Thoughts on New Patch Strategies

[CYZ]

I wonder if they're not calculating the cost of attacking over a river? I've seen a handful of suicide attacks after the patch, but now that you mention it they have all been attacks over a river.

I'm pretty sure it is calculated. I've had a damaged warrior on one side of the river and a barbarian spearmen kept moving back and forth, without attacking. Kindof as if he was trying to find a weak spot to attack.

 

[Shylock76]

I'm really enjoying the AI behaviour, i'm currently playing as America with China to my north and Ottomans off to the east somewhere. Suleiman DOWed me early on and I luckily had a spearman scouting out to the sotuh east and seen his 6 warriors on the way the turn before he declared, assuming correctly they were meant for me which luckily gave me a turn or two organise a defence and hold him off. He seemed to withdraw some units when he could see he was beaten and offered peace after this.

He later DOWed me again but when i set up my defesive line he stopped his advance and withdrew his forces gradually, i did hit some with archers, and offered me a chunk of change for peace. China DOWed me later and when i bet him back he posted a couple of scouts on his borders assumingly for line of sight.

I'm a 'peacemonger' by heart but i'm enjoying the AI, making the game interesting. Suleiman was declared for a third time and so far not sent forces hopefully I dont see a huge number of units emerge on my eastern border, i think its time to take a city or two off him and teach him a lesson.

 

[Merri]

I just was disappointed by AI. I DOWed Alexander (as he was wondering if I'm about to attack), without noticing he had had a CS ally south of me. I pretty much had most of my units on offensive against him, so I had very minimal defense. The first turn of war, before my moves, he moved an Archer right onto my 6 Iron mine (and I only had 8 total). I was cursing that turn, because I only could bombard from my city and had nothing else available in three turns or so.

Then he did the silliest thing possible. Instead of pillaging my mine the Archer simply walked away. The AI considered preserving the unit more important than removing the edge of my attack force.

So, AI still fails with the big picture. An essentially free, outdated Archer from a CS should pillage an Iron mine if the opportunity rises. Even if it means an almost certain death of the unit the next turn.

 

[LAnkou]

well, that's really, really situational...
Besides, i don't know if in other cases where you have different iron sources, it is that good to pillage one mine.
personnally, i never pillage mines from AI...

 

[Lokos451]

Cavalry denying the enemy resources, and to a lesser extent tiles themselves, is often what makes them so very good. Either that or placing them behind enemies before the infantry attack for the flanking bonus, or sniping ranged units. It's these uses where their advantages really show, and the changes to road infrastructure and general numbers of units.

Of course, pillaging only means something if you're not rolling over the opponent in the first place, but that's fairly boring anyways. I personally think the AI has improved quite a bit...but it still needs some work. Just the fact that they'll look to attack you for simply being weak forces people to play more like they would against living opponents!

I'm going to go ahead and give the recent patch a general endorsement for forcing quite a bit more of strategic thought into your military, and slight improvement in tactics as well. A step in the right direction for sure.

 

[ras9929]

Then he did the silliest thing possible. Instead of pillaging my mine the Archer simply walked away. The AI considered preserving the unit more important than removing the edge of my attack force.

So, AI still fails with the big picture. An essentially free, outdated Archer from a CS should pillage an Iron mine if the opportunity rises. Even if it means an almost certain death of the unit the next turn.

I go back and forth with your statement here.
I'm guessing you are assuming the AI could see your entire map-space, and thus know this was your only source of iron.

One thing I would caution here is remembering this is a city state. Even if you can assume, like a player, it can see all your other city state relations and know they don't provide iron, and less likely that a non-exploring city state would have mapped out your entire territory (or essentially the whole world, since you could have a stray city / puppet out there).

Another aspect to this might be how the AI values its force strength vs your force strength. If you had a large army, even in a foreign territory, its not unreasonable for a small AI / CS to fear losing a unit when your relative force strength is so much stronger.

I'm assuming a CS counts no relative strength from its ally. That's probably appropriate because of some of the responses I've seen to "I've seemed to have attacked..."

All in all, I'm hoping this is somehow because a CS AI isn't omniscient and I really don't think they should be.

 

[roguepro92]

haven't played in a week due to school but a friend who is avid player is reporting that allies CS are still horrible at war and doesn't help you much?
did the change/new patch log show any change in CS?

 

[Lokos451]

From what I've seen so far, a CS will help you fight about as well as any AI does...but they will not go much further a few spaces from their borders to do so. Nor will they pillage enemy lands intelligently even given opportunity. They can hold a flank, depending on terrain and relative army strength of their opponent. They tend to build defensive structures and have a fairly credible city strength relative to era. If you want a city state to help in an offensive action, I think your best bet is to rely on free units from the militaristic ones.

