A simple build-order test...

[Seraldin]

I’m bored at the moment, and can’t play the game, so…

I’m still trying to figure out my early build strategy. I’ve read posts saying that building a settler or worker earlier than city size 3 (or possibly 2) is a bad idea. I’ve read posts saying that it’s a good idea (early expansion) and that it evens out in the end. It looks to me like the general community consensus is that late settler building is generally better. I tend to agree with that from my limited experience, but I’m about as clueless as they come!

So, I’m going to run a little test.

Now, I know that this test won’t be conclusive… it’s certainly a bit contrived, to say the least. Things in Civ games are never “equal” and building strategies (and all other strategies) really need to be rather “on the fly” to take the variable conditions of the game into consideration. Again, I’m a very inexperienced player, but that’s what I’ve seen so far.

The goal of this test is to find out, all other things being as equal as possible, the effects of early build order on civilations.

What I’d like from you guys (if you’d like to help):

1) Has anyone already done this kind of controlled test? If so, care to share the results or point me to the thread? I’m sure almost everyone here has an opinion formed based on experience, but I’m trying to remove anecdotal evidence from the equation.

2) Do you see any obvious flaws in the plan? I’m a total Civ newb, so I may be missing something completely obvious.

3) What leader should I use for this test? Someone that has no obvious production / commerce / cultural benefits (although I think cultural benefit may be okay for the test)?

4) What would be a good order of technology Research for this test?

5) Is this really worth doing? I’ll probably do it anyway, but I won’t bother posting the results if no one seems to think it’ll be interesting to see…

THE TEST:

Pick a leader than has no obvious production or commerce benefits. Of course, the test will probably be just as valid regardless of the leader since all of the test cities will get the same benefit. (Thoughts?)

Manually create a world that has exactly 1 Commerce, 1 Food, and 1 Hammer for each and every tile. All plains, no rivers/hills/forests/etc. If it isn’t possibly to do with no rivers/hills/etc., then make sure that each city has the same benefits. Again, not realistic at all, but we’re going for control here.

Include space in the map an initial Settlers and 1 new city.

Build one of: Settler (S), Worker (W), and Warrior (M).

Run the test 6 times, each time with a different build order.

Cities names will be based on the order of production to help with processing the results. For example, SWM will indicate Settler, then Worker, then Warrior. The six build order combinations: SWM, SMW, WSM, WMS, MWS, MSW.

After a city completes building, build a Barracks, followed by Granary if there is time, followed by Warrior production from that point on.

When a Settler is created, immediately move him 5 tiles to the East and found a new city.

When a Worker is created, immediately improve in this order: Farm, Mine, Farm, Mine, etc. (Better ideas?)

When a Warrior is created, simply fortify the city.

Start building and record the results per turn. Continue until all production is completed… and several turns after that. Record the final state of all cities.

Also of note in the results: the order of technology research. I’m going to go with the “basic” early techs to start with… but I’m not 100% sure of the order. (Thoughts here would be greatly appreciated)

. . .

Hopefully I’m going to do this tonight (if my friends bail on me for multiplayer). If not tonight, then Wednesday for sure. I’m just kinda curious what results (if anything) will come out of this.

[Sciolist]

It seems to me that exluding forests matters here. Starting techs vs. workers also matters - no point building a worker if he can't do anything.

The worker then settler on pop size 1 strategy basically relies on being able to chop a couple of forests down. It also relies on being able to get bronzeworking before the worker spends any idle turns.

It seems to me that it's going to be hard to decide which order worked best. What criterea do you use? I would think that it comes down to what your overall strategy is that decides which is the best order.

[Jcg316]

The worker then settler on pop size 1 strategy basically relies on being able to chop a couple of forests down. It also relies on being able to get bronzeworking before the worker spends any idle turns.

A worker to start is worth it if you have, or are getting, the tech for a resource near you. Worker/settler, Settler/worker and chops are different. If you want to expand certain criteria can mean eaither worker and settler order will work best for an early second city. Starting with mining and going to bronze working for early chops is another seperate way to get far more production towards whatever you have follow the worker in production. Military, settler, barracks, or anything else.

