Before beginning some write-up, which is only there for backing up my theories, I'll keep my tips the most pragmatic.
Firstly, I recommend warrior first and more to reinforce. I'm no expert at all generally speaking in Civ4, but regarding worker stealing, I did that multiple times on marathon in the past, thus I can predict worker movements as well for settlers (but that's another story). Yes, this is Epic speed, so worker stealing is a tad more difficult and prudence is required.
Normally, a worker takes 24 turns to train, making a long wait without growth, exploration and impact on the world. If we opt for warrior multiple train instead, we are free to stunt expansion of others ,easing future conquers. In addition, one discovers earlier the risk of attacking such civilization before another one with the intel of strategic resources. Nonetheless, choking must be done intelligently. Choking within BFC without being necessary gives the AI a reason to produce more garrison.
Here are the tips:
Begin with warriors and steal workers as much as possible. Do not begin with a worker, it is not necessary, we are all playing a HoF game, not a XOTM (or Succession game).
On immortal, extorting a cease fire is given in most cases. Abuse cease fire because the AI is enough stupid to work again the same tile multiple times. Not much of a tip because if people are talking lots of minus diplo points, they're hinting me they are already abusing of cease fire.
It is possible to see after the second or the third stolen worker, the AI beginning a settler, so be careful. Blue circles are a indication of where they will settle, if ever you have settler.
Don't fear to fortify on the tile they will settle on because the AI just avoid enemy tiles and go back. Then, just fortify on settling tile if they're approaching it and move to and fro to an adjacent tile to freeze the AI into a sisyphean circle. On the other hand, if you want that badly that city location, just let the AI settle that city at the opportune moment.
Don't forget to max the numbers of opponents.
Furthermore, choose wisely your foes. I really wonder which leaders you have chosen,
Agey. Perhaps, some are completely wrong and impedes your path to victory. My choice is sometimes impertinent too. Wang Kon was chosen because he produces 20 % gold and Churchill because he has a good WTYABTA of fifteen techs. De Gaulle because he's just a bad early warmonger...and I always put him as a baddie. Worker stealing him is just too easy with him.
On a side note, of course, woodman II warrior can be use, but that requires amount of luck and wasting hammers because of bad luck can be irking. If ever that occurs, cease fire becomes useless and abuse of forest patterns in a way you're capturing workers from two tiles away. Settler capture are viable tactic with woodsman II warriors.
If ever some tips are crying for further detail, I fain would explain more. I'm not like
Obsolete with his deity-info-leaking-restrictive-elitist habits.
Flame me!, I won't respond.
Still, warfare is depending much on luck (BFC resources disposition for worker stealing, BFC quality of captured cities, generous RNG outcomes bringing soldiers to survive longer and more experienced, etc.)
An early rush that went well ... doesn't mean the player is extraordinary.
Just to bring back
Misotu words.
[...]does it make it perfect? Or laudable? Or amazing?
Even I was quite surprise of the success of the rush.
-------------------------------------------------------------------------------------------------------------------------------
Very succinct Write-up.
3900 BC (Turn 4):
Of course, I could've been luckier been on a PH and working a forested PH, but four gold tiles plus a riverside corn isn't often seen. I moved my initial warrior southward along the coast to avoid jungle as it can be uninhabited.
3700 BC (Turn 12):
I waited Isabella for working that gold tile. Unfortunately, as usual, she ignored the BW tech path later.
3600 BC (Turn 16):
The last turn, I stole Gandhi's worker. I am moving my warrior onto that rivery tile to be protected from archer retaliations. Then I can move the worker onto that flood plain tile.
3500 BC (Turn 20):
Wang Kon is working on a gem tile. I stole the worker. I stole many workers from him...repetitively. I got lucky by killing a lion at ~33% (explaining the early woodmans II promo)...but later that warrior got killed at ~33 by another lion...without using the said warrior effectively.
3400 BC (Turn 24):
Churchill is another one who provided me a bunch of workers.
3000 BC (Turn 40):
My stats say I have five workers without training a single one. I build a barrack before training vultures. Copper hooked on that turn.
2600 BC (Turn 56):
My first city capture with four vultures and a warrior along for moping. The most risky move, but it was the nearest city. I have nine workers by the time.
2425 BC (Turn 63):
A good road network can do marvels. That is why I march fast towards one civ after one. That is why worker stealing like beast is powerful at the expense of some diplo points in some cases. See the quantity of high commerce tiles. That is one powerful gift from RNG Gods. It lessens the arrival of new cities. I decided to go after Gandhi because he was more dangerous than Isabella without Mining yet. To be honest, I captured her capital much later. I was worried that the others might expand too much.
Later, I DoWed De Gaulle at the east of Delhi because two workers were working on the copper tile. I marched towards the east to discover it was his only copper tile. His capital was nice with another commerce tile (gem).
Another army was marching towards Wang Kon's capital. I let him settle his second city for more workers. That said army, then, was marching towards Churchill.
Meanwhile, the other stack wiping out De Gaulle was returning towards Delhi for the final smash into Kmer territory with the fusion of both stacks (the other stack has finished to grab main Churchill's cities).
Here's a summary of my marches. A bit a mess I know.
Hope your next early rush will turn out to be a perfect one,
Agey!
Don't forget, a good worker steal can impede an AI so much its conquer later is a piece of cake!