When Should You get Your First Worker?

[MarigoldRan]

So in a rich strategy game, with a lot of variety to offer, you've got one script that works, sometimes, if you play one of a small handful of leaders, on a map that naturally caters to that strategy. Often on marathon (which is broken). And you think this is a demonstration of skill.

You are some newb. Suck it up.


Two bits of news

(1) The AI that you are playing against SUCKS at war. Even more so than it sucks at the other aspects of the game.

(2) The AI is deliberately gimped in the opening, such that it cannot do unto you what you are doing unto it.


Seriously, you've been playing since Civ II, and this is the choice of play you prefer? How are you not going insane from the tedium of it?

This is in fact the choice of play that I PREFER. I'm sure that you have a certain way of playing the game that YOU PREFER TOO. Notice how I'm NOT criticizing YOUR SPECIFIC CHOICE OF PLAY BECAUSE THAT WOULD BE SOMEWHAT RUDE.

Otherwise, I think you missed my point. I'm quite aware that the AI sucks at war. That's why you go to war with them early, often, and fast. Hence, warrior first.

 

[MarigoldRan]

They're being facetious.

I'm sorry. I guess my sarcasm meter is broken.

 

[MarigoldRan]

Speaking for me, i didn't, and if you felt like i did so i'd like to apologize. Even the best players go wrong about something sometimes, and i strongly disbelieve that, for instance, TMIT would think someone calls him a newbie just because they try to answer the question as understandable as possible. Just remember, there are people here (including myself) that are reading this forum and lots of it's threads to improve their play, and they're not necessarily IMM/Deity players. Your question is a good question after all, and as far as i can see all posters tried to contribute to the discussion in a way that alot readers can get something out of the thread.



This won't always work If you actually HAVE a strategic resource in your vicinity, opening up with warrior first and choking a neighbour can be very good. If you do NOT have a strategic resource, you may end up choking with crappy units for a very long time. Also remember that choking one neighbour will enable the other AIs to settle his land, improving the chances of a runaway AI with 12+ cities in the BCs.

There are more things to consider.

a) What will you do when your target is very far away? It's pretty likely your worker will either take ages to move back to your land, or that it will be eaten up by barb animals
b) What will you do on mapscripts where you might end up isolated? Warrior first is a sure waste here, and you don't know if you're isolated @ continents or fractal.
c) How do you want to effecivly choke a Deity AI that starts off with 2 cities, which increases the chances that it has a strategic resource right away? Some of the AIs even settle ON TOP of the resource, in which case you're totally screwed.
d) you declare on an AI in order to choke, then you realize that the AIs BFC has little forrests. You're pretty much screwed. You can avoid this with proper scouting before you start your attack, but this isn't always possible. Sometimes the AI also guides his workers early on, what will you do in this case?

There are alot more things to consider. To not go for worker first isn't necessarily bad, but you will need to know what you're doing, and even then bad luck may strike you. With worker first, you're alot saver in this regard



Yeah right, just "keep the economy afloat". That's the problem: it's not guranteed you actually get your worker, get it back to your land and maintain the choke until you're able to kill your target off. You might end up without workers for a long time, and the other AIs will still be in the game and far ahead if something goes wrong.

As i already mentioned, that is not viable for any mapscript anyway: You will want to elaborate strategies that will make you win most of the time, not only some time and on one particular mapsetting. If you're fine with starting new maps over and over again until your strategy works out i'm okay with that, but that kind of play is pretty situational, and a classical has nothing to do with "taking away the killer instinct" - it's just the best way to win most of the times.

You make legit points.

In order to broaden my game, I've been playing the Nobles, Monarch, and Emperor challenges that are occasionally posted on the forums. However, what I've noticed is that with most of the maps, going warrior first, stealing a worker, and then rushing is a really easy way to win the game by a mile.

Take the Carthaginian: Emperor one. Try worker stealing off some of the neighboring civs. It works VERY well. Or the Monarch: Alexander. Or Emperor: Napoleon. Or Immortal: Ragnar Earth 18.

