Home Trending Latest threads New posts Developer diaries
Jump to latest Follow Reply
  • We have updated our Community Code of Conduct. Please read through the new rules for the forum that are an integral part of Paradox Interactive’s User Agreement.
Stellaris 53097 Bug Reports 34469 Console edition 1306 Suggestions 21148 Tech Support 3149 Multiplayer 391 User Mods 4771 Stellaris AAR (After Action Reports)
Previous dev diary Next dev diary
Show only dev responses

Stellaris Dev Diary #366 - Announcing Stellaris 4.0

Happy New Year! It’s good to be back!

I want to start by welcoming all of the new Stellaris players who joined us during the Winter Sale, and to our Chinese community, which has grown so much over the last year, 欢迎光临。

Next, I want to draw your attention to several feedback threads that have been running for the past few weeks. These threads have forms you can fill out to share your thoughts.
Your feedback is essential in shaping Stellaris's future, and I’m extremely grateful for the strong response we’ve received so far.

For some time I’ve been hinting that the Custodian team has been working on something big, so now let’s look at what they’ve been up to and what we’re planning for the first half of 2025.

A Moment of Prophecy?​

A long, long time ago, I was asked when we would move on to Stellaris 4.0, and I answered “Definitely not until we get to release Update 3.14”.

Psionic Event Art

Little did I know how prophetic that joke really was.​

Announcing Stellaris 4.0​

The Q2 Stellaris release, currently expected sometime around our Anniversary in May, will be the Stellaris 4.0 ‘Phoenix’ update.
It will be released alongside our major expansion for the year.​

While designing the plan for the Stellaris 4.0 release, the Custodian team had the following major priorities:
  • Performance Improvements
  • New Player Guidance and Game Pacing
  • Quality of Life Improvements
As much of this is still very deep in active development, I don’t have too many screenshots to show off yet, so I’ll go over some of what we have planned and provide more in-depth details in future dev diaries. As they get closer to completion, some of these features will likely change as we iterate on them, and it’s possible that some may end up very different from how they were described in this dev diary, be delayed, or even cut altogether - these are some of the risks of sharing plans in an early stage, but I feel that the benefits outweigh any potential drawbacks.

Performance Improvements​

Stellaris has many moving parts, and an incredible number of calculations are performed every month. Many of those calculations rely on others, forcing them to be performed sequentially rather than in parallel. This causes the game to slow down as the number of calculations increases throughout the game and is especially noticeable in large galaxies - more planets and empires means more pops filling more jobs, producing more resources, with more pathfinding for the fleets, and so on.

Pops and Jobs​

The Pop and Jobs system introduced in Stellaris 2.2 ‘Le Guin’ have always had major performance implications in the late game, and we’ve been working on incremental improvements ever since.

The Tech Pope Speaks​

Last year I mentioned that we were exploring a Pop Groups prototype, and showed you a horrifying placeholder screenshot in the last dev diary of the year. Our initial experiments have been promising, so in the Stellaris 4.0 ‘Phoenix’ update, we’re changing the way Pops fundamentally work. Pops will be grouped together into Pop Groups based on species, strata, and ethics, and these Pop Groups will produce Workforce that is used to fill (or partially fill) Jobs. As part of this change, we’re changing the overall scale of Pops - most things that previously affected or manipulated 1 Pop would now affect or manipulate 100.

These changes will significantly impact other systems, such as Pop Growth, Migration, and many others. I’ll dedicate a full dev diary to more details before the Open Beta.

Trade​

The current Trade system, with its constant calculations around pathing and pirate generation, is another that has a disproportionately high impact on performance compared to the benefit. We’re simplifying that one significantly and making Trade act as a standard resource. Trade will also be used to represent general logistics capability and as such, will likely become available to gestalt empires for these logistical purposes. Again, we’ll cover this in a future dev diary.

Additional Comments​

Fleets are the remaining system I’d highlight for having a major performance impact. While 4.0 will have some general fixes, we’ve got our hands full with these changes so we’re expecting to focus more on them in a future update.

New Player Guidance and Game Pacing​

Much of the feedback we’ve received from newer players indicates that Stellaris has become overwhelming in the early stages of the game, providing a flood of decisions and a seemingly endless barrage of notifications. They have trouble identifying which of these choices are important for long-term growth versus which are primarily flavor, and the constant interruptions make it difficult to form both short-term and long-term goals.

