Install Steam
login
|
language
简体中文 (Simplified Chinese)
繁體中文 (Traditional Chinese)
日本語 (Japanese)
한국어 (Korean)
ไทย (Thai)
Български (Bulgarian)
Čeština (Czech)
Dansk (Danish)
Deutsch (German)
Español - España (Spanish - Spain)
Español - Latinoamérica (Spanish - Latin America)
Ελληνικά (Greek)
Français (French)
Italiano (Italian)
Bahasa Indonesia (Indonesian)
Magyar (Hungarian)
Nederlands (Dutch)
Norsk (Norwegian)
Polski (Polish)
Português (Portuguese - Portugal)
Português - Brasil (Portuguese - Brazil)
Română (Romanian)
Русский (Russian)
Suomi (Finnish)
Svenska (Swedish)
Türkçe (Turkish)
Tiếng Việt (Vietnamese)
Українська (Ukrainian)
Report a translation problem
Discussions Rules and Guidelines
Report this post
What's not historically accurate, however, is AI allies behaving differently when their fellow ally is AI vs when their ally is human. That's a telltale sign something's wrong — when pure-AI alliances all stack up but human-AI alliances can't co-ordinate.
What's also not historically accurate is allies not helping you when they stand to lose by you losing.
John III Sobieski came and pwnt the heck out of the Turkish army (while having plenty of Muslim soldiers and even Muslim military chaplains in his own army, which few people actually know about) because you can't be safe in Cracow with a Turkish Vienna, for example.
The one thing the ally shouldn't want to do is lose its all armies. Allies don't want to overpay the price of helping you, but they don't want to overpay the price of not helping you either. Allowing you to lose doesn't automatically help them. In fact, it harms them because if you get ousted from power you can no longer help them when they need help. If their supply of sons and daughters is not finite, they should prefer to keep in-laws in power.
My point is not that AI shouldn't take big risks for you. My point is that AI shouldn't be dumb as a brick losing all its army just after you lose all your army, simply because AI doesn't want to or can't co-ordinate with a human player even though multiple AIs can co-ordinate.
The problem is that a total AI alliance will all gang up into a single stack, while an AI-human alliance will spread apart and fail to achieve even the most basic co-ordination.
Clearly, there is a difference depending on whether AI co-ordinates with other AI or with a human player.
damn dude you dragged your nuts across this guys face with accurate history lmao
if you have a strong army crusade is the best part of the game, because that is the only time where strategy in war matters, how you move your army around, and deciding which castle to siege so you can keep your armies supplied,
and most importantly maneuver armies in a way that you are cutting of his reinforcements, fighting and destroying smaller armies so you can gain small edges in the final engagement, and some time due to the large area you fight in you can even get multiple large engagements in a war, it is where a strategist dream come true.
I had a crusade, every single one of the enemy united to join a battle.
Every single AI ally sat in fields watching.
Its some confidence that out of all those hundreds of AI leaders every single friendly had a screw it personalty and all the enemies had a die with honour personality. 4 crusades in a row.
He referenced a commander from the 17th century, far beyond the timeframe of CK3. By then, armies were no longer the personal property of landowners, but professional troops paid, equipped, and trained by a central government. Very different matters.
Nations (especially America) will levy their forces on behalf of an ally all the time. How often, by contrast, do people tell their sons to arm up and help a neighbor whose house is being broken into?
Fans want us to believe personality is the hidden factor behind those bad decisions. Or perhaps they just want to give P-dox the benefit of the doubt, either out of sympathy or just to be fair. But in my experience this belief doesn't explain all of the situations in which AI acts in ways that seem to have 'AI design problem' written all over. It's not difficult to tell a human style from an AI style, and those bad AI moves can't just always be the good roleplaying by AI of a bad commander. Usually it looks like bad AI design.
For example it looks like AI struggles to properly take into account the game's own rules (details of CBs), move distances, paths, co-ordination, reduced combat ability after disembarking, and more. It also has plenty of indecision locks, changing commands back and forth and becoming unusable due to being stuck in a loop.
Maybe it was rushed? Didn't get enough concept manhours to do properly when racing against the development deadlines? It does look like it didn't receive anywhere near the number of manhours it ought to have received in the development cycle.
In any case, I could certainly believe in the occasionally well-roleplayed bad commander but can't possibly chalk all AI blunders down personality roleplaying. AI behaviour looks like a sort of stupidity that doesn't come from roleplaying. More like from genuinely not knowing what to do.
Sometimes they attach but only rarely.
AI will attach to other AI, although generally to a primary attacker/defender. Even if this means coming up with a 20K stack taking attrition in 10K or 5K supply provinces.
But AI won't attach and stack up when it's in a coalition with a human player. It will spread to adjacent provinces and often even adjacent-to-adjacent, resulting in a carpet coverage of the map with 5K splinters. And then comes the enemy with a 20K stack. You may have like 35K total forces among you and the AI allies, but you just won't be able to funnel them into the battle fast enough.
The sole fact that AI-only coalitions stack their armies and AI-human coalitions don't loses the human player many of his wars.
This is less annoying when we're talking about offensive wars for which you made a daring or risky decision. But it's more annoying when we are dealing with a 9K viking invasion vs your 3K army as a duke, whereas even Karling kings have little more than 4K per kingdom.
But in any case, it's always annoying when you lose to death funnels simply because your side won't stack while the enemy side will.
Introducing a couple of simply commands for the human player to give to his AI allies (which they could execute well, execute badly, execute partially or ignore, depending on personality and to some extent RNG) would solve this problem.
I was responding to the Siege of Vienna being pointed out to me as a specific example. The decision to talk about Vienna wasn't made by me. Are you slower today or just not paying attention while reading, or just really that biased?