Yeah, I prefer certain map scripts. It's a weakness in all-around play. But my point is that on most maps, warrior first is more optimal than worker first (not Archipelago, for example. But if the setting is Archi, then you know you're on Archi and so obviously you don't go warrior first).

There are exceptions, like isolation starts. But these are rare.

Going warrior first is a gamble. But I think the odds are worth it, since most of the time you will be on a continent with other civs. Also, if you don't find anyone early, THEN you build a worker. Yes, it slows you down a little, but that's the risk of gambles. There's always a good chance you find someone early, and the gamble pays off, at which point you're already close to victory because you've canceled out the comps' starting advantage.

 

[MarigoldRan]

It sounds to me like you're (TC) advocating a Warrior-first opening for a Worker steal and choke, which I believe is a fairly accepted strategy.

That does not mean that it's best in all cases. In isolation or when the nearest Civ is too far away are two cases where getting your own Worker up without worrying about the other Civs is optimal.

I think the main advantage of Worker first is that you know it can help you from Turn 0. You won't know until after you've already gambled with your Warriors if you're in isolation, and by then you're many turns behind in development during the crucial early game.

And I think you over-estimate the power of early scouting. I think most vets here would agree that the initial unit is plenty to scout around the immediate area of the capital and find some good second and third city sites.

I'd be interested to see you (or anyone else) do a comparison of that game you mention with a Worker-first opening. I'd also be interested in seeing a comparison in a LHC-style game, and see how far behind the 3xWarrior opening will leave you with no Workers to steal.

EDIT: The killer instinct you speak of is also slightly overstated. Since you won't be Warrior-rushing any Archers when you get to Monarch, all you're "killing" is their economy (not their cities) and making an enemy, which can definitely hurt if they're still around and their friends decide they're not happy with your diplo plan.

My point though is that building a warrior will help on turn 1 too. You're going to need warriors anyways to garrison cities and to fogbust, so you may as well get it out of the way. EDIT: Also, there's always the chance that your scouting warrior might die. It's good to have back-up.

Especially if you're prone to building settlers, those early warriors really help.

My argument is an attempt to change the mindset of: get a worker out early. There are enough exceptions to the worker-early rule that I think it's a false rule.

Yes, if you don't manage to steal a worker you're in a little bit of hot water. But if you do, then you've almost won the game. That's why it's a gamble. And that's why I think the gamble is worth it.

Think of all the things that you could do if you didn't have to build workers or get worker techs early.

 

[ahcos]

Nope, it's just the other way round: there are so many exceptions to warrior first that it's a false route in general.

But feel free to start a game series to show off your arguments on fractal/random leader and your favourite difficulty (although i don't believe there's much use for anything below Emperor for what you want to show off). It's not that we are the ones that are limiting ourselves regarding strategy, i'm pretty sure that most of the Imm/Deity players here know how take advantage of choking etc. and have done that before, YOU are the one who limits himself to a warrior opening for no good reason.

I'm with you if you say that warrior first + choking can be a viable opening, but as a opening in general? No, Sir. You limit yourself to one single path that may go horribly wrong: yes, it's a gamble, but Civ4 is also about exploring strategies that make gambling less of an issue. Otherwise i'd always go for oracle -> Feudalism on Deity. You want to be consistently winning.

Another (pretty common) scenario you might run into: warrior first, you go out to search for a target, and the only available target is called Mansa Musa, Hammurabi, Sitting Bull or Pacal, all with resourceless UUs that will eat your choking warriors alive. You might run into a Civ that founded and early religion and is creative, making it virtually impossible to steal workers. You run into Ghengis, declare, steal a worker - and realize that you're the worst enemy of not only Ghengis, but also Shaka who, thanks to warmongerrespect, was already pleased with him, setting you to -3 diplo right off the start. Best of luck if Shaka decides that you're worth a BC dagger.

Worker first lets you get an overview before you lay your plan down, whereas with warrior -> choking you're set to this single plan with everything that might go wrong here.

On a sidenote, you're free to change what you want to build in your city when you run into someone who's a good target to choke on turn 3 (which means that the AI is very, very close). Switching over from worker to warriors is still useful by then.

 

[Dhoomstriker]

That's silly. You don't want to scout with an early worker. A warrior stands a chance of staying alive. A good chance too, on forests and hills. A worker simply dies.

