• 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 #154 - UX Design in Stellaris

Hello all, this is Hildor Anduv, the UX Designer on Stellaris. I’ve been with the team since December, and have really enjoyed working on such a great game. With my first Dev Diary I’m going to be sharing how I have been approaching UX Design on Stellaris.

Disclaimer: We're not quite ready to announce our plans just yet (we're saving that for the big stage, PDXCon), so this is going to be more process oriented without delving into specifics about any new feature. If that's not your thing, you can safely give it a miss, but if you're interested in the nuts and bolts of how games are designed and made, let us lift the curtain a bit.

Whiteboarding: The UX Corner
Whiteboards are a great place for getting your ideas out quickly. When whiteboarding I try to get team members from multiple disciplines and rough out ideas, together. It’s a great way to discover gameplay and player needs, as well as find any potential development pitfalls early on.

We wanted to create a permanent space for this, and thus the UX Corner was born. Here we come together and iterate on ideas in real-time as a group. Having the permanent space has helped facilitate conversion and increased cross-discipline communication within the team,.


ijod8Lid9WhNsBPs1t7irsX6Oj2eStn5GmfQrAEOYW6DejLrt5yeDzY58pR-y6JQIRkYDx8i6gfnXtx9bLJvETvV87DwKzP9kH-SfaWvsKcvRls1zffzlpAQdZGAZneSeGyeF_co



Wireframing & Prototyping.
From the whiteboard sketches, I then move onto creating more detailed wireframes, the skeletal mockup of a UI design. Keeping your wireframes simple lets you iterate quickly, but more importantly allows you to nail down what content should be displayed, and how best to lay it out visually for the player.

I typically try to nail down one key screen, and from there I build out a User flow of the entire experience.

When Designing the Archeology UI for Ancient Relics, I had some key pillars I wanted to nail.

  • The stories are linked, and I can revisit past chapters for context. (Also experience the great artwork again!)
  • When I come back here, I have a sense of the progress my scientist has made.
  • The challenge of the site feels ‘game-y’.

Once I nailed an overall layout for the Archaeology UI, I then broke it out into a User Flow to capture the full breadth of a player’s journey while excavating a site.


HXhqPAfmMvuAI8JXM6Zuu0FlVFKfbu4lVxH7nJwOBJ_FRGbEz1YGR-8OdqJHh1q6rNNq6EYb4DcMhrtyTOTxCqraq7f_dRRdMnBFZxxcdHnDakrXXe-dIxU-BKPzK7CsO12AxtT5



With the User flow complete I then link it up and create an interactive Prototype. The prototype allows team members to experience what we are trying to create before we set out to develop it. It’s a great low cost way to identify issues in the flow, and fix them here before investing in development.


IomaaXyi4recbbyNf_jM2z1tGRjrD0OkkWiG08OfKh4lSaN3qVh4I7Y1oS3F4-WtCcBbCULNa9FIMxd49JddaKcRzABnQWzfXoCInqzipkHan_HioASIhQ4djhmRvTlQQ-sG0TKM



After we agree on the UX design, we then get to finally create it in game. As you can see below, the design mostly held through to release. The text window was increased to accommodate larger story text, and ‘difficulty’ was changed to ‘breakthrough chance’ to help better understand the ‘die roll mechanic’.


ullUaf8HKlwVIzUWNbKN-J3_m7AAyzmTal-6C_AnRhBaVNB4rEQx8NbYItxI9ivkRi9n8yhxMdYyTDM1bgdel1ocmo97efIGZZZFxAIOBjgy4ObtkyNK20jUUdS2CXIzq-zUnEhZ



User Testing
So, even though the design has now gone through many gates, we must test our designs with players and iterate based on player feedback. Luckily Paradox has a great User Research department, and so we were able to run playtests and get vital feedback which we could respond to before launch.

----------

Well, I hope this was an informative window into how UX Design is done here on Stellaris.

Also, I read the forums and look out for your feedback as well. You all have been very helpful, so thank you for that!
 
  • 1
Reactions:
i sincerely hope so, beats paying for churned out window dressing that cant quite hide the core inadequacies.
If you believe the problems in Stellaris are insurmountable, and that these problems are significant enough to make the game unenjoyable, then why do you buy the DLC? Do you believe that buying DLC for a game will encourage Paradox to work on a sequel?
 
So.. help me to understand this. You're basically saying that, one of you put the idea of "Move the Building throgh the slots (and the planet itself so) like K.Kardashian move her B-side" on that Whiteboarding, and when you saw it (help me here really hard), you didn't laugh out loud... but instead you've discussed it (guess maybe even for days couse that wasn't like that at the begining) and then after all the talk you've really(really guys, really?!) decided that this garbage we have now, was the best path to follow? Couldn't you just leave the slot priority like we have on Starbases? I mean, personally, if i build a building in a specific slot is couse i want to have all checked out at the first glance, i don't want to play to hide and seek if i'm playing stellaris. Just.. how peculiar.. xD

OFC take this comment lightly, i'm just exaggerating to make things funnier... but seriusly guys, why?
 