This is probably a good thing, otherwise you could see players buying several simultaneously right before declaring war, and then winning simply off the tide of units that would effectively generate. Proxy wars from halfway across the map by players going for diplo wins, or gold crazy high difficulty AI's would suck....well coordinated invasions as the meatshield moves in would operate as a win more mechanic.

 

[Merri]

ras9929: sorry if I wasn't clear, but the unit was Alexander's Archer (gifted by the city state). I wouldn't expect a CS to hit my iron mines, but an Iron mine with 6 is quite automatical for AI that is in (or should be in) defensive mode. I guess the problem here was that I had a slightly weaker army than Alexander, thus Alexander didn't want to sign a peace treaty until he suggested it himself and my military advisor stated I was losing the war, when in fact I had just conquered three cities - very much winning in my eyes!

This does relate to another AI "stupidity". I wouldn't have started a war against Alexander in the first place, but he placed a city about two tiles away from a spot where I had planned to have a city. That doesn't sound very bad in itself, but the next closest city of Alexander was over 15 tiles away with a mountain range partially in-between, with a lot of hill and forest tiles, which means there is no way he can pull off a good offensive let alone defend his newly found city. There was plenty of territory for safe cities.


I'm partially playing my current game to see the limitations of current AI. At the moment I haven't been impressed. I have Washington who is still in the Warrior era, who've maybe had his next-to-his-capital Silk improved for 100 turns, game currently in turn 682 (year 1574). His capital is in a good spot yet it is only of size 5. This is the result of raging barbarians & lots of space for war camps, combined with "incompatible" AI personality Wu Zetian and Gandhi are also struggling.

I guess current AI does much better in tight maps where space is an issue. Barbarians are currently an issue for AI players.

Spoiler :

 

[CossackProblem]

I think the AI gets messed up on occasion. Last game Elizabeth went OCC for inexplicable reasons. However, London was huge and she was actually a very relevant power for most of the game. On the other hand, Washington look over the entire other continent by games end. This was not a total swoop, either, he slowly just gobbled up the powers along with Napoleon until they split the territory 50/50. Then the two butted heads and Washington came out on top.

 

[snarzberry]

Also, I've found that late game the AI has still been woeful. Random artillery roaming around like headless chickens entering my lands one at a time and units diving into the water for a turn just for the hell of it when they should be attacking.

On the plus side, I've been pillaged more often and in the early game (which I think it handles better) the AI is cleverer about retreating and focus attacking.

 

[grandad1982]

Haven't played through beyond the early game yet but it seems that the more aggressive barbs and AI have forced me to maintian a large military at the expence of REXing. In fact both games I have started I got DoW very early. Hang on both times it was by Sulieman as well. The little @&%#! Any way first time I had REXed like pre patch and only had 4 warriors for 4 cities and 5th city on the way when I was DoW and then a CoD turned up on my door step. Last game he sent about 6 warriors and I saw them of with 4 warriors of losing 2 (one got caught away from home when the DoW occured). A nice touch is that his diplo status after peace say you have been at war but he holds no grudge.

So as it stands build more units early and keep them handy.

This is on Emporer.

 

[jagdtigerciv]

@Kevin J

Currently, the AI only launches it's units at your front line if it expects to do enough total damage in one turn to kill your unit (aka odds of killing it >50%). All the way to artillery, this can prove as playing in your favor because if you hold your line tight, you can slowly push their melee units backwards and then fire them slowly with range once they're cornered (obviously that's granted you make sure your units are always on their favored terrain so that their strength, in the calculation of odds to kill, is to it's best potential.

Once your opponent picks artillery up, you should avoid getting into their territory for so long as you are not absolutely sure you will survive range:3 mass bombardment. A good way to tell is to send a worker out to scout their artillery position. If you move in such a way that only one of them can hit your unit, it won't be able to kill it and you should be able to move your own artillery accordingly to blow theirs one by one slowly pushing.

It's a much bigger strain and slows domination game pace significantly if you don't win before range:3 units but ultimately it's a serious improvement to the game.

Other than that, they have done changes to tactical AI and operational AI to increase chances of "sneaky" DoWs. The exact 2-3 lines of interest, I believe, were posted previously in this thread but basically whenever your total military strength is lower than that of an opponent AI, there are chances it will DoW and launch an attack on you, regardless of current relationship (although it is very, very not likely to break a declaration of friendship, it will very often break RAs/trades and go from friend status to war and back to friend status if the war didn't involve "too much" city swaps).