Can't believe you'd recommend something like that. I thought better of you.

But, is it really so different from what you are suggesting?

"Build a couple of Warriors first. Expect a couple of them to die. Steal a Worker. Bring the Worker home."

I'm saving you the trouble of building and losing the Warriors, while you can still get in some exploration. You can also bring the Worker back home sooner than you could bring home most captured Workers.

I admit that my message #29 was me just being silly (although it would be abusing the fact that you don't have to worry about Barbs until you settle, so it wasn't completely in jest), but I don't see what's wrong with Worker-first, temporarily ignoring Worker techs (say, by, as you suggested, beelining Alphabet), and thus still making use of the Worker for early exploration.

If you play on Deity, where Barb Animals only spawn for 1 turn (between Turns 5 and 6, if I recall correctly), your Worker should be safe because all of the two-movement units should have been erased by the time that the Worker exists--that is, it will be safe if you are careful about how you scout with it (1 turn of movement per turn on flatlands unless you know that nothing is there thanks to your initial Warrior or to your Cultural Borders revealing the nearby area).

In contrast, a scouting Warrior may have to end up fighting fights that it may lose--even when defending from a Grassland Hills square, a Warrior can lose. But, if you don't have to fight at all, then you have a decent scouting unit, while also having the "equivalent" of an early Worker steal, but without having an AI breathing down your back and without the Diplomatic penalty of having declared war.

 

[TheMeInTeam]

Settler first is also a very (very) rare possibility with IMP leaders, but that's mainly for people who like to gamble.

There's no gamble if you know what you're doing.

@ Marigold:

Even my sorry micro will destroy a start that puts off workers to build 3 warriors (excepting below monarch, where warrior x5 = settler on an awesome city spot).

VoiceOfUnreason showed you straight up why warrior first sucks with most starts.

Also, much as you like bronze working, chopping constantly can barely keep up with...

...

Working a pastured cow.

Improve your tiles.

 

[ahcos]

There's no gamble if you know what you're doing

Yeah right, and then Bear Cavalery (c) (tm) shows up and eats your guarding warrior who's somehow stuck @ the coast for whatever stupid reason, and lolowolf or roflopanther decides to feast upon your settlerparty and all of sudden...

Basicly you're right, but one has to be sure as hell that settler 1st is the best choice to not call it a gamble at all, imo, and to be that sure you're either a fantastic player or actually don't know what you're doing. I'd say that for 99% of the players, settler 1st is a gamble.

 

[cbucks]

If you grow to size 4 working unimproved tiles, how long does it take to get a settler out?

 

[PeterWolf]

If you grow to size 4 working unimproved tiles, how long does it take to get a settler out?

I tried doing that in worldbuilder, someone could check it out or do the maths to see if I'm right.

Just take the best possible situation at the beginning of the game:
Settled on a plains hill, for the extra hammer.
3 tiles with 3F each in the BFC.
At least 2 forested plains hills (3P).
There could be some commerce, but that doesn't matter for this case.

Working the 3 3F tiles, the city gets to size 4 on turn 19.
From then, if you work 2 tiles with 3F and 2 tiles with 3P ( let's say you're IMP for the 50% bonus ) you'll finish the Settler in 9 turns, on turn 28.

Of course that's the best possible situation... without such tiles it would take way longer to get to size 4, and without IMP if would take way longer to build the Settler.

 

[Kaosprophet]

This is in fact the choice of play that I PREFER. I'm sure that you have a certain way of playing the game that YOU PREFER TOO. Notice how I'm NOT criticizing YOUR SPECIFIC CHOICE OF PLAY BECAUSE THAT WOULD BE SOMEWHAT RUDE.

Actually, you are.

But that's beside the point. You asked if anyone had done a cost-benefit analysis on worker-first over warrior-first. They have, and it favors worker-first for the general case. That this finding goes against your preferences doesn't change this in any way; nor does citing the existence of 'intangibles' that cannot actually be accounted for in the cost-benefit analysis you asked for.

If your real goal was to promote the worker-steal & choke strategy, you would have done better by skipping straight to the point instead of asking a question that can only touch on the idea.

