Hello everyone and welcome to another Stellaris development diary. Today, we're going to continue talking about the Distant Stars Story Pack that will be accompanying the 2.1 'Niven' update, on the topic of exploration and new things to find in the galaxy. As with the previous dev diary, we're going to be fairly light on the details to avoid spoilering too much of what's out there to explore in Distant Stars.

New Anomalies
Distant Stars adds several dozen new anomalies to the game, with everything from simple discoveries to complex event chains (including the opportunity to make a new and very different kind of friend...) Overall, Distant Stars increase the anomaly count of the game by about 50%. We've also taken the time to go back and improve some old anomalies and fix up others that did not spawn correctly, so that even those without Distant Stars will be able to experience new discoveries in the Niven update.

New Systems
Distant Stars adds around 20 new, unique systems to the galaxy, similar to Sanctuary or Zanaam in the base game. Most of these systems have unique encounters, event chains or anomalies related to them, and some are actually several systems with a common background or event chain tying them together. Overall, though none of these systems are guaranteed to spawn in any individual game, every randomly generated galaxy in Distant Stars should contain some new and potentially rewarding discoveries for your empire to make.

New Leviathans
Distant Stars also adds 3 new Leviathans to the galaxy. Two of these Leviathans work in a similar way to existing ones, while one is altogether different from anything you will have seen before. As with previous Leviathans in the Leviathan Story Pack, these new Leviathans have curator dialogue and interactions related to them, which means that the Curator Enclave will appear in the galaxy if you have Distant Stars but do not own Leviathans (Traders and Artists do not spawn unless you have Leviathans, however).

That's all for today! Next week we're going to talk about the Niven update, on the topic of Space Creatures and Strategic Resources.

 

[I am Goot]

@TheDungeon

And the name of your mod is?

 

[117Killer]

I have an idea for a paid DLC, it would add a feature we all are really lacking and most of us really would love to have in the game: Proper AI

Patch Name: Mühlen-Schulte

Features:

• Working Sector AI with tons of configuration options for specialisation, optimisation and automation including sector defense budgets, Space Station focus, Ship construction based on what has been built in the fleet manager
• New carefully rewritten AI with capable planet management, fleet analytics and intel based fleet composition for whoever it fancies to go after, highly capable diplomatic ai that crafts alliances, federations based on geographical and relation based factors.
• Battle hardened AI based upon pro gamers input how to play properly and how to avoid being a shamefull mess
• Proper Federation management including votes, fleet management, fleet composition, budget management throughout several involved federation ai's
• Carefully crafted strategies for all kinds of different ethics, civics and race compositions which take into account starting locations and surrounding aswell as neighbours and their preferences and setup
• and a ton of new scripted and dynamic ai events and triggers to make it more dynamic and fun...

(if you find some irony in here you can keep it ;-) )

if paradox tried to charge me, or anyone else sensible on this forum for that dlc, they would find themselves under a consumer siege. making me pay for a fix of their beyond terrible AI, id revolt, and so would almost everyone here. that's a terrible idea, they need to fix the AI, for everyone, for free. cause as it stands the AI is braindead. and your game can be interesting and pretty as all hell, but when the challenge in the game is smart as a box of bricks, then something needs fixing. the reason they won't fix it and haven't, is because they know a dlc like this would cause an uproar, and so, since they can't charge us for it, they won't do it. not unless we keep bringing it to their attention like many people are always trying to do. kick up enough fuss, they have to fix it, able to charge for it or not.

 

[TheDungen]

@TheDungeon

And the name of your mod is?

My fixes, I only mod for myself.

 

[dosam01]

I'm really really looking forward to this new expansion pack coming out and as soon as it comes out I will clearly call all my friends to recreate new races and play a game! please do not stop the development! continue adding more content like for ex : complex systems of diplomaties etc.!

 

[thekaje]

if paradox tried to charge me, or anyone else sensible on this forum for that dlc, they would find themselves under a consumer siege. making me pay for a fix of their beyond terrible AI, id revolt, and so would almost everyone here. that's a terrible idea, they need to fix the AI, for everyone, for free. cause as it stands the AI is braindead. and your game can be interesting and pretty as all hell, but when the challenge in the game is smart as a box of bricks, then something needs fixing. the reason they won't fix it and haven't, is because they know a dlc like this would cause an uproar, and so, since they can't charge us for it, they won't do it. not unless we keep bringing it to their attention like many people are always trying to do. kick up enough fuss, they have to fix it, able to charge for it or not.

