case, it would be
22: Size 2, 2 food, 8 production (pasture 2)
24: Settler at 16, pasture finishes, 11 production, needs 8 more turns, I guess you get decay with a 5 resource tile.
If you grow to 3, you hit it at turn 26, mine in 3 turns. 12 x 3 + 13 x 5
So you gain 2 turns with the new city, and waste 4-2 non-settler building turns at size 2.
I outlined for a 3 city empire. Also note your second city (and most later cities) is probably growing to a second resource, building a 3rd warrior, and later on, if need be, a chariot/axe. And of course once you hit your 4th city, your base empire-wide maintenance increase is ~2 gpt without distance maintenance, say if it's at distance 10.5 you'll total about 5 gpt. You get one city commerce,so you have to offset 2 non-riverside mines with 2 riverside cottages. In which case your 100 yield settler investment is outputting 5-4=1 to 7-4=3 yield first pop. 5 yield + mine - maintenance loss will be about
Your second city, on the other hand, is going to have 1 less empire-wide maintenance from number of cities alone, will tend to be closer and have 2 special resources.
This thread is really delving into the science of improving land. Thanks! I hope I didn't sound, pushy with my opinions.
My bad, I ought to have been clearer that I was sketching and estimating, rather than allowing the inference that I was trying to draw up an optimal schedule. One can't be precise without a map to measure against.
Are you sure? My memory is that pasture is 4 worker turns and farms are 6. So if the worker is available in the city on turn 15, and the corn is in the outer ring, then 15 is a move, and the corn is farmed before the end of turn 21. Which would be 6 turns working an unimproved tile (15-20). Move to the cow on turn 22, and the cow is pastured before EOT:26.
Yes, you get a different number if you only burn half your movement (or if you have a fast worker).
More to the point, though - I'm not yet trying to define an answer, but rather to define a question, and using the rough sketch to illustrate some concerns that I have, in the hopes of getting some feel for whether the concerns matter in any real sense.
I didn't make that at all clear in my earlier post.
Note: I didn't follow your shorthand at all.
But roughly we're looking at
A turns working an unimproved resource
B turns working corn
C turns working corn + uninproved resource
D turns working corn + cow.
If we're going to be pumping settlers at size 3, we'd like to (a) cross 22 food as quickly as possible, (b) reach 46 food as quickly as possible, (c) convert as much food to hammers as possible without compromising a and b.
A pure food approach might look like
A = 5 (15/22) +3F tile alone
B = 2 ( 5/24) +6F wet corn
C = 1 (12/24) +7F wet corn + unimproved grass cows
D = 2 ( 4/26) +8F wet corn + pastured cows
By my count, that gives you 10 hammers from the city, 4 hammers from the cow, and 4 extra food that you can convert to hammers. If that's right, it should mean 4 turns that you can work a grassland forest instead of a 3F tile to get the hammers you need to finish the warrior, and maybe have some overflow in the capital to speed the settler.
Is that off somewhere?
Separate names with a comma.