 

[TheMeInTeam]

Yeah right, and then Bear Cavalery (c) (tm) shows up and eats your guarding warrior who's somehow stuck @ the coast for whatever stupid reason, and lolowolf or roflopanther decides to feast upon your settlerparty and all of sudden...

Basicly you're right, but one has to be sure as hell that settler 1st is the best choice to not call it a gamble at all, imo, and to be that sure you're either a fantastic player or actually don't know what you're doing. I'd say that for 99% of the players, settler 1st is a gamble.

In doing settler 1st, you don't scout normally, you pull your scout into a "scout the settlement area" role. What you are really doing is preventing 2 move animals from sniping your settler by surprise; anything else the settler can avoid via his 2 moves and simply settling before the threat attacks. Settler 1st assumes a site pretty close to your capitol is viable; you won't have much to see in order to know he's safe when you're settling 3-5 tiles away.

The problem with this is the instant research hit, unless you also have a trade network of some kind formed (possible along coast with culture border touching or similarly along rivers).

I can't remember the last time I lost a settler 1st settler to anything, or a settler at all for that matter regardless of opening. That let's play where I lost my city despite defending it with an archer is actually more frequent than settler losses over the past...oh...50 games or so at least.

It's bad to say there is "no risk" I suppose, only because there's always some risk in a game and you're just minimizing it. Theoretically, in starts where settler-first is appropriate it is *less* risky however.

 

[MarigoldRan]

Problem with the math is the intangibles. I just got this game on Emperor as Hannibal: (Huge Continents, Epic speed, random events and huts).

Check out your neighbors

The thing I'm criticizing is the "worker-first-or-workboat-is-the-only-option" mentality.

I think I'm going to post this as another game in the Emperor series.

I think my opening was: warrior, warrior, warrior, warrior, barracks, warrior, warrior, warrior. I'm going to try standard next time and see how it works. This one is a good test case.
EDIT: Moved to other thread.

 

[TheMeInTeam]

The thing I'm criticizing is the "worker-first-or-workboat-is-the-only-option" mentality.

The best option is rarely the only one.

If you can pull a worker steal, great. However, that carries risks of its own...you may never manage to steal a worker (some AI arbitrarily cover their worker with a unit...or you simply don't find one in time) and you ruin diplo.

 

[ahcos]

In doing settler 1st, you don't scout normally, you pull your scout into a "scout the settlement area" role. What you are really doing is preventing 2 move animals from sniping your settler by surprise; anything else the settler can avoid via his 2 moves and simply settling before the threat attacks. Settler 1st assumes a site pretty close to your capitol is viable; you won't have much to see in order to know he's safe when you're settling 3-5 tiles away.

The problem with this is the instant research hit, unless you also have a trade network of some kind formed (possible along coast with culture border touching or similarly along rivers).

I can't remember the last time I lost a settler 1st settler to anything, or a settler at all for that matter regardless of opening. That let's play where I lost my city despite defending it with an archer is actually more frequent than settler losses over the past...oh...50 games or so at least.

It's bad to say there is "no risk" I suppose, only because there's always some risk in a game and you're just minimizing it. Theoretically, in starts where settler-first is appropriate it is *less* risky however.

That first paragraph was me trying to be funny. So yeah, you're right, but i didn't question that, it was just my lack of humor.

Regarding the "risk" - the last time i've done a settler-first was @ emperor difficulty, i've never felt confident enough in my ability to rate the risks accurately on higher difficulty. You say it's less risky - how come? When i think about it, the risks seem quite high to me as it sets you back on your first worker, and the point where the fast 2nd city is a net plus might be after the point where barbs show up @ Deity ... Haven't done the math on this, but it "feels" like it. IMP leader starting on PH with 3H forrested PH in BFC = 15 turns settler @ size 1, right? Another 12 turns for the first worker, that's turn 27 - turn 40 barbs will come to steal your candy. Archery seems mandatory in order to survive, roading into the strat. resource, mining/pasturing it, and then starting to build the actual unit sets you close to turn 40. So, Archery regardless? Then it might be useful, but you're forced to tech to Archery with a lower techrate overall (as you've mentioned) etc.pp ...

