Thanks for your guide.
As a Monarch/Marathon player, a few comments:
* "This is a tactic primarily reserved for Monarch level and above." True. Winning on Noble and below is easy enough you don't have to pester the AIs at all.
* "It is actually slightly more difficult to steal a Worker at Monarch difficulty than it is to steal one at Emperor or Immortal." Interesting. Though while you detail the Monarch case (the roving Archer) you do not do this for the Emperor/Immortal case - if you're saying the AI doesn't scout with an Archer at those difficulty levels, perhaps you should write that down in the article?
* "I suggest building at least one or two spare Warriors" I've found Warriors to be useless at nearly everything. Obviously, if you start with a Scout you need to build one (or abandon Worker stealing altogether unless you can pop one from a goody hut). I usually make do with the Warrior I've been given at start. The best use of a Warrior imo (except for Worker stealing, for which you usually only need one) is as cheap garrisons for unthreatened cities.
* Knowing Your Enemy. Sound advice. Worker stealing actually is excellent against the civs with good resource-dependent UUs (Romans, most notably): this puts them in a state of war, which slows them down. As you nearly always worker-steal your close neighbours only, slowing down these civs can be crucial (giving you time to prevent them from hooking up their resource)
* A slightly lesser known feature of AI leaders is that they have inherent “personalities.” Again, good advice. Some leaders you shouldn't DoW unless you're prepared for an eternal war for survival!
You could add that the AI's readiness to make peace is dependent on two important factors:
1) does it have any units close to your cities? This lowers the chance for peace. Conversely, if you have a choke unit parked next to their capital, this helps you get peace.
2) who's done the most damage to the other? Pillaging improvements and killing units are great ways to get an otherwise unwilling AI to agree to peace. This is something you should always do, if for no other reason than as insurance against that surprise Archer showing up on your undefended doorstep (if you aren't ahead by then, its presence might dip the AI below the threshold where it will agree to peace).
In step 3 "Planning the Heist" you discuss concerns about your worker-stealing warrior getting attacked back. I've found the AI incredibly hesitant to attack my warrior even when I would clearly have done so myself. Obviously, if the AI has only managed to build a single Archer, it will never leave its capital undefended. But my experience is that the AI must have three and even better four Archers before it's willing to risk an attack.
Obviously stealing the Warrior in the second BFC ring or on a forested square is better, as this brings down the AIs combat odds, which often means no attack at all.
What I don't see in your guide is the incredible use you'll get out of Woodsman II visavi early wars against the AI (wars you'll almost never have unless, of course, you are practicing worker-stealing). This promotion enables you to run in rings around the AI: 1) steal all his workers (until he starts escorting them at all times) 2) never get cornered by AI Archers 3) present such odds on forested squares the AI will almost never attack, and you standing a great chance of winning if it does.
So go out and explore with that first Warrior of yours! Chances are pretty good you'll amass enough experience fighting animals to reach Woodsman I/II.
You also say the Monarch AI starts with a Worker. Are you sure? I have never seen AI Workers appear during the first 20-30 (marathon) turns or so.
* "Warriors make poor choices in this role, due to the difference in movement speeds. An escorting Scout is really the only choice of escort." A Woodsman II Warrior is often an excellent worker escort.
Regarding manual movement: you're right in that manual movement is vastly superior to auto-movement, and I quickly gave up on the automove to carry my stolen Workers home (even with slow Warrior escorts). But this isn't (in my experience) due to a "magical" reduction in Barbarian interference.
Instead the main difference is that when you move manually, you can test the waters so to speak (unless the terrain is all hills and forests with no clear ground anywhere):
Move the worker one step first.
If you see a barb animal, move the Warrior defensively (onto a forested hill, a forest, or a hill in that order; preferably closer to home, but don't be afraid to stand still to let the animal walk past you, or come to you to attack you on a favourable tile) - and move back the Worker.
If you don't see any threats, move the Warrior closer to home, always preferring forested tiles to ones out in the open.
If this uncovers the threat, and you believe the Warrior is at risk of losing (being caught out in the open vs two Lions, or on top of a bare hill vs a Bear) don't follow with the Worker. If the Warrior loses, your Worker is dangerously exposed of course, but you still have a chance of making a run for it.
