Instalar Steam
iniciar sesión
|
idioma
简体中文 (chino simplificado)
繁體中文 (chino tradicional)
日本語 (japonés)
한국어 (coreano)
ไทย (tailandés)
Български (búlgaro)
Čeština (checo)
Dansk (danés)
Deutsch (alemán)
English (inglés)
Español de Hispanoamérica
Ελληνικά (griego)
Français (francés)
Italiano
Bahasa Indonesia (indonesio)
Magyar (húngaro)
Nederlands (holandés)
Norsk (noruego)
Polski (polaco)
Português (Portugués de Portugal)
Português-Brasil (portugués de Brasil)
Română (rumano)
Русский (ruso)
Suomi (finés)
Svenska (sueco)
Türkçe (turco)
Tiếng Việt (vietnamita)
Українська (ucraniano)
Comunicar un error de traducción
Normas y directrices sobre discusión
Denunciar este mensaje
This story is loaded with inconsistencies. Aztecs do not have Javalineers, and are a Medieval Era culture, so you cannot be Aztecs AND English.
The Aztecs UU is a Melee unit, called Jaguar Knight, not a Javelin unit.
♥♥♥♥♥♥♥... You're right, it was Maya. I always get them mixed up in my head. It's even right there in the screen shots and I mixed them up. Leading with "loaded with inconsistencies" makes me figure you probably didn't meant to point out an honest mistake that could have easily been head-corrected by glancing at the screencaps, but it was helpful anyway.
By the way, what other inconsistencies are we talking about, here?
Oh I got
But you playing on metropolis doesn't prove what he said wrong.
the AI does cheat on metropolis... it's given plenty of advantages and that's just an objective fact.....
You didn't prove anyone wrong by posting a picture of a game where you ended up snowballing over the AI early because it's been quite clearly established already that the player can snowball before the AI
also him saying that the ai score explodes if you're under it (talking about early eras) doesn't mean he is arguing that you can't eventually surpass that lead later in the game, especially on easier difficulties.
I've never played with Midas, what are his traits? Maybe one or more of those is the culprit. If I can just remember who it was when it happened to me, I can cross reference...
I'm fairly certain that unless an AI leader has a fame multiplier attribute, they get the same amount of fame that you do. They have additional resources based on difficulty I think, but I don't think they get to cheese the fame score. Regardless, its easy to be in the middle of the pack for the first 2-3 eras and then explode when you hit your power spike. It sounds like you're drawing conclusions, and then giving up early. The game actually has a lot of room for combacks.
Also with 60 hours in the game, I have been pretty impressed at how viable almost all of the early civs are tbh. I think babylonians, hitites, and to a lesser degree the olmec, are fairly weak, but the rest of the 7/10 are all pretty solid at something. Some are definitely stronger than others, but this is a game about known how to leverage situational strength.
Another note: I think gold econ is actually the strongest econ option in the game right now. Its not as immediately useful as the other resources, but taking an econ faction lets you get a bunch of trade routes via reselling, and rack up early merchant rep, which leads to a ton of map vision and keeps you relatively insulated from hostility, since people are reliant on your resources.
I recently broke the game over my knee using a purely econ strat. Phoenicians->Carthaginian->Ganain had me generating like 15-20,000 gold a turn by turn ~150 on normal speed on Empire difficulty. I switched to science and started researching 3-4 techs a turn, getting a science victory within like 20 turns. Even before I switched to science, I was able to buy out enough stuff every turn that I was easily eclipsing anything an industry faction could put out at the same point.
The AI DOES get cheat advantages on Metropolis. It gets a flat 5% to all FIMS output to all cities. Not to mention over discussed 'special awareness' behaviours also covered earlier in the thread.
If you think thats even up for dispute then you really need to read up on the fundamental basics of how the game works first before trying to disprove the already firmly established.
I've already linked it in this thread but as it seems people are diving in to throw around ignorance as fact based on nothing but "Look I beat the game so all I say is true".
https://humankind.fandom.com/wiki/Difficulty
I really don't get this thread. It started off covering, discussing and further establishing a legitimate aspect with the early AI behaviour that allows for the AI to snowball to varying degrees, in some cases as extreme as >5000 fame points by turn 110 with a tier 4 religion and achieving the early modern era.
The reason this snowball aspect to early-mid game AI potential tapers off and doesn't pose a end game issue is largely influenced by later game concepts being handled very very poorly by the AI.
The AI is fine handling primitive and medieval era mechanics fine, but once things like ranged artillery and other more nuanced building combinations start coming into play that go beyond "Just build this to get better", the AI just crumbles.
But after about page 4 or 5, it feels like the thread largely got taken over by seemingly randomness with people deciding to draw their own lines in the sand and going "Hey look, I can beat the game so there can't be problems!"
Beating the game was never presented as a issue, just that in certain situations that are relatively common enough for quite a few people to have encountered them at least on multiple occasions, the AI is capable of truly crazy progression leaps well beyond anything outlined possible by players.
For all of Dimitri somewhat derailing the thread and pushing this incorrectly more into a "Can you beat the AI?" experiment, they're spending a whole lot of time illustrating that even with a whole lot of good luck and gaming the design they can only touch on the very bottom end off what a AI empire has been clearly shown capable of in certain situations.
The actual experiment is, what combination of factors is allowing the AI to snowball like it can in fringe cases. Is it largely map/world factors? Culture factors? Or even AI archetype behaviour factors?
(I will add though I don't think Dimitri quite intended to completely derail the thread and draw others into further doing so, they just got caught up in their own exercise and didn't consider if it was at all the same thing.)
On Civilization difficulty and above, the AI begins to get flat out stability bonuses on all cities (+45 stability on all cities on civilization difficulty, and +60 on humankind difficulty for those that refuse to click links to get information).
This allows them to support significantly more districts before suffering any kind of stability loss (assuming the AI adheres to basic rules like stability that apply to the player).
This is also further augmented by the AIs +2 output also given on Civilization difficulty per every population on all tasks applied to the AI given that population limits are directly dictated by distracts.... the more districts you can support the higher your population cap per FIMS category. Higher your capacity the more that bonus output per population stacks to greater and greater leads in output potential.
Testing what a human can build, really can't replicate anything the AI can potentially achieve once silliness like that steps in.
That said, the >5000 fame by turn 110, early modern era with a tier 4 religion case, that was a AI snowballing on Metropolis difficulty, on normal pacing.
I understand the frustration about the AI's bonuses, but the reality is that making AI for this kind of game is really hard- take it from a game developer who has been writing AI recently. Even on the harder difficulties, superior decisions making can let you totally eclipse the AI given enough time. I too am frustrated about the AI, but not because its too hard. As a veteran endless legend player, I've found myself a little frustrated that by turn 150, I am 100% sure I will win, since I have out-optimized the AI, and playing out the rest of the game at that point is more of a chore.
Hopefully either the devs make the base AI better, so they can reduce the FIMS bonuses they get with difficulty, or the community gets together and makes another project like the Endless Legend Community patch (which totally overhauls the AI and makes it much more enjoyable to play against)
Ah fair enough.
You didn't run that test game as an actual experiment, you ran it in a vain attempt to prove OP wrong and show him that it was actually that "the AI is just transcending silly, this is totally fine".... which you did and only to embarrass yourself proving the OP's point