Immortal NA Cyrus.

[Qactus]

Not totally sure myself, but I've seen that myself. Sometimes if I'm stealing from one AI, the others will protect their workers..sometimes. I've seen AIs protect their workers even before any aggression on my part.

My guess is it has something to do with a units coding when built. Some units are coded to defend (you may have noticed a city's defenders not attacking your warrior if next to the a city, even though it is a sure kill). Others are coded to attack, explore or defend workers/settlers

About archers escorting workers:
The same question was raised in G-M 137, which was on Marathon speed so everybody relied on worker stealing. Unfortunately it's such a huge thread and I don't remember who gave the answer, but the answer was: archers and workers move independently, only when a wandering archer "stumbles" upon a worker, it will stay with him to protect him.

 

[MegaLurker]

About archers escorting workers:
The same question was raised in G-M 137, which was on Marathon speed so everybody relied on worker stealing. Unfortunately it's such a huge thread and I don't remember who gave the answer, but the answer was: archers and workers move independently, only when a wandering archer "stumbles" upon a worker, it will stay with him to protect him.

I believe AI troop movements are also affected by proximity of your unit(s) to their cities. There is a specific "city threatened" behaviour, which is why you can have, for example, a warrior next to a city with 2 or 3 archers defending it without being killed. The AI won't move them off the city tile even at 99.9% odds, I think the number of units required before one will attack out of the city might depend on attack courage. The AI also has a problem with units assigned to escort a settler that scatter and leave the settler unprotected when you threaten it. I imagine this has to do with the difference in movement points.

 

[elitetroops]

I believe AI troop movements are also affected by proximity of your unit(s) to their cities. There is a specific "city threatened" behaviour, which is why you can have, for example, a warrior next to a city with 2 or 3 archers defending it without being killed. The AI won't move them off the city tile even at 99.9% odds, I think the number of units required before one will attack out of the city might depend on attack courage.

They will attack warriors if they have 3 archers, sometimes even with 2, as long as they have good enough odds. But what I noticed (on immortal) is that with only 3 archers, they don't attack the turn you declare, so you can always move in to steal a worker and then you have one turn to move to safety before they attack. On the second turn they will attack. It has even happened to me that two out of three archers attack if the first attack fails.

 

[MegaLurker]

Here's what I've got on the second attempt, to 1200BC:

Spoiler :

Went mining->masonry->wheel->mysticism->pottery->hunting->AH->archery

Did so to make the Great Wall a safe bet, finished it in 2400BC. Southern city was settled first, chopped a totem pole, it was able to work the improved corn and stone until the border popped and I could improve and work the gold. Ivory will be connected next turn.

Stole 3 workers. Built 3 Dog Soldiers, two of which are now Woodsman III, the other is Woodsman II. Starting warrior is also at WIII. Just popped my first GG a handful of turns ago.

Just under 400 hammers into the 'Mids.

Very tempting hill spot that gets the clams/ivory/plenty of green to the north. No great defensive spots to the east, really. I guess the corn on it's own to the southeast is decent enough.

Strategy spoilers

Spoiler :

I'm thinking about BiC style rapid expansion at this point, weighing my options.

Pros

1) good land within reach
2) no shortage of workers to steal to get it developed
3) less land for AI expansion


Cons

1) the best land is toward Babylon with it's Bowmen
2) the investment in settlers and defenders would mean I probably can't tighten the choke-hold on the two nearest AIs

I'm leaning towards just settling one or two more spots. With the 'Mids in the equation there's less need to try and grab and then defend the gems to the west. Rep scientists with plenty of food close to the capital is plenty of beakers for, dare I say, elepults? I think I can keep the two nearest AIs entirely without metal and largely without improved tiles at all. It seems like if I take out Peter and Mehmed there might be nobody else to the south or east. I would need to do more scouting to confirm that, too busy stealing workers and such. Even if there's another AI that way, still a strong move I think.

 

[BornInCantaloup]

@ Mega :

Spoiler :

I think your pic makes it clear what the hammer cost of the stone wonders really is.
Your research rate is decent but you'll have a hard time raising it, because of the low cap on your population. I.e. you've evaded maintenance to support your tech pace but couldn't develop your commerce much.
The spy may pay off nicely tech wise.
Your production is low with only 3 cities.