If you believe the problems in Stellaris are insurmountable, and that these problems are significant enough to make the game unenjoyable, then why do you buy the DLC? Do you believe that buying DLC for a game will encourage Paradox to work on a sequel?

Honestly? - Misplaced hope i think, i do really love the idea of Stellaris, and i always just felt that one more patch or one more DLC might address the issues, or at least improve them, but the AI has never recovered from apocalypse, and got worse with megacorp. and performance flits between "slightly better" and "much worse" with every patch.

Each patch seems to re-break basic shit with the game. i think we've had 6 months of reliable crisis in the last 2 years? other than that they are somehow bugged.

I recently got HOI to play with my friends, the second paradox game i've actually tried to give time to, and it made me realise that paradox are great at ideas, but horrible at implementation.
 
it made me realise that paradox are great at ideas, but horrible at implementation.
This is actually so true.

The issue with Stellaris's economy is absolutely insurmountable. The assignment problem of allocating the correct job to the correct task has an np-hard solution, to my, albeit naive, knowledge, the most effective solution method is an exhaustive branch and bound search. To me it speaks that the team in charge of implementation and the team in charge of design are two separate groups.
 
I
To me it speaks that the team in charge of implementation and the team in charge of design are two separate groups.

Yeah it feels like when the mil/eco reworks were being done, there wasnt someone going "yeah our AI will never be able to handle this, please rethink it to make it more friendly".

There are several parts of this game that suggest the design idea exceeded the resources and/or expertise present to implement them.
 
This is actually so true.

The issue with Stellaris's economy is absolutely insurmountable. The assignment problem of allocating the correct job to the correct task has an np-hard solution, to my, albeit naive, knowledge, the most effective solution method is an exhaustive branch and bound search. To me it speaks that the team in charge of implementation and the team in charge of design are two separate groups.

Since jobs' output is multi-dimensional, not only is allocation np-complete, but optimal requires additional information that aren't collected/used at this time.

Take the simple case where a colony has 30 pop composed of 9 pop of A and 21 of B.. Pop A is +1 at mining and pop B is -1 at food production. There are 12 food jobs, 6 mining jobs, and 12 other jobs. What is the optimal placement? How does the answer change if you are told the empire is running a surplus of minerals, but is short food?
 
Last edited:
The assignment problem of allocating the correct job to the correct task has an np-hard solution.
Why on earth would you require an algorithm that gives the optimal result? You can make some very quick heuristic algorithms that will deliver a result from the "good enough" category. In the very simple cases where a human can immediately see how the optimal distribution would look like it would propably give the optimal solution anyway.
 
Why on earth would you require an algorithm that gives the optimal result? You can make some very quick heuristic algorithms that will deliver a result from the "good enough" category. In the very simple cases where a human can immediately see how the optimal distribution would look like it would propably give the optimal solution anyway.
Because humans, being what they are, will look at a good-enough solution and say to themselves "I can do better than those foolish devs! IThat pop should be here amd this pop should be there!"

And then they start hunting for the manual allocation button. That doesn't exist.
 
Why on earth would you require an algorithm that gives the optimal result? You can make some very quick heuristic algorithms that will deliver a result from the "good enough" category. In the very simple cases where a human can immediately see how the optimal distribution would look like it would propably give the optimal solution anyway.
This. The ai mustn't play optimal, but competitive.
 
Hi all, I'm Jamor and I'm the producer on Stellaris. I run the team's time and budget, the boring biz bit, while grekulf is the creative and design lead.

I publicly acknowledge there are AI and performance issues in the game right now.

I want you to know that maintaining things like AI and performance is a big and neverending struggle, especially on a game that is growing as large as Stellaris has. There's no one-button magic bullet solution that will fix it overnight. In some patches it is stronger, in others it lags behind. I have to follow my agreed release schedule and juggle the not-unlimited time and personnel I have to deliver that, but I can tell you that whatever I can possibly spare goes in to bug fixes and improvements. We're Paradox, we don't release a game and patch it twice and call it a day, we are in it for the full marathon. I was a PDX fan for a decade before I worked here (HOI 2 and onward) and maintaining that long term commitment is my personal guarantee to you, as a fellow fan.

The reason there's been no public news about fixes or improvements on AI and performance is simple: there are none yet. I'm not going to pull the wool over your eyes and promise something until it's done and dusted and in my hand to give. That's not my style. When we have it, you'll see it in a patch note. Until then I don't want to string people along with false optimism and empty promises.