More Meaningful Events​

The Content Design team has been reviewing events and notifications to ensure that any interruptions are meaningful. Events should generally not be purely informative – you should have a choice that has an impact. A substantial number of purely informational events, such as the discovery of Terraforming Candidates or new Strategic Resources, have been converted into toasts or notifications.

As an example, during your first steps to the stars you’ll find evidence that life is surprisingly common out in the galaxy. While this used to simply have an acknowledgment, you’ll now have choices based on the nature of your empire.

Simple Forms of Alien Life event, now with potentially useful event options

Event options should help guide the way your empire grows.​

Anomalies are a wonderful content delivery vehicle during the exploration phase, but having a window pop up in your face every time one of your science ships finds anything interesting is decidedly less wonderful. We’re moving the popup to a Toast - you can click it or a notification to open the full anomaly window, or get to it through the Situation Log.

Anomaly Toast, including difficulty and flavor text

Anomalous readings registered! ​

Certain event chains that are not particularly loved have had (or will have) a bit of adjustment as well.

The Divine Glory-class Battleship from the Radical Cultists event chain

Radical.​

Message Settings​

Speaking of Toasts and Notifications, the Message Settings system has been expanded to give you more control over how different messages should appear.

Message Settings configuration: Notification, Toast, Popup, and Auto-Pause can each be toggled

We’re doing a pass on the default settings for each as well.​

The new Message Settings should allow you to customize your notifications to suit your preferences – whether you want a popup that automatically pauses the game or to turn certain notifications completely off.

Leader Trait Frequency​

Empire Leaders were cited in your feedback as feeling very needy, like they’re constantly clamoring for attention to select new traits if you owned Galactic Paragons. We’re looking at merging the first two tiers of leader traits and reducing the number of levels that you make trait selections at - this has the net effect of increasing the overall power of leaders a bit (as they’ll start with what was formerly a tier 2 trait, and if you select a new trait at level 3 instead of upgrading their starting trait, you’ll have two formerly tier 2 traits), but makes the experience with them a bit smoother.

Fewer trait selections do put you at greater mercy of the random selection of options, so we’re increasing the number of option draws by 1. This should reduce some of the risk of getting a “dead trait” without diminishing the benefit of +1 Leader Trait Option effects too much.

Galaxy Generation Updates​

As Stellaris has grown, so has the number of pre-scripted systems. Many of these unique systems were set at extremely high weights to appear, causing most of them to appear in every game you play. Since these special systems usually contained one or more habitable worlds, it inflated the number of such worlds well above the expected number, especially since they did not respect the Habitable Worlds slider from your settings.

We’ve done a normalization pass on the weights of these systems - many should still appear in each game, but it shouldn’t try to stuff all of them in. They also now respect the Habitable Worlds and Pre-FTL sliders from galaxy setup if appropriate, and should generally no longer appear in the immediate vicinity of Empire homeworlds.

This change yields general benefits to game pacing and indirectly, an improvement to performance in general.

Empire Focuses​

The Focus Trees in some of our other Grand Strategy Games do a great job of outlining possible ways you could take your country. In Hearts of Iron, for example, you already know the general “plot” - the different factions will behave as you expect until World Tension reaches a certain level, after which the world descends into war. The differences that will occur from game to game are largely due to how the events play out, and your interference in history lets everything spiral out into an alternate resolution. The Focus Trees not only provide a great way to create butterflies that can change history but are fantastic at providing new players with short and medium-term goals.

We decided that static Focus Trees were not appropriate for Stellaris though - our sandbox and 4X nature with a mysterious universe require any such systems to be more adaptable to what’s happening in this galaxy. Instead of trees, we’ve decided to go with suggested tasks that fall into Conquest, Exploration, or Development aspiration categories - these can range from investigating an anomaly to building a Dyson Swarm, or at the highest ranks, even becoming Galactic Custodian. You’ll be able to select your empire’s focused aspiration, which will skew the offered tasks towards your choice.