You don't need to settle 10 cities to call it RExing.
Just need to settle 3-4 before 2000 BC. Not all maps allow 4 cities by 2000 BC, so let's say 4 cities by 1800 BC is still RExing.

At any point, you can call your early expansion phaze done and start focusing on growth & infra.
In general, 4 (good) cities by 2000 BC with Pottery + BW would make for an extremely strong position.
It would allow to either tech up or keep settling at a very high pace.

Cities pay off quick.
The thing with early cities is, well, they've got plenty of turns to pay off.


Final point : I'm not sure how important it is to choke Arabia. HRE seems a lot more crucial to me.

And yes, the hill corn to the SW is a very good spot, not requiring any culture to develop.

 

[MegaLurker]

You might still be thinking about this too much like a regular game. Benchmarks like X number of cities by Y date don't necessarily apply like they normally would. Only cities you can hold matter. I don't imagine I'll build more than 5 settlers total in this game.

I'm not sure your flatland settles will survive the catapult phase, let alone knights. It will certainly be a lot more difficult without all the extra super-defenders the Great Wall provides. Around 700BC I am over half-way to my third Great General, the Great Spy freely scouting the map is just the cherry on top of an already delicious cake.

As for which neighbour is more important to choke I'd say: why not both? I'm bringing in probably an average of around 7-8gpt just from pillaging. Our neighbour to the east keeps building cottages. On a good turn I'll clear 20 pillage gold. So far this only requires 4 units total. I used my first Great General to create a super-pillager.

 

[BornInCantaloup]

Yes, if you can choke both, that's great.

On Wonders : I'd think the Great Wall is a whole lot better than the Pyramids.
It's a good deal to trade a settler for the GW.
Is it good to forgo 2 settlers for the Mids ? I really doubt it.


It really isn't the hammer cost of the GW that put me off. It had to do with the tech path. I find that Masonry, here, is one really awkward tech to research.


re : cities :
Not all cities should be targets for invading stacks. Maybe I'm wrong and the AIs like to wander and pillage around... Or maybe they mostly attack the first line of defence.

 

[MegaLurker]

Fair enough. Dead-end tech.

The 'mids are doing nothing for me yet, but it's a long-term payoff. I'm about ready to switch to rep to power through HBR and Construction, then massive police state chops into elepults. If that works I'm all good. If it doesn't then I've just kind of maneuvered myself into a war of attrition I probably can't win. The final sweep of the map is going to require thousands of hammers of units, and I'm hoping that won't take so long that I'll have already unlocked police state the old fashioned way. Having two AIs knocked out early would go a long way toward the possibility of basically turning off the slider after Steel and Rifling.

As far as I can tell large stacks will not pillage at all. They'll either suicide or camp defensive terrain until the conditions are met for their attack courage. Which is good, until the AIs start showing up with stacks at the same time. Then you get a cascading effect of the highest attack courage AI suiciding, giving the lower attack courage AIs the condition they need to attack. If you don't have Drill IV super-defenders by the time catapults show up things go downhill fast.

 

[dutchfire]

AI will usually attack the same city again and again, right? I suppose it makes some sense to try to place cities in such a way that the AIs mostly have different target cities. That prevents all of their stacks to pile up next to one city.

Masonry is a dead end tech (in the short term), but you will want walls anyway, and it allows you to build a stone quarry which is a relatively high yield tile.

 

[MegaLurker]

AI will usually attack the same city again and again, right? I suppose it makes some sense to try to place cities in such a way that the AIs mostly have different target cities. That prevents all of their stacks to pile up next to one city.

In my game each AI has individually targeted one city so far, but that's still resulted in 3 different cities being attacked. And I suspect their target can change when you settle a new, closer city.

AI behaviour is certainly a bit different in AW. For example an AI that can only build archers will attack a city with a stack of archers. In a normal game, unless I'm mistaken, the AI will not plot war if they can only build archers.