About the process focused dev diaries: we don't want to go completely radio silent in those times before official announcements. Right now, we have our biggest max visibility PR event on the calendar coming up, PDXCon, and we naturally want to get the most eyes on us when we reveal our future plans. We need to get the word out with highest public impact so we can keep supporting this game and others, for ever. So, in moments like these, dev diaries and official communication will have to be about other things that won't spoil our planned reveal. Process stuff, an under the hood glimpse at how games are made, is interesting to some, but not all. If you don't dig it, just please bear with us for the official announcement and then you'll get a ton of in-depth feature DDs after that. The alternative is for us to just go completely quiet during those pre-announce high intensity dev periods, and I don't think anyone wants that.

We're here, we're working every day to make this thing better for you, our fans and supporters. We tell you as much as we can consistent with the needs of basic business security, and really prize our unusually close contact with our community, certainly unprecedented for a developer of our size. We want to keep that, so please be decent to my staff, and each other. Thanks for sticking with us through this journey, and there's much more to come.
 
About the process focused dev diaries: we don't want to go completely radio silent in those times before official announcements. Right now, we have our biggest max visibility PR event on the calendar coming up, PDXCon, and we naturally want to get the most eyes on us when we reveal our future plans. We need to get the word out with highest public impact so we can keep supporting this game and others, for ever.
I honestly don't get how people don't understand that. The annual BIG REVEAL event is coming up. Dev Diaries are going to be all about "look at the cool nifty stuff" right now and nothing about content or even really bug fixes because a bug fix Dev Diary would get people screaming "WHEN IS THIS PATCH COMING OUT?!111"

The choice is fluff diaries or no diaries. Give me the fluff. Sometimes it'll be fluff I find interesting, sometimes it won't.
 
Stellaris has a great UI, though there is always something that can be improved. I created the designs for the updated Planet View, for example. I'm on the look out for community feedback on features, though. You all help give direction on what to improve, and how.


Something that would actually improve the game was automatic migration working properly for everyone instead of having to resettle pops manually in the lategame. We should get a better UI for resettlement where you decide on where you want the pops to move (Ringworlds most of the time). You could create some kind of chart that shows arrows and gives clear visual information where your pop growth is going. In the lategame, its obviously going from your fully developed worlds into your Ringworlds. Since you guys have not provided us with automatic resettlement, we are left with doing this manually. Managing 100 planets and having to manually resettle pops away from 30 of them onto several Ringworlds is a very tedious task and takes away any enjoyment of the game. Because at this point players are constantly getting bombarded by unemployment events since you can't resettle pops fast enough.

Not everyone might play up to 100 planets, but it does happen. And a game thats supposed to call itself "grand strategy" shouldn't prevent players from getting to the "grand" part just because its unbearable to manage an empire with 100 planets.
 
Last edited:
OP, thanks for the peak under the hood. I specialised in UI/UX design and development many years ago, but haven't had any recent experience in the industry with it. So it's good to know that proven methodology hasn't been chucked in the bin. I do think ClauseWitz is limiting better UI flow, but I can see from recent additions and changes to Stellaris' UI that you're always trying to extend it.

@ evry1 asking "where's my performance/AI fixes?" Here:
EDAI https://steamcommunity.com/sharedfiles/filedetails/?id=1805681002

It does change game-state quite heavily and isn't achievement compatible, but I hope @Jamor will have the team look at it. I've also relayed a bunch of actual AI fixes in it to PDS via a new company hire who is a colleague and I hope some of those will make it into an update sooner rather than later.
 
Hi all, I'm Jamor and I'm the producer on Stellaris. I run the team's time and budget, the boring biz bit, while grekulf is the creative and design lead.

I publicly acknowledge there are AI and performance issues in the game right now.

I want you to know that maintaining things like AI and performance is a big and neverending struggle, especially on a game that is growing as large as Stellaris has. There's no one-button magic bullet solution that will fix it overnight. In some patches it is stronger, in others it lags behind. I have to follow my agreed release schedule and juggle the not-unlimited time and personnel I have to deliver that, but I can tell you that whatever I can possibly spare goes in to bug fixes and improvements. We're Paradox, we don't release a game and patch it twice and call it a day, we are in it for the full marathon. I was a PDX fan for a decade before I worked here (HOI 2 and onward) and maintaining that long term commitment is my personal guarantee to you, as a fellow fan.

The reason there's been no public news about fixes or improvements on AI and performance is simple: there are none yet. I'm not going to pull the wool over your eyes and promise something until it's done and dusted and in my hand to give. That's not my style. When we have it, you'll see it in a patch note. Until then I don't want to string people along with false optimism and empty promises.