Completing these tasks gives no immediate reward, but progresses you down Conquest, Exploration, and Development tracks, and if you get a task that you’ve already completed that’s fine - it’ll immediately complete and you can get a new one. We don’t want you to sit there waiting to build your Interstellar Assembly, after all. Reaching certain milestones will grant abilities like Form Federation (which will be moving out of the Diplomatic Traditions), or give guaranteed research options for critical technologies, reducing your reliance on random pulls from the technology deck for techs like Cruisers, Colonial Centralization, or Mega-Engineering.

Veteran players already know how to play the game and are already adept at forming their own goals. We expect that you’ll already be completing these tasks naturally as you play - they’re primarily intended to teach new players how to play like you and guarantee that you’ll be able to force access to those important technologies.

Empire Timeline​

Accessible via a new tab within the Situation Log, the Empire Timeline is a real-time chronicle of your empire’s journey. From humble beginnings on your homeworld to the heights of galactic dominance (or the depths of ignominious defeat), the timeline will automatically document key events and milestones as they occur.

We aim for the Timeline to serve as a practical ledger, allowing you to retrace the pivotal decisions and moments that have shaped your game. It will also provide a rich narrative framework, transforming your gameplay into a story worth remembering.

We look forward to sharing more details on the Empire Timeline in a future diary. For now, we invite you to prepare your empires for posterity – and to ensure that your name echoes across the stars.

Quality of Life Improvements​

Many of the other changes also fall into Quality of Life Improvements, but two I want to highlight in particular include improvements to the Species Modification process and the Colonization flow.

Colonization Process​

Colonizing worlds had a few quirks that we’re smoothing out to make for a better experience, especially if you use Colony Automation. We’re changing the “Colony” designation to a modifier that will exist for some time after initial colonization, and letting you pick a Colony Designation and even turn automation on when you give the colonization order. This should prevent a common situation in the mid to late game where you would colonize a planet, but would have to pick and choose between using automation or losing out on the amenity and stability bonuses of the default designation.

The new flow also helps out Automation significantly since you won’t end up in a situation where Colony is no longer a valid designation and it falls back to an auto-designated selection.

Species Modification and Assimilation Targets​

We’ve gone through the genetic modification process to remove many pain points and make the overall flow much smoother. You’ll also be able to set a template as the species default, and can set sub-species variants to automatically integrate over time into the species default template.

New Species Tab showing Sub-Species Integration Species Rights

The Species tab is generally more helpful as well.
Note: This branch does not include the pop changes.​

Ship Designer​

As we did with Species Modification, we’ve gone through the Ship Designer to improve the general process of creating new ship designs.

Ship Designer, showing Ship Roles selection window

And the Auto-generate designs checkbox won’t stop you from saving a new ship design!​

The Next Few Weeks​

There’s a lot more going into this update as well - I’m hoping to challenge Lem for the Patch Note Crown.

Next week we’ll go into more detail about some of the changes coming in the Stellaris 4.0 ‘Phoenix’ update that are possible to show, including some things I didn’t go into above like Precursor Selection and the Stellaris Databank.

See you then!
 
  • 203Love
  • 126Like
  • 10
  • 5
  • 4
Reactions:
Sort by date Sort by reaction score
Jumping back into this, I don't get why people are getting so worked up about the focus trees. If they work exactly how they are suggested to, all they do is offer stuff for completing the tasks all you would likely have done anyways. Don't forget that the stuff is either things you could get from other sources (guaranteed research options) or stuff that was previously locked away behind less accessible means (form federation, which is currently part of the kind of trash diplomacy tradition tree).
 
  • 5
  • 2Like
Reactions:
passeris said:
@Thiend 's argument was originally that choosing a given tech would lock you out of others (ie. no chance). Yours was that randomly drawing others would lock you out of the one you want (no agency). You've now both agreed on a concern (the latter). But in theory this can happen with the current system: you just get unlucky and can't draw the T1 tech you want for 200 years, against astronomical odds.