 

[elitetroops]

AI behaviour is certainly a bit different in AW. For example an AI that can only build archers will attack a city with a stack of archers. In a normal game, unless I'm mistaken, the AI will not plot war if they can only build archers.

They can't plot, but they can attack your cities with archers if you are at war. It has happened to me sometimes in normal games, when worker stealing goes wrong and I can't get peace, that the AI shows up with a stack of archers.

 

[obsolete]

AI will usually attack the same city again and again, right?

LOL... NO!

If only it were that easy/simple... The AI's can see into your cities due to spy points once all your cities have been spotted.. .they know to target the weaker ones when that happens.

 

[Rusten]

I started playing this and went to look here and I'm really surprised at being the only one who settled 1 north.
The way the river bends means that any attacks coming from west or north will have to cross it and you're also on a hill making it much easier to defend in the early game. I sent the warrior 1 NE to start with to make sure that there were other hills for production there. Very few downsides.
Furthermore there's a visible hill in the initial BFC which means that you will have your second city spot covered no matter what else you find (can share corn).

Image to illustrate:

Spoiler :

 

[BornInCantaloup]

@ Rusten :
I settled 1N on first try as well... but then I played the start two more times and didn't bother
Map knowledge tells both spots are quite comparable, so losing a turn isn't very appealing.

Iirc, I first moved, mostly, to get the hilled capital. I didn't think much about river shenanigans.


Good to see you try your hand at this game, btw.

 

[MegaLurker]

I started playing this and went to look here and I'm really surprised at being the only one who settled 1 north.
The way the river bends means that any attacks coming from west or north will have to cross it and you're also on a hill making it much easier to defend in the early game. I sent the warrior 1 NE to start with to make sure that there were other hills for production there. Very few downsides.
Furthermore there's a visible hill in the initial BFC which means that you will have your second city spot covered no matter what else you find (can share corn).

Image to illustrate:

Spoiler :

The thing is that early defense is really easy, it's free XP farming essentially. The first double-digit metal stacks will not show up until close to the ADs, ditto for any siege. And by that time you would hope to have buffer cities soaking up the stacks. Settling the capital based on defensibility is IMO not important.

I might be wrong, though. I'm not far enough in to know if the AIs will bypass closer cities to continue targeting the capital. On one hand they seem to always target the closest city, on the other hand I don't know for sure that they will change the target if they've already started attacking your capital before you settle a city closer to them.

Also you put stone outside the BFC, which I assume means you don't intend to build the Great Wall? Huge mistake IMO. It's hard to put a value on the extra GGs and the fact that you start accumulating them earlier. Let's not forget that GG-attached units upgrade for free. You can reasonably expect super defenders to last over 100 combats, and potentially the entire game.

Re: stone/masonry for walls. Do walls even matter? By the time you need that extra defense the AI will be bombarding them away, so we can consider them of no defensive value at all. At best they'll buy you an extra 4-5 turns before the stack hits (the AI always seems to bombard thoroughly before attacking once they have siege, correct?), by the time lots of siege is involved they might only buy a single turn. If that ends up being the difference between success or losing the city you have to assume that you didn't prepare well enough.

 

[MegaLurker]

They can't plot, but they can attack your cities with archers if you are at war. It has happened to me sometimes in normal games, when worker stealing goes wrong and I can't get peace, that the AI shows up with a stack of archers.

Right. So I guess what I meant was: an AI in a normal game is no threat until they get >3 strength units (does the AI consider chariots as good enough to start plotting?). Whereas in Always War because the DoW is automatic they will build archers and assign them to attack cities if that's all they have.

 

[Rusten]

The thing is that early defense is really easy, it's free XP farming essentially. The first double-digit metal stacks will not show up until close to the ADs, ditto for any siege. And by that time you would hope to have buffer cities soaking up the stacks. Settling the capital based on defensibility is IMO not important.

I might be wrong, though. I'm not far enough in to know if the AIs will bypass closer cities to continue targeting the capital. On one hand they seem to always target the closest city, on the other hand I don't know for sure that they will change the target if they've already started attacking your capital before you settle a city closer to them.