Many players have been switching their total dom attempts into diplomatic victory attempts because the changes make it much more challenging to go for total dom as you basically need to pause your wars from as soon as an opponent gets artillery...all the way up until you get BOTH artillery and infantry (so that your front line doesn't get killed the very turn it moves one hex forward to scout their lands so you can shoot from range:3 as well)

In this particular setup, the relative value of powered-up classical/medieval siege weapons with the +1 range promotion becomes a huge advantage. Enough to the point of allowing you to keep pushing with rifleman front line and range:4 artillery upgraded. Obviously the better way is to always be first on the next important military tech but that's pretty hard to achieve, especially on higher difficulty levels. It thus becomes important to attempt to avoid hurting your diplomatic relations with the civs you plan on attacking later in the game. This way, chances are they won't DoW and tear you up right as soon as they get a slight tech lead(aka not what I did in the game I'm mentionning below).

My current deity game I took 14 cities without any casualties on an ottoman janissary->rifle upgrade rush only to catch up to the Iroquois and England with about 25 turns of research ahead of me (the first had artillery, the later had infantry) Iawatha DoWed me and slowly mowed through my very very promoted front line (each had 3 terrain special, march & blitz atop the janissary baseline promotions) with his 10 artillery..."game over" (I can still win as diplomatic quite easily but the total war is pretty much done, as by the time I recover, tech and units wise, nukes & aircrafts will be out and doing about the same those artillery did to me.

This is a fantastic post. I've noticed very similar things and i'd recommend giving this a read if you're having trouble with higher difficulties and the new AI aggressiveness/strategies.

 

[Louis XXIV]

You are correct, no upgrading in foreign lands anymore. This goes for open border agreements with the AI as well.

I'm not sure what the logic behind the change is - is it an exploit to upgrade in friendly foreign territory?

Yeah, that's an odd one. At least upgrading in City-State territory would be nice. I had one game where having a foreign base was essential for my victory. It was diplo victory, so I didn't want foreign conquest, just liberation and having an army there was essential for this.

 

[Lyoncet]

Yeah, that's an odd one. At least upgrading in City-State territory would be nice. I had one game where having a foreign base was essential for my victory. It was diplo victory, so I didn't want foreign conquest, just liberation and having an army there was essential for this.

I'm going to assume that this was changed to balance out city state beachheads, since it was so laughably easy to slap 500g onto a CS and have your safe haven on a distant continent.. At least this way, you have to make investments (even if it's just capturing a city and then razing it) to get that effect. I can see how that would make your scenario much more difficult, but I think that's OK really.

 

[nokmirt]

Could someone tell me what tech blocking is? Thanks

 

[Eat_My_Brains]

Could someone tell me what tech blocking is? Thanks

http://forums.civfanatics.com/showthread.php?t=407683

 

[nokmirt]

http://forums.civfanatics.com/showthread.php?t=407683

Thanks for the help. In my current game I have RAs turned off just to try it. The next game I'll try using one of the blocking strategies.

 

[Deau]

well, that's really, really situational...
Besides, i don't know if in other cases where you have different iron sources, it is that good to pillage one mine.
personnally, i never pillage mines from AI...

With regards to how the AIs go with zerging your frontline units (as I explained in my previous post in this thread), pillaging a resource can do miracles. Here's the picture, if you manage to push their strategic resource count to -1, every unit using this particular resource will have the combat effectiveness penalty. As such, there are little to no chances they will send their swords/longswords all to one of your units because their expected dmg goes way low in the calculation. On top of that, their defending catapults will do half the dmg, much less threatening your frontline during siege. Then, instead of plowing through their longswords, just let them slowly retreat as you push to ensure they always retain that resource penalty. No casualties war at "even tech" simply by pillaging enough of a strategic resource to bring them to <=-1

*edit* in fact, if you have extra strategic resources, it may be wise to sell them to the AI you intend to go after 5-10 turns before DoWing them. This way, they will build extra units to reach the cap and will turn into negative strategic resource, allowing you to do as mentionned above right off the bat, not even needing to move around to find the first resource to pillage. It is one of my favorite trades to do when I have horse sources on longswords rush

*edit* thanks Jagdtigerciv for the comment, much appreciated

 

[roguepro92]

is it me or are all post patch wars have basically turned into a stalemates?
i can't seem to get a decisive victory but i never really LOSE a war where i HAVE to make peace to continue.