Stellaris Dev Diary #258 - 3.4.4 ‘Cepheus’ Information

[Cadzon]

I stopped playing stellaris few months after the pop revoke, can't even remember what the reason was but the pop migrations and crowding etc was insufferable.

I just tried the game again and wow, it's light years better in every way possible. Holy shit, some genius have really figured this out. Thank you so much, and very very good job, I can't say it enough.

 

[UltimateTobi]

Any word on military AI improvements in the near to mid future?

 

[PlasticFacecrops]

one system scholariums become entirely pointless.

I'm okay with that. It feels pretty arbitrary and gamist to want a vassal not because of what they produce or perform, but because they have some special magic designation that gives you bonuses 'because the game says so'.

 

[zZander56]

Wish we'd see a Branch Office Expansion Planner on that list of summer projects.

Sorry, I'm contractually obligated to say it at least once every dev diary.

 

[Nevars]

I’d like to have another look at some of the unity jobs, if I have the time over summer.

Will you take a look about unity output between ethic?

Anyway to make spiritualist actually good at unity generation like they are suppose to be good at, instead of them being middling at best right now?

Or if it's really impossible (considering materialist technocracy dominated them before and now egalitarian, xenophile and xenophobe outperformed them easily now I'm starting to think that spiritualist are cursed) then we might need to rethink about what spiritualist suppose to be good at.

Right now the only good point of spiritualist is relatively easier access to psionic ascension (or guaranteed but that need to sacrifice origin slot) than other ethic.

 

[Ferrus Animus]

Worth keeping in mind, that the holding still gives +3% research speed buff.

That sounds like a mess of a design tbh.

 

[Enero]

Will automatic research pick dangerous techs should they be available (as it currently does)?

I would suggest prohibiting dangerous techs selection.

 

[Sutopia]

That sounds like a mess of a design tbh.

Yeah, why is only scholarium special holding granting a significant empire wide bonus while prospectorium is stuck in a few pieces of scrap metal at a huge loyalty penalty? It’s never worth using.

 

[Franton]

• Added Hydroponics Station designation for habitats. This designation removes 2 housing from Habitation Districts in exchange for adding 1 farmer jobs.

Just for clarification:
- Is that -2 housing for every single habitation district?
- Is that +1 farmer job added to each Hydroponics building? As written it sounds like it's +1 job to the entire habitat, or +1 farmer job per habitation district, which would be kind of ... unusual.

 

[Chuunibyou Imouto]

Any word on military AI improvements in the near to mid future?

They just did this last patch, and have already said a full combat rework is something they'd like to do in the future. The problem is they tried to make the AI better at combat, but then released Hype Relays which broke path finding. The AI is worse at war now than before IMO, because the hyper relays seem to make them *really* bad about suciding in their fleets.

I'll have 15K worth of fleets sitting one system over, and they will send in three 7k fleets one at a time and end up losing the war because of it. I'm assuming they are trying to group their fleets together, but the hyper relay makes the pathfinding ignore dangerous enemy fleets I guess.

Setting myself a reminder to get this done when I have some free time.

Unrelated but since Merc Enclaves seem to have been your baby, is there any chance of them getting a UI pass in the near future? Not even the sweeping major changes that would be nice to have but a lot of work, they really just need some minor updates to increase their functionality like 9,000 percent

Simple things like

• A pointer on the Enclave and on the fleet, to go back and forth between the two so you can find the Enclaves fleet, and find the Enclave that owns the fleet. Just a button at the top of the fleet to take you to the Enclaves base that owns them would help a LOT
• A way to auto renew the contract on expiration so they don't go rogue on you mid war when the contract ends and an AI auto rents it in 0.0124 pico seconds
• A pop up warning if the system detects one of your own merc fleets as hostile. Just a little skull at the top of your screen saying "Hey maybe you should bribe these guys" would be very welcome

would go a loooooooong way towards making the Merc fleets more usable and less of a liability. They are so freaking cool RP wise, but after getting absolutely ruined by one of my own merc fleets because I couldn't find the enclave that it belonged to (because I had 5 enclaves total), I'm pretty iffy on building them again.

Will automatic research pick dangerous techs should they be available (as it currently does)?

I would suggest to prohibit dangerous techs selection.

Better yet, can we just get an option to black list certain techs from the auto research? Just a button next to the tech to flag it for "do not research" would be fantastic. You could then turn on auto research once you felt like you'd flagged all the ones you don't want to research

Yeah, why is only scholarium special holding granting a significant empire wide bonus while prospectorium is stuck in a few pieces of scrap metal at a huge loyalty penalty? It’s never worth using.