Well, if you're not willing to pay for it, I think the chances are high they will never fix it. Johan himself said on Twitter that "as long as the game is selling, we're not doing anything wrong." Paradoxically, they will also abandon the game if it stops selling, so I think we're never gonna get those AI fixes.

 

[Eelectrica]

Wiz has already stated they are fixing AI bugs. I'm hopefully in the final 2.05 patch we'll see some good improvements.

Definately looking forward to an AI that can put up a fight.

 

[KingHoot]

if paradox tried to charge me, or anyone else sensible on this forum for that dlc, they would find themselves under a consumer siege. making me pay for a fix of their beyond terrible AI, id revolt, and so would almost everyone here. that's a terrible idea, they need to fix the AI, for everyone, for free. cause as it stands the AI is braindead. and your game can be interesting and pretty as all hell, but when the challenge in the game is smart as a box of bricks, then something needs fixing. the reason they won't fix it and haven't, is because they know a dlc like this would cause an uproar, and so, since they can't charge us for it, they won't do it. not unless we keep bringing it to their attention like many people are always trying to do. kick up enough fuss, they have to fix it, able to charge for it or not.

The problem is that too many people refuse to vote with their wallet. I haven't bought any DLC because I didn't think it was worth it (but I did get 2 free DLCs from the giveaway they did).

The AI needs to be fixed badly and should be priority one, all other priorities rescinded. It needs to be competent enough to present a challenge, even if that challenge only comes at the highest difficulty. But as has been mentioned, people will keep buying the "waffer-theen" DLC packs which only encourages the company to make more of them.

Also, these DLC packs, for two years, have added hardly no new systems. Only content. Content is all well and good, but if it's events, a few lines of text or just some research or structures you can build that don't really change the way the game plays out, then as far as I can see they're too vapid to be worth the coin. The new race and government types / Ascention perks are a good example of a system that is actually an interesting edition. If they added more entire new systems, (e.g. intergalactic trade routes / civilian traffic and domestic factions that need to be managed / an espionage system / a more in depth logistics and invasion system i.e. if you want to invade a planet you bloody well better have a supply line to it because invading a world should be a massive undertaking) then it would make the game far more interesting, because right now, especially with this weak AI, after the early game it's just boring as hell.

I mean all of these things, and so much more, could be done within the scope of the game. But what's actually being released is 50 new anomalies that I'll read the first time then click through every subsequent time for the +% bonus or whatever it gives me.

The amount of times I've not only posted but read other users post things like "Can we get the AI fixed as a priority before new content, please?" means that Paradox cannot fail to be aware of it.

But it comes to my mind that a large portion of people playing Stellaris LIKE the AI being broken. Either because it let's them feel "awesome" at a game they would otherwise be only "ok" at, or simply because they like just chilling and carving out an empire and painting the Galaxy their colour without any massive roadblocks (this is fine by the way, it's why "Easy" difficulty exists, I mean it can be very relaxing to play this way). I'm not disparaging any players here, I just feel like making the AI robust should be regarded with the utmost urgency and effort.

 

[BlackUmbrellas]

The AI needs to be fixed badly and should be priority one, all other priorities rescinded.

That's a completely ridiculous approach to take.

 

[KingHoot]

That's a completely ridiculous approach to take.

No it isn't. Your comment adds nothing to the discussion.
Stop straw-manning every post I make, please.

Thank you.

 

[AlanC9]

The sensible version of his objection is that the paid content pays the salaries of the devs working on the free patches Asking for just the free content isn't realistic.

 

[Murray Rothbard]

The sensible version of his objection is that the paid content pays the salaries of the devs working on the free patches Asking for just the free content isn't realistic.

And yet there are plenty of games that have been in EA or official release for far longer than Stellaris has been released, which have incrementally improved their systems including AI, and have done this for years without charging extra for anything. The price of the product may go up during EA, but no one is being asked for an extra dime after they've bought a copy. At PDX we are paying to license a (fully-revocable) copy of a game and then asked to make installments for extra content, and the AI never seems to improve.

Being upset about the state of the game is a perfectly acceptable position. Being frustrated should be taken as a sign that we want a great game that plays well. More content that is easy to draft and implement and earns revenue might be PDX's priority, but it's should not be the highest priority.

 

[BlackUmbrellas]

No it isn't. Your comment adds nothing to the discussion.
Stop straw-manning every post I make, please.

Thank you.

That's... not even a strawman argument...

Expecting the entire team to drop everything and focus solely on AI is both completely ridiculous and completely unrealistic. The fact you're so convinced it's either/or in regards to "fixing the AI" and "fixing other bugs/creating new content" just shows you haven't thought this through at all. They can do both! They are actively doing both!