Also you put stone outside the BFC, which I assume means you don't intend to build the Great Wall? Huge mistake IMO. It's hard to put a value on the extra GGs and the fact that you start accumulating them earlier. Let's not forget that GG-attached units upgrade for free. You can reasonably expect super defenders to last over 100 combats, and potentially the entire game.

Re: stone/masonry for walls. Do walls even matter? By the time you need that extra defense the AI will be bombarding them away, so we can consider them of no defensive value at all. At best they'll buy you an extra 4-5 turns before the stack hits (the AI always seems to bombard thoroughly before attacking once they have siege, correct?), by the time lots of siege is involved they might only buy a single turn. If that ends up being the difference between success or losing the city you have to assume that you didn't prepare well enough.

Forgot that most of these tries were not first attempt. Stone wasn't visible before moving so I didnt know I'd miss it. Even so I think I'd do the same. I'd rather get early dog warriors out to choke neighbours than make the GW. This is a NTT game so you just need to get a good start going and you're set. Free upgrades to longbows-rifle-infantry won't make that much of a difference. I like playing these with tech trade on TBH.

I went straight to BW and whipped a worker into dog warrior which is going to choke Peter (forested hill). Considering putting my second city 1E of the horses as my capital can be protected easily from north and west.

Spoiler :

IMO 4-5 turns extra can mean a lot if you're attacked on multiple fronts. At least personally I'm planning on having mobile forces in addition to static ones. Can use spies to scout for incoming attacks and move them accordingly.

 

[MegaLurker]

You make a good point there, I forgot all about that even though I mentioned it just the other day. Different AI stacks camp out, highest attack courage stack shows up, suicides, cascading odds until even the lowest attack courage AI will go for it. Game over in one turn.

 

[Rusten]

So far so good. (225 BC)

Spoiler :

I did end up settling 1E of the horses. This city would take care of any HRE stacks (if there were any) while my capital would take care of any western surprises as it's the obvious target.

Spoiler :

I couldn't choke Peter completely, but I have been able to pillage tiles setting him back including this copper tile so I only had to deal with archers.

Spoiler :

Stacks are finally starting to show up from the west, but they are no match for my horse archers and soon I will get longbows. My HAs already killed 4/5 units in the stack next to my capital and I have a GG medic ready to heal them up immediately.

Spoiler :

The horse archers mobility comes into play as I gobble up Arabian territory.

Spoiler :

My strategy for this game was to be pro-active rather than hole up in cities with "super-defenders" and wait for AIs to suicide. Having a number of HAs means that you don't suffer from pillaging as you can attack the enemy in the field, and they are capable of roaming around. Not suffering from pillaging allows making cottages, which was my favoured improvement here given the settings. Cottages also increase in value when you're forced to make fewer cities without realying on overlapping.

I went on the attack early and choked Peter to remove his metal and as a result he never posed a threat. It turned out Arabia didn't have metal so I decided to attack them with the HAs instead of Peter.

Monarchy is in and with Arabia's cities I will have more health resources so I can finally grow my cities into the double-digits.

Obviously this game is far from over, but so far it's looking pretty good. Definitely winnable.

 

[BornInCantaloup]

@ Rusten :

Spoiler :

Well played, this is a really nice showcase !

Initially, I'm surprised at how good you did considering the mere 3 cities.
I suppose that, settling that few, helped a lot with the commerce/maintenance on the way to HBR. Also it must have helped with the early military, chokes, etc.
It's very interesting to me, too, how you settled city 3 that far south. I guess Peter being dealt with makes this a strong play.

HBR looks like a very nice military tech considering the map has : bowmen, skirmishers and 1/2 no-metal AIs.
HAs were mostly chopped, right ? The map allows it.


You mentioned it a couple times already but it would be hard to over-emphasize the importance of :
a) Early proactive play ;
b) A mobile task force.

And it's not just that simple as a) + b)... These maps require most excellent unit control.
When we played IMM AW, a few years back, Snaaty came in and broke the maps open. He, too, mentioned and made great use of harrass units as well as mobile, advanced defence (jungled woody 3 axemen, in those cases).
I seem to always play too much of a builder & reactive game for these maps.


So, thanks for backing up strong reminders with an actual shadow game