Prospectorium as a whole need a buff. It's the worst specialization by far to actually play as, and there's not really that much reason to even have them under you either. Tributaries and Vassals can get you mostly as much money from what I've seen, without you having to pay them anything, and none of the unique Prospectorium stuff is that crazy for the overlord compared to Scholarium. I almost always end up with 1 or maybe 2 Bulwarks (because they are expensive AF), and 1 or maybe 2 Prospectorium, and then just a dozen scholarium / protectorates / tributaries

 

[UltimateTobi]

They just did this last patch, and have already said a full combat rework is something they'd like to do in the future. The problem is they tried to make the AI better at combat, but then released Hype Relays which broke path finding. The AI is worse at war now than before IMO, because the hyper relays seem to make them *really* bad about suciding in their fleets.

I'll have 15K worth of fleets sitting one system over, and they will send in three 7k fleets one at a time and end up losing the war because of it. I'm assuming they are trying to group their fleets together, but the hyper relay makes the pathfinding ignore dangerous enemy fleets I guess.

That is exactly why I ask. They said they hadn't had the time to make substantial changes to the military AI, but would afterwards. So I wanted to know how long that would take approximately.
Obviously until after the summer break. But I hope not a year+.

 

[Starfury]

Better yet, can we just get an option to black list certain techs from the auto research? Just a button next to the tech to flag it for "do not research" would be fantastic. You could then turn on auto research once you felt like you'd flagged all the ones you don't want to research

Even better, can we not have techs you shouldn't research? (Not sure which ones you mean though, apart from some techs being far more important than others, ultimately I'll research everything)

If the dangerous techs actually were, that might be interesting too.

 

[Nefezer]

They do get their bonus, minus one stage. The issue with vassal rebellions is like three fold

1. They lose 1 stage of their difficulty bonus, and part of that bonus is stability. So their stability goes down a bit instantly even in a perfect situation
2. Usually the player is levying taxes against them too, which will further lower their resources. And I am 99.99% sure those are the gross resources produced before accounting for what they are consuming, so they might be making 1900 energy, but were only actually getting like 500 per month after all their expenses. Suddenly the player comes in and takes 60% of that 1900 and tanks their economy, AND their economy also goes down one stage too because of the difficulty bonus going down
3. The AI is still bad at managing their worlds, but most especially Primitive Planets. They invade primitive planets which then are hit by stellar shock, and the AI can barely manage to hold on to those in the absolute best of times. Then suddenly a player comes in and repeatedly kicks the AI in the testicles while they are on the ground, and the primitive planet / habitat (which they are also A W F U L at managing) starts to rebel and the AI can't stabilize it while their entire economy is on fire

And that's not even mentioning the devastation their planets probably have after the player beat the crap out of them and subjugated them a lot of the time.



The third point especially is what leads to a ton of chaos. There were stories in the first week before the hotfix where a single primitive planet would rebel against an AI, rebel against itself after it was freed from the AI, rebel against a player who took it, rebel against itself AGAIN, rebel against the original AI when they took it back, etc. They turn into black holes of rebellions because of the nonstop devastation and stellar shock and opposing ethics each time someone new takes it etc.

Another issue I've personally found when inspecting my vassal worlds that keep rebelling, is the AI has WAY too many pops that are unemployed and nothing to do with them. They will have 40+ pops sitting there not working, with all the districts and building slots maxed out / no money to build new stuff, and then the planet rebels. My vassal last game had literally like 8 or 9 rebellions in a single game, because of over population. I eventually just took their planet and rebuilt it from the ground up to not be hot garbage and then gave it back so they would stop dragging me into wars against the rebels every 10 years


I'm quite sad that the patch doesn't add any new tools for Overlords to help their vassals stabilize their empires. We really, really need a panel that lets us see that our vassals are having major stability / economy issues so we can build a holding that helps or lighten their taxes etc

¿Uh?... As a Resolute Assimilator, I never felt the need for vassals... but after reading about how your vassals go through so much hardship, I'm tempted to at least have one... just for the sheer joy of watching it burn

 

[Sutopia]

Can the wall of text people take their 1 vs 1 debate elsewhere, kthanks




They do get their bonus, minus one stage. The issue with vassal rebellions is like three fold

1. They lose 1 stage of their difficulty bonus, and part of that bonus is stability. So their stability goes down a bit instantly even in a perfect situation
2. Usually the player is levying taxes against them too, which will further lower their resources. And I am 99.99% sure those are the gross resources produced before accounting for what they are consuming, so they might be making 1900 energy, but were only actually getting like 500 per month after all their expenses. Suddenly the player comes in and takes 60% of that 1900 and tanks their economy, AND their economy also goes down one stage too because of the difficulty bonus going down
3. The AI is still bad at managing their worlds, but most especially Primitive Planets. They invade primitive planets which then are hit by stellar shock, and the AI can barely manage to hold on to those in the absolute best of times. Then suddenly a player comes in and repeatedly kicks the AI in the testicles while they are on the ground, and the primitive planet / habitat (which they are also A W F U L at managing) starts to rebel and the AI can't stabilize it while their entire economy is on fire

And that's not even mentioning the devastation their planets probably have after the player beat the crap out of them and subjugated them a lot of the time.