 

[Peter Ebbesen]

I have an idea for a paid DLC, it would add a feature we all are really lacking and most of us really would love to have in the game: Proper AI

Patch Name: Mühlen-Schulte

Features:

• Working Sector AI with tons of configuration options for specialisation, optimisation and automation including sector defense budgets, Space Station focus, Ship construction based on what has been built in the fleet manager
• New carefully rewritten AI with capable planet management, fleet analytics and intel based fleet composition for whoever it fancies to go after, highly capable diplomatic ai that crafts alliances, federations based on geographical and relation based factors.
• Battle hardened AI based upon pro gamers input how to play properly and how to avoid being a shamefull mess
• Proper Federation management including votes, fleet management, fleet composition, budget management throughout several involved federation ai's
• Carefully crafted strategies for all kinds of different ethics, civics and race compositions which take into account starting locations and surrounding aswell as neighbours and their preferences and setup
• and a ton of new scripted and dynamic ai events and triggers to make it more dynamic and fun...

(if you find some irony in here you can keep it ;-) )

I would buy that in a heartbeat, irony or no.

Heck, I'd buy a DLC that said "we locked our AI developer in the dungeon and only released him two months later when the AI could provide a modest challenge on Grand Admiral, as verified through testing by expert players, and that is ALL you get in this DLC. And sure, it cheats to achieve that, but so what? The important thing is that the AI isn't a complete pushover and that when you beat it nevertheless and see its planets, you see that they are well developed. When you look at its expansion in the early game, it does so as rapidly as a good player if you play on a high difficulty. And when you fight it in war, it stands a chance. We've kept most of the cheats under the hood rather than in your face, because we understand that players prefer the illusion of competence creating challenge to the reality of incompetence creating none."

 

[117Killer]

I would buy that in a heartbeat, irony or no.

Heck, I'd buy a DLC that said "we locked our AI developer in the dungeon and only released him two months later when the AI could provide a modest challenge on Grand Admiral, as verified through testing by expert players, and that is ALL you get in this DLC. And sure, it cheats to achieve that, but so what? The important thing is that the AI isn't a complete pushover and that when you beat it nevertheless and see its planets, you see that they are well developed. When you look at its expansion in the early game, it does so as rapidly as a good player if you play on a high difficulty. And when you fight it in war, it stands a chance. We've kept most of the cheats under the hood rather than in your face, because we understand that players prefer the illusion of competence creating challenge to the reality of incompetence creating none."

I'd buy it.

And yet there are plenty of games that have been in EA or official release for far longer than Stellaris has been released, which have incrementally improved their systems including AI, and have done this for years without charging extra for anything. The price of the product may go up during EA, but no one is being asked for an extra dime after they've bought a copy. At PDX we are paying to license a (fully-revocable) copy of a game and then asked to make installments for extra content, and the AI never seems to improve.

Being upset about the state of the game is a perfectly acceptable position. Being frustrated should be taken as a sign that we want a great game that plays well. More content that is easy to draft and implement and earns revenue might be PDX's priority, but it's should not be the highest priority.

and this is the point. everyone seems to forget, stellaris is $40. it's not free to play with DLC, it's $40 bucks a pop. when I pay $40 for a product, there is a certain level of quality assurance you can expect. a somewhat capable AI is a part of that. the AI being braindead shouldn't come from DLC, cause this game has an upfront cost. and with that upfront cost it is expected that the vast majority of the games works up to a decent standard. and in 2 years i'd guess that less than half the games base systems work well or have been balanced and changed to fit what has been added.

the problem paradox always has is that, they never, ever focus on the base game, despite the fact that they always have a cost, the base game is always neglected. its why people say getting CK2 the game is only getting a quarter of the game. why EU4 looks like a different game when I see it played with all the DLC's. but at the end of the day, I payed $40 for a game. that game at bear minimum, has to function decently and provide some level of entertainment. and stellaris can do that, if your a player not seeking any kind of challenge and happy to ignore the braindead AI being incapable of even basic tasks. for the rest of us, i'll continue pointing out the flaws on the forums and adding my voice to those demanding a response from paradox.

to see just how bad this AI stuff is starting to get to people
https://forum.paradoxplaza.com/forum/index.php?threads/so-about-that-ai.1090227/

 

[BlackUmbrellas]

Paradox has consistently been improving on the AI for Stellaris.

2.0 was a setback, yes, but its not as if they've just given up. We'll see it get better.

 

[Jorrhast]

Paradox has consistently been improving on the AI for Stellaris.

And has consistently been failing to make it challenging, even at highest difficulty.

 

[BlackUmbrellas]

And has consistently been failing to make it challenging, even at highest difficulty.

