• 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 Dev Diary #314 - Pi in the Sky Ideas

Hi everyone!

It’s been exactly 100 dev diaries since we introduced the Custodian initiative alongside the Lem update. With that milestone met, and 314 being funny math number, I thought it would be a good time to review what the Custodians have accomplished, some of the process they use, and where we could go from here.

To review, the original idea behind the Custodians Initiative was to do some of the following:
  • Tweaking game balance
  • Adding new content to old DLC
  • Polishing existing content
  • Bug fixes
  • Performance improvements
  • AI improvements
  • Multiplayer stability
  • UI and quality-of-life improvements

So what have they accomplished so far?​

In my view, the Custodians have been a pretty solid success.

We’ve done a few things:









I promised bad math jokes last week: Why did the Blorg cross the Mobius Strip? To get to the same side.

So how do the Custodians work?​

Ideally, a third of their tasks are dictated by my directives, a third are taken from community requests, and the remaining third are individual developer passion projects. In reality, these segments overlap quite a bit, since we all share many of the same desires.

Before each release, we hold a “Custodian Pitch” meeting, where everybody on the team can write up proposals, give a two minute presentation on what they want to do and why they want to do it, and then a brief but spirited discussion is held debating the merits of the pitch as well as highlighting any concerns or suggestions.


Photo from a recent Custodian Pitch Meeting

Accepted pitches go onto a team board, get prioritized, and eventually end up on the schedule. Unlike tasks slated for DLCs, Custodian tasks are intended to be able to slip to a later release if they’re deemed not ready to go live yet, and sometimes accepted pitches wait for a while before getting worked on. Rejected pitches sometimes resurface at a later date, revamped to resolve whatever problems the original pitch had.


Part of the Custodian Team Board

As you can see, far more things get pitched and approved than we can actually do, so prioritization is extremely important. More things will move into 3.10 and 3.11 over time.

The Development Oubliette off to the side is for things that we started working on, but had to pause for various reasons.

Following Overlord, we instituted a rule that expansion work cannot absolutely depend on Custodian work, since we intentionally want Custodian work to be able to slip to a different release if necessary - the Situations system came in a little hot and that had negative effects on some of the systems we planned on using them with.

So what are some of those notes?​

From the “In Progress” section under 3.9 Caelum, we have the Diplomacy and Trust changes that we’ve mentioned a few times. We’ll have a dev diary dedicated to this next week, since these have been accelerated to be included in the 3.9.3 release.

Two weeks from now we’ll be providing more details on the Leader Consolidation that we’ve also talked about in the past. Currently these tasks are in the To Do and In Progress sections of 3.10 Pyxis, grouped by dark blue dashed lines that you can barely see in this image.

Other things on the board without planned release dates include a variety of possibilities. Some are larger scope tasks that involve a large number of Custodians, while others might be the type of thing that can be knocked out quickly.

A handful of them include:
  • Espionage enhancements
    • Make Espionage more impactful, but have systems in place to prevent “dog-piling”.
  • Continued work on concepts and nested tooltips
    • Provide information in a clearer way, avoiding walls of text.
  • Improve tutorials
    • They’re really not good at introducing the concepts new players need to learn the game.
  • Improve the outliner
    • Make the Outliner easier to use during different phases of the game.
  • Do somethingwith the Megastructures UI
    • As the number of large space constructions grows, it gets harder to find the ones you want to build.
    • It’s also very confusing to have things that require Mega-Engineering alongside constructions that don’t.
  • Continue after losing
    • Pop up the Select an Empire UI from Multiplayer if you lose a Single Player game, in case you want to continue your galaxy’s story, even if your first empire’s has ended.
  • Improve species modification
    • Address micromanagement and tedium in species modification.
  • Pop performance
    • Pops are one of the greatest sources of late game performance issues. Find ways to reduce their impact.
  • Ship performance
    • Ships (and fleets) are another performance issue to investigate.
  • Investigate number of habitable worlds (and rebalance if necessary)
    • As more scripted systems get added to the game, the habitable worlds slider becomes less accurate.
  • Make AI personalities matter more
    • Review the existing personalities, make them show up in AI weights more often, and differentiate them more.

Next week​

Having come full circle, next week will be about Diplomacy and Trust.