Weighting doesn't create this problem, because the possibility is always there, and weighting can be adjusted besides. No matter how 'light' the weighting is (even zero), you would always be able to make this complaint when probability is involved. Either you are misunderstanding, or are fundamentally opposed to, a probabilistic system. I think it is the latter, which is why I'm concerned that players won't tolerate Focus Trees as a compromise when what they really want is to be rid of the card-based system.
Click to expand...
Preemptively, I notice you chose not to quote my post that directly pointed out two reasons something is a bad idea pointed out by different sources is not "mutually contradictory." According to your inaccurate summary of our objections, we actually made the same argument, you've just labeled them differently, making them mutually in agreement. Both are that random draw will determine what fields are available, and then your forced choice of those fields will reduce the chance of receiving options you didn't get in the previous draw. Both of us are and were talking about the fact that, if we want (as an example) Propulsion and don't have a Propulsion tech available that round, the chance of a Propulsion tech will go down - with no way out of the cycle.

"In theory a probabilistic system can screw you over with any percentages"

"Getting a losing result increases the chance of a losing result"

If you don't understand the difference between these two statements, there is nothing more to say.
 
passeris said:
@Thiend 's argument was originally that choosing a given tech would lock you out of others (ie. no chance). Yours was that randomly drawing others would lock you out of the one you want (no agency). You've now both agreed on a concern (the latter). But in theory this can happen with the current system: you just get unlucky and can't draw the T1 tech you want for 200 years, against astronomical odds.

Weighting doesn't create this problem, because the possibility is always there, and weighting can be adjusted besides. No matter how 'light' the weighting is (even zero), you would always be able to make this complaint when probability is involved. Either you are misunderstanding, or are fundamentally opposed to, a probabilistic system. I think it is the latter, which is why I'm concerned that players won't tolerate Focus Trees as a compromise when what they really want is to be rid of the card-based system.
Click to expand...
Choosing a given tech locking you out of others and randomly drawing others locking you out of the one you want are not identical complaints in all contexts, but in the context of the card system (limited card slots, probabilistic weighting) they are two different ways of phrasing the same fundamental complaint.

Yes, the chance of rolling player goal snake eyes repeatedly on the tech draws is a core issue with the probabilistic system with the current tech cards and yes I don't like it. I think there's much better ways to do what the Stellaris system does, and deliberately leaning into what-i-consider-a-weakness is not in my wishlist. But given the tech card system is here to stay having a second method that shows, slow, somewhat steady, observable accumulation of Development/Exploration/Conquest Points toward however many are required to gain Megastructures as a guaranteed tech is exactly my jam, even if each individual point is obtained semi-randomly.

E: I think I'd quite like your system if it was applied to AI empires though.
 
  • 1Like
Reactions:
Zaacplot300 said:
Jumping back into this, I don't get why people are getting so worked up about the focus trees. If they work exactly how they are suggested to, all they do is offer stuff for completing the tasks all you would likely have done anyways. Don't forget that the stuff is either things you could get from other sources (guaranteed research options) or stuff that was previously locked away behind less accessible means (form federation, which is currently part of the kind of trash diplomacy tradition tree).
Click to expand...
It's understandably hard to resist the allure of Fanatic Pessimist.
 
  • 9Haha
  • 1
  • 1
Reactions:
Enero said:
It is a pinnacle of "gamey" - a totally separate from the gameplay system that gives you bonuses out of thin air.

We have enough "rewards" presented in the form of relics, and other things. They are at least woven in the gameplay, unlike that Focus Tree.
Click to expand...
I agree that I don't need bonuses out of thin air. But if the system is built in such a way that there will be no bonuses at all, but only tasks, then this is a good guide for beginners. If the possibility of federation appears due to the fact that you are trying to improve relations and conclude all possible agreements with any country, then this seems logical. The problem is whether every such possibility is logical. I would suggest that such actions open up access to the tradition of diplomacy, and it is up to the player to decide whether to spend the empire's resources on this.
 
Have you tested the changes with larger galaxies?

I always wanted to play CK3 in space, but the amount of empires and galaxy size you can support is too low performance wise.
 
Critical Ethics said:
Stellaris has about a dozen gameplay loops. They all boil down to invest time, resources, and effort -> get stuff. This is no different. Why can I build a bunch of priests to make blue mana and then spend it on making my buildings cheaper? Why does filling out the discovery tree let me take an ascension perk that increases my edict capacity? Why does observing a collapsing star make me research better computers faster? Why will getting into a lot of wars allow me to tell my scientists to invest their time into researching a better warship? Why is it only the last one that sounds weird to you? Why does the last one sound weird to you at all?
Click to expand...
Exactly, all the questions are strange. I want there to be fewer strange questions, not more. I hope there won't be any new bonuses, just tasks.
We are talking about this so that the developers know that we are worried about this issue. It's worth telling about it before others.
 