Which isn't surprising in the least. AI for strategy games is hard to make challenging, because humans are so much better at it.

Challinging isn't and shouldn't be the issue. Competent is. The AI was competent enough in 1.9- 2.0 shuffled the board around mechanically, and the AI hasn't been made to catch up yet.

 

[Peter Ebbesen]

Paradox has consistently been improving on the AI for Stellaris.

2.0 was a setback, yes, but its not as if they've just given up. We'll see it get better.

Given that AI controlled enemies are now less of a threat to the player than they have were on launch back in 2016 or any time since, and given that this follows a simple and fairly predictable pattern of patches since launch, I think your "consistently been improving" needs to be taken with a grain of salt.

Say rather, "consistently working on and improving some areas of the AI (especially those directly in the player's face such as planet management) while overall enemy AI performance as a credible threat decreased", which would have the virtue of being the truth.

The problem here appears to be simple: Developing new features sells DLC so that's obviously a high priority in the post-launch development model. It has to be. Developing AI capable of playing well is hard and takes a lot of time and iterations with testing.

Time that the DLC development schedule just doesn't allow for. It is AI technical debt with a vengeance: You don't have the time to either satisfactorily address old issues or make new features work well enough (or ensure you don't by accident slightly break something that was already working) - and that's before time spent on tracking down and fixing outright AI bugs (new as well as old) is factored into the equation.

The result is what we've seen: With every feature/DLC patch the enemy AI performance drops, sometimes drastically (Banks/Utopia and Cherryh/Apocalypse). With every non-feature/DLC patch the enemy AI performance improves a bit. Thus an overall downwards trend.

And that's despite the occasional outright elimination of features the development team gave up on writing AI to handle and despite some old AI idiocies being eliminated along the way.

 

[BlackUmbrellas]

The result is what we've seen: With every feature/DLC patch the enemy AI performance drops, sometimes drastically (Banks/Utopia and Cherryh/Apocalypse). With every non-feature/DLC patch the enemy AI performance improves a bit. Thus an overall downwards trend.

Nonsense.

I remember the trend. It was upward. By 1.9, people were frequently remarking that the AI had gotten... pretty good, actually. The jump in competency was noticeable (IIRC, a big thing was that the AI was expanding way more aggressively).

I really don't understand how people have developed a sort of selective amnesia about this.

 

[Peter Ebbesen]

Nonsense.

I remember the trend. It was upward. By 1.9, people were frequently remarking that the AI had gotten... pretty good, actually. The jump in competency was noticeable (IIRC, a big thing was that the AI was expanding way more aggressively).

I really don't understand how people have developed a sort of selective amnesia about this.

Because they remember their own experiences, not "what people were frequently remarking in the forum?"

1.9 was certainly a high point compared to 1.5 in terms of enemy AI opposition, but not compared to 1.0, and the major cause of that can be traced to the force-multiplier that unity-traditions introduced with 1.5 - something that any competent player could abuse the hell out of but the AI had no clue how to use.

As one example of that (and not the only one), anybody competent playing the game in single player soon realized after 1.5 that they could effectively double their science output throughout the first century or so simply by rushing unity to open up a few discovery traditions and focusing heavily on science ships and keeping up exploring while they did whatever else they wanted to (gradually building up a huge stored surplus while doubling spent tech, then eating of that surplus for years to come once the player had explored the galaxy). So even if the player chose not to focus on science with regards to buildings or gaining techspeed modifiers and plowed resources into production infrastructure and ships as a priority, the player would stay competitive with all but the Fallen Empire AIs in terms of tech so long as the science ships were out exploring. The AI, even if it took the same traditions, would stick to one or two science ships and gain little for its efforts.

The result of that particular opening on the part of the player was that all enemy AIs fell behind the player even faster than before 1.5. So they nerfed the stored tech gain to 1/10th in a patch, and it was still not enough. Which is why that particular tradition was completely reworked in 2.0.

Compared to force-multipliers like that in the hands of a competent player which the AI can't even begin to exploit, something like "the AI expands somewhat more aggressively" is small beans that has little impact on the overall level of challenge the enemy AI players present to the player. And that's just traditions - don't get me started on the ascension perks and the synergies between particular tradition and ascension perk choices as part of a broader strategy.

(Now in 2.0 that particular opening gambit doesn't work any longer since Discovery now provides a tech-assisted unity rush, something that is very easy to abuse by focusing on low-cost techs and research agreements, but at least it isn't as badly broken for balance purposes as the exploration-assisted tech rush was and if the AI does end up with those traditions it can't help but benefit as it doesn't require it to do anything but research.)