 

[TheMeInTeam]

Tech is only slower if you don't get the trade network up (if you do it's actually a bit faster for a while).

T40 barbs gives you time to make workers in each city 1 and 2 and a few warriors; depending on the map this is often enough even on deity but pretty frequently below.

Less risky when it is optimized is because of the improved production relative to a standard worker 1st if you really won't be able to make a lot of tile improvements with him right away; you have more production and at least as much fog busted/spawn locked thanks to earlier 2nd city.

 

[Habitus]

An example when Settler first might be optimal and less risky would be something like a riverside Plains Hill SIP with Dry Rice + Banana start with a near fully forested BFC, an empty spot thats likely to be either Horse/Copper/Iron as Cyrus@Persia. Your warrior can see Gold+Flood Plains to south on the same river. Atleast that would be a time i would consider it, and most likely use it.

Doing something like Mining-BW-(AH or Pottery depending on if Gold site had AH food), and once my Capital had done Settler-Worker i'd probably even chop out 2-3 warriors quickly for barbs.

 

[ahcos]

Long story short: you have to be in the one-out-of-hundred (well, let's say out of 20) starts to do this. I'll try this one out with regenerating maps until i get such a start, although i usually prefer to do the worker -> settler @ size 1 opening rather than the other way round.

 

[gspchamp999]

I'm with you if you say that warrior first + choking can be a viable opening, but as a opening in general? No, Sir. You limit yourself to one single path that may go horribly wrong: yes, it's a gamble, but Civ4 is also about exploring strategies that make gambling less of an issue. Otherwise i'd always go for oracle -> Feudalism on Deity. You want to be consistently winning.

Another (pretty common) scenario you might run into: warrior first, you go out to search for a target, and the only available target is called Mansa Musa, Hammurabi, Sitting Bull or Pacal, all with resourceless UUs that will eat your choking warriors alive. You might run into a Civ that founded and early religion and is creative, making it virtually impossible to steal workers. You run into Ghengis, declare, steal a worker - and realize that you're the worst enemy of not only Ghengis, but also Shaka who, thanks to warmongerrespect, was already pleased with him, setting you to -3 diplo right off the start. Best of luck if Shaka decides that you're worth a BC dagger.

Worker first lets you get an overview before you lay your plan down, whereas with warrior -> choking you're set to this single plan with everything that might go wrong here.

On a sidenote, you're free to change what you want to build in your city when you run into someone who's a good target to choke on turn 3 (which means that the AI is very, very close). Switching over from worker to warriors is still useful by then.

Well put.

 

[MarigoldRan]

Check out SirCheezy's recent Emperor-Vicky game. This is an excellent example of a game to start without workers. I posted my game and commentary there. Spoilers below.

P.S. Slave-rushing Wonders is a good idea most of the time. If you aren't sure, rush it. That way you can guarantee it.

P.S.S. I've found that most of the Emperor, Monarch, and Prince-series games go wonderfully without early workers.

Spoiler :

Builder order: Workboat-workboat-warrior-warrior-warrior-warrior-warrior-settler-settler-settler-warrior.

BW first, followed by a beeline to Alphabet. Back-traded with Hatty for a lot of early techs.

Completely skipped workers and worker techs because of the large number of good sea spots. Food is NOT an issue on this map. So why work the pig or the wheat plots? Especially since happiness is an issue if your cities get too big. Since you're imperialistic, it's better just to spam settlers and warriors, and take a crapload of cities early, at which point you can do more shiz.

I currently have five cities. All of which was settled before my second worker.... Currently I have three workers (380 AD) and they're running out of things to do.

Currently spamming Wonders. Since I have a lot of cities, I can do that!

I have the GLH, Pyramids, Collossus, and Oracle. The Pyramids is almost as important as the GLH because of Representation (happiness is an issue). Aiming for the Great Library, Mausallos, and the Parthenon. Once that's done, probably get a few more cities, get some defenses, go for Rifling, and then... WAR.

I've got the tech lead. The only way I can not win this is if Toku chooses to attack me. But his worst enemy is Hatty, so....