About the process focused dev diaries: we don't want to go completely radio silent in those times before official announcements. Right now, we have our biggest max visibility PR event on the calendar coming up, PDXCon, and we naturally want to get the most eyes on us when we reveal our future plans. We need to get the word out with highest public impact so we can keep supporting this game and others, for ever. So, in moments like these, dev diaries and official communication will have to be about other things that won't spoil our planned reveal. Process stuff, an under the hood glimpse at how games are made, is interesting to some, but not all. If you don't dig it, just please bear with us for the official announcement and then you'll get a ton of in-depth feature DDs after that. The alternative is for us to just go completely quiet during those pre-announce high intensity dev periods, and I don't think anyone wants that.

We're here, we're working every day to make this thing better for you, our fans and supporters. We tell you as much as we can consistent with the needs of basic business security, and really prize our unusually close contact with our community, certainly unprecedented for a developer of our size. We want to keep that, so please be decent to my staff, and each other. Thanks for sticking with us through this journey, and there's much more to come.
I'm glad to see that the performance issues are being publicly acknowledged.

The reason that I have been pushing in DDs for acknowledgement that performance problem is being addressed is that this isn't a new problem. It isn't a 2.3.x problem. It started for me in 2.2.0 and has not noticeably improved in any patch that I have tried since that release. That is 9 months, 1 major release, and 10 patches (2.2.1 to 2.2.7 and 2.3.1 to 2.3.3) with zero improvement from my perspective.

I hope that you, the team, and other internal stakeholders put Stellaris performance as the highest priority, because at this point it is a long lingering problem.
 
@HildorAnduv One clear weakness that the Stellaris UI has is that many of it`s UI tooltips and pop-ups don`t offer you all the necessary information in themselves. For instance a leader pop-up might just say that a leader gained a level or a trait, but it doesn`t tell you that the leader X got the trait "logistician" for example which means exactly this in the game (+Y % to this) or that he/she/it achieved level 4 which means exactly this (+Z to something).

Also one great thing that older space 4X games like MOO 2 had was a general planets screen, which could be used to both viewing what each individual planet in your empire where producing or building at the time, but also to make changes to those things if needed. Someting with that functionality could make managin your planets require fewer clicks overall. As you could take a look at the big and smaller pictures simultaneously.

_kolonien.png


And as one final thing it would be great if the UI had a button, which could be used to send idle science vessels from the research projects screen staight to the target, instead having to exit the screen, search for the project in guestion on the map and then search for an idle or nearby science ship to do it.

Hi all, I'm Jamor and I'm the producer on Stellaris. I run the team's time and budget, the boring biz bit, while grekulf is the creative and design lead.

I publicly acknowledge there are AI and performance issues in the game right now.

I want you to know that maintaining things like AI and performance is a big and neverending struggle, especially on a game that is growing as large as Stellaris has. There's no one-button magic bullet solution that will fix it overnight. In some patches it is stronger, in others it lags behind. I have to follow my agreed release schedule and juggle the not-unlimited time and personnel I have to deliver that, but I can tell you that whatever I can possibly spare goes in to bug fixes and improvements. We're Paradox, we don't release a game and patch it twice and call it a day, we are in it for the full marathon. I was a PDX fan for a decade before I worked here (HOI 2 and onward) and maintaining that long term commitment is my personal guarantee to you, as a fellow fan.

The reason there's been no public news about fixes or improvements on AI and performance is simple: there are none yet. I'm not going to pull the wool over your eyes and promise something until it's done and dusted and in my hand to give. That's not my style. When we have it, you'll see it in a patch note. Until then I don't want to string people along with false optimism and empty promises.

This is good to hear, but I hope people even higher up than you realise that you cannot sell new DLCs endlessly to a base game that has big unsolved problems. Because if those problems aren`t addressed properly sooner or later you will start to loose the player base that buys those DLCs.
 
Last edited:
Well I'll try keep things on topic for interface stuff though gameplay will likely bleed in.

Things we could use:
A way to assemble armies similar to the fleet manager. It's hard to keep track of how many armies you need when you're building them across many worlds, so it would be convenient to have a screen where you can select what you want built and where to gather and have the game start building wherever they can get there the fastest.

Resettlement details.
There's some inconsistencies when a planet has slaves and jobs slaves can't have, showing open employment when there isn't any, or robots and biotrophies which have two very different types of employment and no overlap.
Could also use a new way to move pops where you select a pop and the game shows you places with positive housing, amenities, and available jobs, ordered by habitability.

And is anything in the works regarding pops and the jobs they have, possibly getting a means to assign them? We get all these genetic tools to make pops good at jobs but they're not consistent in getting those jobs.