The third point especially is what leads to a ton of chaos. There were stories in the first week before the hotfix where a single primitive planet would rebel against an AI, rebel against itself after it was freed from the AI, rebel against a player who took it, rebel against itself AGAIN, rebel against the original AI when they took it back, etc. They turn into black holes of rebellions because of the nonstop devastation and stellar shock and opposing ethics each time someone new takes it etc.

Another issue I've personally found when inspecting my vassal worlds that keep rebelling, is the AI has WAY too many pops that are unemployed and nothing to do with them. They will have 40+ pops sitting there not working, with all the districts and building slots maxed out / no money to build new stuff, and then the planet rebels. My vassal last game had literally like 8 or 9 rebellions in a single game, because of over population. I eventually just took their planet and rebuilt it from the ground up to not be hot garbage and then gave it back so they would stop dragging me into wars against the rebels every 10 years


I'm quite sad that the patch doesn't add any new tools for Overlords to help their vassals stabilize their empires. We really, really need a panel that lets us see that our vassals are having major stability / economy issues so we can build a holding that helps or lighten their taxes etc

A random question, did you modify the pop growth rates or used some mods to disable AI from building habitats? I have never seen AI having a lot of unemployment.

 

[currylambchop]

You can't run a project the size of Stellaris in a constant state of emergency. And that's what releasing every few weeks would be.

Plenty of games have updated every few weeks, league of legends for one

 

[KNakamura]

Plenty of games have updated every few weeks, league of legends for one

I wouldn't call LoL and Stellaris similar.

 

[Starfury]

Plenty of games have updated every few weeks, league of legends for one

LoL is a service game, designed for constant engagement.
Those games must deliver regular content and their teams are designed from the ground up to do so.

And why do they need to update every few weeks? You'd think they'd have fixed everything by now

 

[Chuunibyou Imouto]

Even better, can we not have techs you shouldn't research? (Not sure which ones you mean though, apart from some techs being far more important than others, ultimately I'll research everything)

Some techs just give dead end buildings that you just generally aren't going to want to build in any situation, or are flat out bait even. Stuff like upgraded commercial buildings if you aren't a trade empire, or gene clinics which are still very controversial, or weapon techs if you aren't using that weapon type etc. Especially repeatables like kinetic if you are full energy

A random question, did you modify the pop growth rates or used some mods to disable AI from building habitats? I have never seen AI having a lot of unemployment.

No, they were just not able to expand any further, and just weren't building many habitats. The AI is absolutely terrible at using habitats as well, so even they do build them they usually turn them into a crime against all sentient life everywhere with what they build on it. I actually think a lot of the rebellions that happen are because of habitats with horrible stability causing the entire system to rebel

 

[pmchem]

LoL is a service game, designed for constant engagement.
Those games must deliver regular content and their teams are designed from the ground up to do so.

And why do they need to update every few weeks? You'd think they'd have fixed everything by now

in case you haven't noticed the Paradox DLC business model, Stellaris and most of their other games also rely on delivering regular content in order to make money.

But, the size/budget of the Stellaris team is... nowhere near comparable to a game like LoL, so having releases every few weeks is not realistic.

On a related note, I fully admit I don't know all the overhead required for a steam beta.

Nevertheless, I think that prior open beta feedback for Stellaris has been extraordinarily useful for catching balance issues and bugs. The Stellaris QA team is the first to admit they won't catch all the bugs in weeks of internal testing that users can catch on day one of a wide release. It's a simple matter of human-hours interacting with the game: there's a lot more players than devs.

Given that a release before the PDX/European summer holiday season is intended to be exceptionally long-lived, I think it would have benefitted from an open beta. For example, if the "AI's leave federations because of their subjects" and hyper relay pathing bugs are not fixed in 3.4.4, well, I'm sure an open beta would have provided emphasis regarding the importance of those to the playerbase. I guess we'll see what the final changelog looks like next week. Certainly the devs are aware of those particular bugs, as both of those classes of bugs were highlighted in multiple reports with save games and sirblack's thread. And earlier in this thread.

Finally, 3.4.4 is adding some entirely new things. It's not ONLY bugfixes. So if one of those new things accidentally introduces a new major issue -- we're stuck with it for many months. An open beta might lead to a more chill summer for both players and devs.

 

[zZander56]

There's something about this new patch that just feels lazy to me. I've seen too many bugs, too many half-efforts like the thousand new things wrong with subject AI, Mercenaries, Hyper-Relay effects, the new worthless resolutions, the anemic shroud enclave, poor holding balance, broken new events and text. When did the new game director come in? Was this just a bad cycle, or bad management? How could such an unbaked patch be acceptable to release before a summer break?