Growing before building any stagnant growth unit (worker or settler) is not as effective as the total yield does not increase by enough per population growth. A worker will make up for the yield difference and then some, while if you want a second city very early you can use a settler first.

[Seraldin]

Hmmm...

Ah, the Forest chopping makes sense to me now... it just hadn't gelled in my mind. Like I said, I'm a newb and still learning (I thank everyone on this forum, it's invaluable to me!).

I think I'll do the test as described above as a baseline. What I'd like to get out of it is a simple "here's how the cities compare at a later point in time with no real variance or complications".

Later, I may do the same test with a forest and appropriate techs.

As you guys said, starting with Worker will be mostly pointless since I won't have any way to improve anything at first... but I'll still do it as a point of comparison.

Thanks for the input.

[LzPrst]

the test is ultimately flawed since the game is about responding the best possible way to unknown start variables. by that I mean what your map looks like. it might be better to produce worker settler warrior in a forestrich area but if theres flood plain a settler might as well be better. adapt to the variables, dont create a set strategy that might not work in a different environment. geez, dont you read history? adaptation = PWFN!

[khumak]

Another variable that most people don't even consider is whether you're starting on a coast with fish/clams/crabs. In that case you can build a workboat or two to increase growth/production without interrupting growth. It's purely food/commerce but food is just as fast as shields for pumping workers and settlers and the extra commerce speeds up your research considerably.

[ducki]

I think the best place to get info about the settler first vs worker first vs others is Vel's Strat thread at the top of the forum. Yes, it's a loooooong thread, but folks have crunched numbers and run both mental experiments and in-game (anecdotal, typically) tests, so that's the best starting point.

But like others are mentioning, because of all the possibilities, you'll be hard-pressed to come up with a single, always-best build order strategy.
Terrain, difficulty level, barb level, map type, etc., all contribute to the wonderful complexity and interesting decisions. Min-maxxing will never go away, but I think C4 does a good job of putting it in its place.

[xxFlukexx]

Yeah Vel's thread is a good read although long.

The gist of the post is that there are 3 pretty good starts:
1-If your land sucks(no resources or floodplains or trees) build a settler first. It's risky with barbs and animals walking around, but you can grab territory much faster.
2-Worker first then settler, do this if you have resources and are getting the techs to improve the land or if you wish to chop, grab bronze working right away then chop like mad.
3-Grow your city to 3-4 then make settlers. This is the best tech strat as you get more pop, therefore more science. This is the best for teching early game and grabbing a lot of religions. This is worse than the above 2 starts in terms of fast growth.

Making a test world that each square has only one food, one hammer and one commerce is pretty silly, since no game will ever be like that. Vel and others already did good tests if you read that thread deep enough(like 3-5 pages I think)

[ducki]

Making a test world that each square has only one food, one hammer and one commerce is pretty silly, since no game will ever be like that.

I tend to agree. I have never had a start where I didn't have at least some bonus available, usually 2 after the first border bump, whether it's food, hammers, happiness, or health. There's always something, because of the mapscripts. Add to that the fact that your civ's traits and starting techs also can generally be exploited/utilized to your advantage in the early game, and I don't think there is a single "optimal" start strategy.

When do you make the cutoff for determining which start is better? 2000BC? 1AD? Endgame? Anything but endgame is arbitrary, IMO, and even that isn't necessarily a good way to compare starts, because everything you do can have an effect on the rest of the game. It's like chaos theory, almost. If you really push on religions and shrines and spreading your state religion, you could have a really stellar bloc of allies making it far easier to trade for techs/resources, avoid/win wars, etc. If you go for the Locust start, you'll probably cow the other AIs for a time due to sheer land area, but eventually they may catch up in other areas and not be so concilliatory. If you go with a balanced start, your margin of advantage, as well as disadvantage, is likely smaller, but it's all in how you use what you've got - terrain, traits, neighbors, knowledge, troops, etc.

Again, I really don't think there's a single optimal first 50 turns because there's no way to verify it. I really think C4 is balanced enough that, while an admirable pasttime, the search for C3's ICSREX equivalent in C4 will, ultimately, prove futile. Not completely fruitless, as we'll no doubt learn plenty from these experiments, but I think what we'll learn is more and varied tools with which we can gain advantages when used properly.