Critical Ethics said:
Certainly... if you get to choose what it is. I want to go all in on gene techs over megastructures because I the player wanted to, not because the RNG decided the entire future of my empire on day 1. Edit: Ending up with a higher than usual number of tech A on a Tech B run is fine, it's the not getting into Tech B in the first place that's the issue.
Click to expand...
You can't make a scientific discovery according to plan. How can you plan to discover dark matter, and what if it doesn't exist? You can plan applied research, for example, to develop a cruiser based on other technologies.
 
Any step towards a reduced micromanagement and a performance boost are always welcome, even after upgrading my PC.
I've limited habitable worlds(and colonization/habitats) for a long time since it has been a microhell to manage plus the performance cost. I had to tinker the difficulty to keep up against the AI as managing fewer planets and avoiding habitats.
Really excited about the species integration update.
The only real loss would be the trade routes/pirate system, which to be honest has mostly been an annoyance.

The rest are all welcomed nice touches.
 
  • 2Like
Reactions:
yay, huzzah for new content!

please give players the ability to purge our main species and modify pops with multiple ascension perk / advanced traits. i want my bloomed gaia world utopian abundance beings to be both erudite and psychic. if i can’t apply that template to distant members of the same species, i’d like to compost them for resources.
 
Conceptually I liked what LeGuin did with pops and I have waited for the better part of a decade for Stellaris to finally become playable on the medium-longer term using that elegant design. It seems that finally PDX decided LeGuin was just too ambitious to manage. It is a shame both that it came to this and that it took so long to come to this if it was going to come to this.

I lament the loss, but here's hoping what we'll get is a truly functional system instead of some gimped version of the design that was already present that doesn't actually meaningfully sort things out. After all this time and with this backtracking, I expect to finally be able to get to the year 3000 on a huge galaxy without the single-core calculation bloat that would cripple every super computer on Earth.

My expectations are no less. Good luck with the redesign.
 
  • 1Like
Reactions:
Alastor said:
Conceptually I liked what LeGuin did with pops and I have waited for the better part of a decade for Stellaris to finally become playable on the medium-longer term using that elegant design. It seems that finally PDX decided LeGuin was just too ambitious to manage. It is a shame both that it came to this and that it took so long to come to this if it was going to come to this.

I lament the loss, but here's hoping what we'll get is a truly functional system instead of some gimped version of the design that was already present that doesn't actually meaningfully sort things out. After all this time and with this backtracking, I expect to finally be able to get to the year 3000 on a huge galaxy without the single-core calculation bloat that would cripple every super computer on Earth.

My expectations are no less. Good luck with the redesign.
Click to expand...
It isn't single-core. That's a pure falsehood that has been spread and debunked for years made by people who wanted to seem important with "the secret to why it won't run well!"

3000 is probably unrealistic, but 2600 is a reasonable expectation both on hardware limitations and gameplay. The default midgame is 2300, the default endgame is 2400, it should be able to run smoothly while you finish all the actual content of the game which should be long since over by 2600 even on all-crisis with a relatively unoptimized player-end, or they should generally have died to the crisis by then. It doesn't need to keep running through every possible part of the map being colonized to the max because the game is designed to end long before that (pop growth changes pending, that may no longer be true?), but it does need to run well up to the natural end point (which, again, would seem to me to be definitively over by 2600, if not significantly before).
 
  • 2Like
Reactions:
Thiend said:
It isn't single-core. That's a pure falsehood that has been spread and debunked for years made by people who wanted to seem important with "the secret to why it won't run well!"

3000 is probably unrealistic, but 2600 is a reasonable expectation both on hardware limitations and gameplay. The default midgame is 2300, the default endgame is 2400, it should be able to run smoothly while you finish all the actual content of the game which should be long since over by 2600 even on all-crisis with a relatively unoptimized player-end, or they should generally have died to the crisis by then. It doesn't need to keep running through every possible part of the map being colonized to the max because the game is designed to end long before that (pop growth changes pending, that may no longer be true?), but it does need to run well up to the natural end point (which, again, would seem to me to be definitively over by 2600, if not significantly before).
Click to expand...
I never said Stellaris is "single-core" but Stellaris does have a tremendous amount of sequential calculations present which are running on one core and that is what degrades its performance the most. That is easily observable by anyone who knows how to use sth as simple as a task manager and what's more the developers themselves have just said so in this very dev diary. In other words, my light hyperbole contains a kernel of truth that is far from "debunked".