Although this is outside of the topic at hand here, you have also just shown that you don't actually have a clue how John Sobieski's army was raised, composed, trained, handled and regarded. ;)
In America or a disarmed socialist welfare state in Europe? You can't compare the two. And you can't compare homeowners and mediaeval rulers. You can compare homeowners in 21st century and homeowners in 14th century, sure.
And even if you are Genghis Khan, the majority of your levies are not your descendants.
Next, you are actually making a very stupid argument by insisting on ignoring the point I made about AI not just fleeing the scene but getting its own army slaughtered just after the human players, out of a refusal to co-ordinate or help. This often actually leads to the sons of AI rulers dying, or at least getting captured and ransomed. In ignoring this point repeatedly you are either arguing in bad faith or just being daft. The point of this topic is clearly not about AI being slow to help you. The point is about AI acting stupidly and getting its own army wiped and your army wiped too, because it preferred your respective 6K armies to face an 8K enemy separately instead of ganging up.
So don't play shenanigans twisting what I said into what I didn't. Argue the points I made, don't misrepresent them into something you have a ready counter for but something they are not.
Crusade that are too far in land from the ocean has problem for AI, because most of them will have problem joining fights if they are low or no supply, that is that is why if the crusader have a easy time taking said Jerusalem, but will fail miserably in Mesopotamia unless YOU win it.
if they have to take any amount of walking to the target after getting out of their ships the crusader is pretty much have to be solo carried by you. you have siege castle in a way that the AI can use to replenish their supplies, you have to be fighting and smashing enemies armies because chances are all your allies armies will have the no supply penality -25% and before AI join in a fight they will consider that. so if you want them to join your fight when their are running -25% penality you need to be seriously winning the fight.
I'm really that biased.
Homeowners are the closest thing to medieval rulers we have in a first world 21st century nation. Homeowners are land owners, and that's exactly what a medieval ruler was.
Nations levy manpower that isn't "theirs", short of their manpower also being taxpayers. It's easier for a modern President to honor a treaty with the full force of his military because he personally isn't risking a whole lot, apart from maybe his legacy. Worst case scenario, if the war is a massive disaster, a President steps down from office and passes the buck to a completely unrelated schmuck and goes into the private sector.
Feudal lords levy manpower that they more or less own. Their fighting men are relatives, tenants, and serfs. When they lose fighting men, they also lose income. If their wars are disasters, they lose the means to defend the thing their whole world revolves around - their land - to greedy rivals. In a worst case scenario, they're passing the buck to their carefully raised son, not some random guy who was a political rival four years earlier.
Feudal lords were not in the habit of risking their forces readily for allies, unless they happened to be very close to that ally on a personal level, or the alliance was somehow incredibly crucial to their security or ambitions.
AI's 4K stack is about to be attacked by a 5.5K stack from the enemy. I move my 4.5K there. Enemy AI walks into it. What does allied AI do? Walk away, allowing me to lose. And without my army, it will lose the war.
Some of you guys may elect to… shall we say, make yourself look less than wise or honest by arguing just to argue, debunking just to debunk and defending just to defend, but you're really going to just keep embarrassing yourselves and not succeed.
I'm not going to continue exchanging snarky remarks. I'll just state the obvious by saying the answer doesn't lend your argument much credit (to say the least). It's perfectly clear that you aren't even trying to address the points you're arguing against. You don't even bother to read them, you just move to churning out low-quality excuses for the current state of the game, which even Paradox devs themselves agree needs fixing (as per DDs). That sort of fanboyism is detrimental to the post-release development of any game, any software product, anything really.
The argument isn't about AI failing to win a crusade for Mesopotamia despite having the theoretical numerical advantage — that would be understandable and expectable in the light of the mechanics as we can easily know them.
The argument is about AI electing to walk and sail to Jerusalem from Syria via Iraq, ships, and Egypt. The AI is clearly trying to calculate attrition and time while perhaps trying to come up with a way of encirclng the enemy and failing badly.
The -25% penalty shouldn't be preventing them from joining with their 4K when your 4K attaches an enemy 6K in an adjacent province. Even with their 4K having a penalty, you are still more likely to win ganging up than attacking or defending separately vs a 6K as two 4K stacks instead of an 8K stack.
The AI's decisionmaking here is neither correct, nor a good emulation of human conduct unless randomness, chaos and incompetence is what is supposed to be emulated.
Finally — 'you shouldn't blame it on ai design because ai behaivor in this game is actually some what predictable.'
What you're saying is effectively that you shouldn't blame bad design as long as it is at least somewhat predictable. That's cutting a lot of slack and being unreasonably lenient.
You're confused. A president is not a one-man decision-making centre. Out of all decision-makers involved quite a lot think in terms of sustainability and preserving the nation's manpower, levy base, tax base and above all independence, and preferably its position in the pecking order as compared to other global powers.
A similar argument as you were trying to make could be made against presidents and other republican leader helping allies — just because the leaders are serving relatively short terms, so if they don't even care about what happens to their nation after they leave office (which is not even cynical but a daft claim to make), then all the more so they shouldn't care about alliances, i.e. what happens to allied nations (which are less important to them than their own).
I have to ask again: Are you daft or just pretending?
I've already told you several times, you are just pretending you don't hear: We're talking about situations in which not helping you costs the AI ally its army. Situations in which the AI prefers your 4K and its own 4K to be wiped in succession by a 5–6K stack, instead of joining together, winning, and obviously losing fewer men.
Not only does your argument not hold water, you are actually being intellectually dishonest in that line of argument.
If you're going to ignore parts you don't like or parts that don't match your counterargument or even contradict it, then don't waste anybody's time, including your own, making a counterargument that's not relevant.
There is no merit in attempting dishonest counters that don't remotely work. Being a loyal fan doesn't justify that sort of thing.