As a new player; Grand Admiral-

[Part-Time Prick]

Okay, had my first game that went... about as far as I could be bothered before I got bored. year 2400 ish somewhere, endgame crisis not spawning, AI are all pathetic cause I was playing on Ensign but my research had stalled out at like the 3k mark pretty much.

Not particularly impressive, I know.

Figured that after buying the new player pack I'd make my next game on the complete opposite end of the spectrum on grand admiral... I understand I'd get cooked more than likely, but is that even a remotely good idea or no?

[PlutonArioch]

Sure why not? If you get stomped you know that you should find your sweet spot somewhere in between. And if not, you'll probably have a less boring game for your second time.

[CrUsHeR]

GA is generally not an option unless you use super-exploitative meta builds exactly designed to handle the challenge.

The AI players get way too many bonuses. +100% resources is insane, considering they not only produce double the alloys per metallurgist job, but also double the minerals to feed twice as many metallurgists. Double research, double consumer goods, food, and so on.

Scaling difficulty is a sweet spot, because it allows you to establish yourself without an AI player sending fleets of 6K your way before you even settled your second habitable planet.


You also may want to consider creating your AI players yourself, setting them forced to spawn. This allows you to at least have a galaxy with a mix of inhabitats you feel comfortable with. And with species you also would want to have on your planets (e.g. no "Noxious" pops, no Clone Army pops or whatever you don't want to see)

[DWI_Suriil]

Grand admiral is not difficult. It just gives AI huge early game boost , giving them 3k-5k fleets early on. If you survive to 2230 with good build you will outscale AI anyway, because AI suck at eco management , AI suck at ship design , AI suck at research priorities , etc etc etc.
Only way the AI can ♥♥♥♥ you up - you spawn near exterminator or devouring swarm and they bully you on year 2210-2215.
In any other case , even if you become a vassal - thats not a pb , you will brake free soon after. If you get friendly civs around , they will not bully you and even can help you , so....

Mid game and late game AI suck ♥♥♥♥ no matter the difficulty. By year 2280 you outscale any AI even with all bonuses they can have.

THE ONLY challenge , that is actually hard in PvE stellaris is GA + x25 crisis all 4 crises at the same time + Khan 2300 (mid game start 2290 option).
Since crises scale on x25 AND difficulty , GM + x25 is like 5million crisis fleets.
And once you beat one crisis , next one is even stronger , etc etc.
Cetana fleets are like 12 millions per fleet :D
And her main base is like 50 millions XD

In this case there is no hope AI will save the galaxy from Khan or endgame crises , they just suck . Its all you vs world. Its pretty fun as a challenge , as you know, timer is ticking , and you have until 2400 (if you set lategame to 2400), to get ready for the end of the world

[DWI_Suriil]

That being said....before all of that you have to understand lot of in game mechanics.
For example by year 2400 you should sit on at least 7k science , 3k unity , 8k alloys , 800k fleet power , 4000 fleet capacity, .....

But in order to do that you need to understand how to
*manage empire size
*rush and pick traditions
*build ship designs, know best weapons
*manage ,specialize and ascend planets
*pick your tech to rush mega structures by year 2300
.....

[Razorblade]

You'll definitely get stomped the first few GA games, but you'll learn fast enough, as long as you're not prone to rage quitting.

On the bright side, the higher the difficulty, the more likely it is that the AI will vassalize you at first contact, rather than annexing you outright (unless it's a Genocidal empire). Back before the Overlord update, the GA AI would vassalize you, then annex you after 10 years. You had to buy your way out with 10 Favors and the +5 Diplomatic Acceptance from the Diplomacy tradition tree to avoid a game over.

[Xaphnir]

With only 3k research per month at 2400, you will get your ♥♥♥♥ kicked in on GA.

[corisai]

I would suggest playing mid-game scaling GA instead, just pick early date for mid-game (like 2250 or even 2225 if you dare) and disable L-gates.

Why it's better then non-scaling GA? Because you're able to play not only meta rushing builds.

Why no L-gates? Because early Khan isn't going to end your game (you can submit to him), but early Grey Tempest could.

[Razorblade]

Originally posted by Xaphnir:

With only 3k research per month at 2400, you will get your ♥♥♥♥ kicked in on GA.

I'm going to drop some controversy on you: Research does not matter in singleplayer. It's nice to have, of course, but I've beaten GA with a single planet and only the single starting Research Lab.

More ships solves pretty much any problem.

[corisai]

Originally posted by Razorblade:

I'm going to drop some controversy on you: Research does not matter in singleplayer.

^^ Second this (in case of vanilla Stellaris).

I don't do something so extreme as having only a single Lab, but would agree that you can win games with very weak research. From my personal experience - using old Crisis perk (still very powerful choice, even if new stuff is stronger).

AI in game isn't self-aware delphic AI so it can't abuse advantage in techs like humans do. Vanilla ship loadouts & fleet compositions are horrible mess too.

[Xaphnir]

Originally posted by Razorblade:

I'm going to drop some controversy on you: Research does not matter in singleplayer. It's nice to have, of course, but I've beaten GA with a single planet and only the single starting Research Lab.

More ships solves pretty much any problem.

It's not so much that it's necessary, but more that it illustrates OP's ability at the game. Yeah, you can beat the AI while mostly ignoring research in the early game, but that still requires playing the game in other ways well.

[DWI_Suriil]

You cant reliably beat the x25 GA crisis w/o solid research though, to get your repeatable done. Fleet only? Well , there are limits on how many you can keep in field and how many resources you got in stock for rebuild after first one is melted.

And you will ♥♥♥♥ up face to the Khan x25 GA in 2280-2300 , w/o solid research , as you need late game ships designs to stand him. The T3 plasma , Arcs , hyper shields and proton armor.
And if you fail face to the Khan , then you lose lot of time to get ready to the x25 endgame Crisis. And you are dead, if you are to face all 4 crises.

At some point Stellaris is not about bullying AI empires , but running against a watch vs endgame Crisis.

[corisai]

Originally posted by DWI_Suriil:

You cant reliably beat the x25 GA crisis w/o solid research though

OP said nothing about x25 crisis ...

[[OCe]Kagari]

You can also turn on "Difficulty Adjusted AI modifiers" beside going GA so the AI will still relevant till the end, without it, GA is just hard in the early game and they will get easier over time