As for what my expectations are, after all this investment, all this waiting, they are what I have stated. Why? Because I do want to add a few mods, I do want the engine to be able to handle comfortably what it is supposed to handle comfortably and have the leeway to push a bit further. I don't want it to just barely be good enough. That's just not enough anymore for me to be happy with the game. That's my expectations, you are welcome to have different ones.
 
Alastor said:
I never said Stellaris is "single-core" but Stellaris does have a tremendous amount of sequential calculations present which are running on one core and that is what degrades its performance the most. That is easily observable by anyone who knows how to use sth as simple as a task manager and what's more the developers themselves have just said so in this very dev diary. In other words, my light hyperbole contains a kernel of truth that is far from "debunked".

As for what my expectations are, after all this investment, all this waiting, they are what I have stated. Why? Because I do want to add a few mods, I do want the engine to be able to handle comfortably what it is supposed to handle comfortably and have the leeway to push a bit further. I don't want it to just barely be good enough. That's just not enough anymore for me to be happy with the game. That's my expectations, you are welcome to have different ones.
Click to expand...
I doubt performance improvements are going to let you run the game for literally over double the timeframe the game's actual content runs for on default settings, particularly when so many people have their games go significantly faster than those default settings.

I would by no means be disappointed if the performance changes were that significant, but I think you're setting yourself up for disappointment declaring in advance that it isn't enough for you unless it's lag-free for 800 years in a 300 year game people often play as a 200 year game.

You're expecting to be free of any single-core calculation blockages, and that just doesn't appear to be what they're saying is happening (nor am I sure it's possible). They're optimizing the bottleneck, but its not going anywhere. Saying you "expect to finally be able to do (thing they haven't indicated you'll be able to do and would be an absolutely incredible and mostly pointless performance leap)" isn't going to make it happen, but it is going to disappoint you if it doesn't.
 
  • 2
  • 1Like
Reactions:
I am so happy thr timeliness is coming! My favorite analy is when you ger the immortal character. Just fun to have one from the start.

Hopefully this comes up in the fleets, but the army builder is hard to use if you have a diverse population. I would like it to show thr most powerful first or have dried owns that I can select a particular species of my thralls

Also, random leaders are nice, but I miss being able to get a general. Most commander traits are, understand ly, fir admirals. I like having generals with my armies (I am not an optimizer, I like role-playing my empires).
 
Tannhäuser Cake said:
A MoO2-similar system for Stellaris could look something like this:
  • Trade Value is considered the analogue of freighter fleets.
  • Trade Value remains a non-stored resource, as currently and similar to Naval Capacity (i.e. the current production is the current capacity).
    • Trade Value could also be renamed Trade Capacity, emphasising its conceptual similarity to its militaristic counterpart, Naval Capacity.
  • Worlds that run local deficits can automatically import resources from the empire's global reserves, at a cost to its global Trade Value.
    • Local deficits, that are not covered by imports, result in negative local consequences.
  • Trade Value can be used to buy resources from the galactic market.
    • To the global reserve, that is. Keep in mind that subsequent local imports also cost Trade Value, as per above.
    • Edit: Trade Value can presumably also be gained from selling resources on the galactic market (minus the market tax rate). This can be thought of as a trade-weak empire using export earnings to pay the freighter fleet to move resources internally.
  • Unused Trade Value is converted to other resources at the end of the month, similar to how trade policies currently work.
Click to expand...
I was going to suggest something like this but this post has covered it in more detail than I would have.

Only thing I'd add is unlockable trade policies could still exist and unlock the ability to convert that excess trade value in to various resources. As trade value can be converted to "normal" resources at the market anyway to make this not redundant the policies would unlock the conversion of unused trade value to resources you can't trade for - research, unity, naval cap (fitting for a mercenary megacorp), etc.
 
  • 2Like
Reactions:
You must log in or register to reply here.
Top Bottom
Back