See you then!
  • 111Like
  • 40Love
  • 10
  • 1Haha
  • 1
The custodians really have done a great job! I would love to see more faction interactions, more factions with specific goals, and more danger if they become unhappy, with a greater chance of rebellions.
  • 5Like
  • 2
ship performance? = making fleet fights actually make more sense and slowed down too see whats going on and allow more tactics and strategy perhaps? .... cos atm its all just a blur and its unclear if combat computers are actually effective? most fights just look like a swarm of ships together like a school of fish.. fighter spam doesnt help either with visuals..

also any chance of weapon/mod rebalance as a whole.. especially event weapons they just feel awful too have as a research project when 90% of them are just trash tier... i even did a thread on them with 'unique' traits ideas .
  • 4
  • 1
The Custodian team is arguably the best thing to ever happen to any Paradox game, and has made my continued support of this excellent game feel very valued. Keep up the great work and thanks everyone on the team for doing such a marvelous job.
  • 34Like
  • 15
I love that the outliner and species modification micromanagement are getting some work!

I suggest two things that I’ve wanted for years, since these topics are on the table:

1. Simply increase outliner scroll speed (or give a slider option in settings)—in the lategame you have to click and drag the scroll bar; scrolling with the mouse wheel is too slow.

2. Allow us to de-modify cyborg pops (gained from refugees or conquered planets). You can make genetically ascended pops into cyborgs if you’ve taken the mechanical ascension perk, so why can’t you do the obverse if you’ve taken the genetic ascension perk? This would provide a lot more flexibility for roleplay, especially for spiritualist empires, plus it makes logical sense that the mechanical implants can be physically removed.

As always, thank you to the Custodian team!
Last edited:
  • 11
  • 2Like
I know it isn't uncommon for Game Devs to talk with Mod Devs and see about pulling ideas from them to add into the game. I feel like for the "Do something with Megastructures." you may want to consider reaching out to the person behind the Gigastructural Engineering & More Mod if you haven't already. There's bound to be some stuff in there that could likely be made into a proper Megastructure in the game, and if you want to keep all that as mod shenanigans, it might just give you some inspiration.
  • 3
The Development Oubliette off to the side is for things that we started working on, but had to pause for various reasons.

Be honest, the Development Oubliette is just another name for Alfray's office, isn't it? :p

Also after last weeks threats, I'm simultaneously disappointed and relieved by the almost complete absence of math puns.
Last edited:
Pop Performance in the endgame would be sooooooooo great ❤️ personally I still think the current mechanic about pops is generically broken, also contentwise. But any change here would be heavily appreciated.
And species mod ❤️ another of my pain in the [censored] topics.
  • 2Like
I for one cannot wait for the improved species modifications. Its fun the first few times you do it in a games, but then it just becomes tedious.
The worst is when you didn't modify an entire species, and then one unmodified pop wanders into your empire

edit- maybe we could have an option in the species menu for sub-species to automatically modify into a selected sub-species (similar to assimilation)
  • 13
  • 2Like
I love that a third of ideas come from developer pitches. Shows how much the devs are players as well as devs.

[*]Improve the outliner
  • Make the Outliner easier to use during different phases of the game.

Omg yes please! Can you consider making fleets sortable also so that it's not so cluttered? Having fleets randomly sort is a huge pain for linear naming schemes and for when you start getting fed/GDF fleets.

Address micromanagement and tedium in species modification.

Another awesome one. I think it's a weird holdover that pop modding requires you to click the population to be modded which starts a time-gated society research project. Feels like the situation system would be much more appropriate.

Being able to set templates as preferred and have pops automod to that (akin to automigration with policy and building modifiers, e.g. chance per month increased by gene labs) would be great. And please let us collapse templates in the species tab. The AI goes nuts with modification so for a xenophilic empire it's not unusual to have dozens of variants per species which takes up a lot of screen space when scrolling.
  • 5Like
  • 1
I love that a third of ideas come from developer pitches. Shows how much the devs are players as well as devs.

This is probably the best part of the Custodian team. Since so many of the developers are fans of the game themselves, and interact frequently with the community, the pitches are super well aligned.
  • 39Like
  • 4Love
The Custodian team is one of the best things implemented by any PDX developed game, and I really hope it will be added to the newer games as well.
Thanks for showing what they've done and a bit more insight in the process!
  • 15
  • 5Like
Paradox. Stellaris is an incredible video game and I love it, I just really enjoy this great work of video game and it really shows that you have invested a lot of time in its development and improvement, thank you for that. However, I would like them to add the much requested robotic ships for the robotic empires and organic ships for the hive empires (in the style of the Flood from the Halo series or similar). And I really liked the Paragons DLC, it was really necessary for more depth in the game, but what I didn't really like was that we can level up the leaders at our own will, this doesn't seem natural, for example a Admiral, we can level it up and we choose whether it is aggressive or tactical, but really, aren't they supposed to be the ones who improve of their own free will? (Similar to the Crusaders Kings games) It would be good if they go to battles as an admiral because they themselves level up to tactical or aggressive and the same for the other leaders. That's it and thanks for Stellaris I really love this game. :D
  • 5
  • 1Love