If the Warrior survives, he's that much closer to Woodsman II, and unless you're right outside of home, it's probably worth waiting for him to heal (yes, the full nine turns in extreme cases) before resuming the homeward trek.
This approach is far superior to automovement (which just blunders on), which is the real reason you'll see many needless deaths when on auto-pilot.
* "More often than not, the AI will attack your Warrior on the IT". On Monarch, I'm very seldom attacked (unless by an UU, but you shouldn't steal workers from these civs). Of course, I'm playing BtS 3.13 Bhruic, and the AI strategy might have changed since vanilla.
* "You want to establish peace for two reasons". It would be useful if you also discussed reasons why you'd want to stay at war. Parking a unit close to enemy capital keeps even a Monarch AI from sending units your way for a looong time. By the time AI manages to build enough Archers to stop chasing you around and actually beeline one or two for your territory you should be ready to kill them (lesson: don't delay Archery yourself unless you get lucky with Horses or Copper!).
"Allowing the AI to recover is key." Well, not allowing the AI to recover is win. I see your point, but in some cases it might be better to keep that particular AI small and weak. Reasons: 1) it hates you, but other neighbours don't. Having strong friends is better than having strong enemies. 2) An AI doesn't need to recover to produce new Workers for you to steal. 3) Staying at war means you can attack its units (and steal more workers) without having to declare war again. The diplomatic penalties are not to be underestimated (and now I'm not talking about the cumulative -3 "you declared war on us" penalty, which won't matter once that AI is dead, but the dangerous -1 "you declared war on a friend").
Often, you encounter your first AI so early it has no friends when you steal its first worker. Making peace should then be avoided as long as possible - the next time you DoW it, you're bound to get several -1 demerits, which can be disastrous in the longer run.
Obviously, to manage to stay at war with an AI for an extended time in the early game, you must ensure the AI is kept in check. What I mean by this is that if you're careless, it goes into what I'll call "extermination mode", which is bad for you (unless you're lucky enough to have an Archer-killing UU available right away, in which case you should probably go straight for the kill instead).
If you settle your second city "too close" to the enemy capital, this can trigger it to send an unending stream of units into your territory. (Often coupled with the unreasonable demand for you to give it that city in peace). Getting the AI to get hung up on a city like that can be incredibly bad, as this seems to increase the bar for you getting a favourable peace by a large amount.
Also (as already mentioned), it's imperative you keep a unit or three threatening the enemy capital. This can complete ruin an AI, which keeps two workers, one settler and eight Archers holed up doing nothing. Imagine if that AI were allowed to "recover" - meaning the next time you steal its Workers, you might face a three-city civ and eight Archers instead!
The fourth reason to stay at war is important enough to discuss separately:
4) to deny the AI a crucial resource. Again, Romans & Iron is the easiest example (you get the idea).
In one of my recent games, my Portuguese started on a peninsula but with an inland capital, facing Churchill and Caesar (and plenty more civs behind them). I managed to find Horses for my second city, but had to keep both civs at bay using nothing but warriors for the ultra-early game. Once I got to Horses, it became easy to deny both civs any resources - I even managed to keep the romans from building an Iron Mine on the tile
next to Rome by threatening that tile with three Chariots.
Why didn't Caesar move out one or two Archers as escorts? A human would easily have protected his mine-and-road building Workers from Chariots, especially as the tile was a hill.
Because I was careful not to steal any roman Workers when I DoWed him! This meant the AI's routine for escorting its Workers were never triggered, which meant it only considered sending out unprotected Workers, which meant it never sent out any Workers at all as my Chariots always threatened that tile. (Caesar did send the occasional Archer to my closest town, but was easy meat for its Garrison Chariot).
Now, I've finally managed to get my own empire up and running, and with Catapults I can finally (two thousand years later) conquer their capitals.
Had I made peace, yes, the Romans would probably not have kept eight Archers for defense, but instead I would have faced half a dozen Praetorians.
In summary, your guide is sound, only I would like you to discuss "Planning the Next Attack" without having to get multiple diplomatic penalties, by remaining at war if possible.
Zapp
PS. What's with the "maluses"? Are you from Germany by any chance, Nares?
Might I suggest you doing a find-and-replace of "malus/mali" for "penalty/penalties" which is much more idiomatic english. Thank you.
.
