• 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 #385 - AI Benchmarks

Hi everyone!

The 4.0.13 update released today with the following changes:

Stellaris 4.0.13 Patch​

Improvements​

  • Behemoth Fury is now available to Wilderness Empires.
  • Improved tooltips for the following civics:
    • Functional Architecture/Constructobot
    • Environmentalist
    • Astro-mining Drones
    • Maintenance Protocols
    • Ascensionists
    • Augmentation Bazaars
    • Brand Loyalty
    • Death Cults
    • Dimensional Worship

Balance​

  • Mutagenic Habitability now counts all planet types as ideal for upgrading Gaiaseeders
  • Dramatically increased the draw chance for the Mineral Purification, Global Energy Management and Food Processing technologies
  • Rebalanced the Pleasure Seeker civics to transform Civilians into Hedonists
  • Logistic Drones are now Complex Drones not Menial Drones

Bugfix​

  • Fixed invaded pre-ftls not becoming biotrophies
  • People once more die when they are put in the Lathe
  • Bio-Swarmer missiles can now be used by all biological ships with medium weapon slots (including defensive platforms)
  • Pops that are being pampered will now be forcibly switched to the correct living standard
  • Replacing a district specialisation no longer destroys CyberCreed buildings that should be kept
  • Corrected a tooltip bug where a planet would display itself as a possible migration target.
  • Fixed capitalisation for resources in trade policies
  • Updated assorted modifiers that still referred to Clerks
  • A Trade deficit now causes Job Efficiency and Empire Size issues
  • Fixed the tooltip for the Polymelic trait
  • Armies now protect 200 pops from raiding, not 2
  • Blocked the Federation Code technology for some empires, for example homicidals. To draw the tech, the empire is also required to be in contact with someone they can form a federation with.
  • Blocked the Development focus task Form a Federation for some empires, for example homicidals
  • Added swaps for some empires, for example homicidals, for the Development focus rewards Federation Code, Xeno Diplomacy, and Xeno Relations
  • Updated the Colony view tab mentioned in the hint of the focus task Enact a Planetary Decision to say Management
  • CyberCreed pops with Ritualistic Implants can now colonise planets
  • Fixed Recycled and Luxurious traits not applying to Roboticists
  • Catalytic Processing Civic now lists correct information regarding job swap
  • Cost for repairing orbital rings when you use bioships is now correctly calculated
  • Gale Speed trait gained from Defeat no longer causes errors
  • Fixed scope for LeaderShipSurvivalReason
  • Fixed scope bug for ruler in leader_election_weight
  • Fixed Worker Coop gaining Elite strata jobs in too many places and tidied up the civic tooltip
  • Updated tooltip for Warrior Culture civic
  • Added a pre-list colon to the Feudal Society civic's tooltip
  • The everychanging stone can no longer cause artisans to have negative mineral upkeep
  • Gave the Neural Chorus advanced authority the pop growth speed modifier that had accidentally been assigned to Memory Aggregator
  • The Planetary Supercomputer no longer has an empire cap of 1
  • The Research Institute/Planetary Supercomputer no longer give scientist capacity
  • Added dashes to Traits tooltips and list items
  • Fixed trigger logic for criminal syndicates and federations
  • Fixed Offspring Bioships not being visible in game
  • Fixed Offspring Bioships not being labelled as non-offspring ships in the ship designer
  • The Machine Uprising will no longer spawn 100 machine pops for every 1 missing housing. However the pop-rework seems to have handled 6 million machine pops okay.
  • Stopped removing occupation armies for bombarded and invaded planets on savegame load
  • Repairing ruined buildings in zones is now always possible.

Performance​

  • Flattened pop job modifier node into planet one
  • Made clearing modifiers a fire and forget job

Stability​

  • Fixed a possible OOS when a player leaves the game.
  • Fix CTD when generating a Cosmic Storm mesh.
  • Fixed a random freeze when loading save with stations containing multiple defence platforms.

We expect the 4.0.14 release will be next week (probably on Tuesday), and is expected to include some fixes to a few infinite loops and some select balance changes (like splitting up Enforcers and Telepaths again). It will be a short work week here in Sweden, so it’s likely to be the only update of the week.

As I mentioned last week, with multiplayer stability largely handled, AI is one of our next focuses. Today I want to talk about AI benchmarks, and have a discussion with you about how we should measure “success”.

What Makes a Good AI?​

The AI in Stellaris has always been designed as very reactive, and AI personality has a massive impact on their behavior. Our goal is for our AI empires to feel like actors in the galactic play - acting in a manner consistent with their Origins, Authorities, Civics, and Ethics rather than always picking the “meta” play.

They do still need to put up a bit of a challenge though, especially at higher difficulties.

The first economic goal we make for our AI is “please don’t collapse in an economic death spiral”, and it’s actually far better at that in 4.0 than it was in 3.x. The current AI does NOT meet the second “provide an adequate challenge” goal though.

One of the fundamental tools we have for our AIs are resource targets in their economic plans. They’ll strive to reach those targets, and many of these are set as “scaling” - if they meet the target, they’ll raise the target the next month. This attempts to ensure that they’ll keep thirsting for ever larger research and alloy numbers (or food if they use bioships!) as is appropriate. This is one of the tools we also use to make them exhibit their ethics - Materialists scale their Research targets faster than other empires, so they’ll inherently be more likely to build more Research specializations, while Spiritualists are more likely to have a lot of Unity specializations.

Ironically, improving AI tends to consume any benefits we carved out through performance improvements. The stronger the AI, the more stuff they have - fleets, colonies, and so on.

Benchmarking​

One way to decide whether or not the AI is performing up to expectations is through benchmarking - what kind of fleet power, alloy generation, and research generation should they have by 2230, 2250, 2300, and so on? Around what year should they hit 10k fleet power?

Then there come questions around whether the benchmarks should differ based on personality type. Should it be different if they’re Democratic Crusaders vs. Peaceful Traders? Or does differentiating them there make the friendlier empires too weak?

I’ve got my own set of benchmarks that come from running 3.14 and from the multiplayer community, and in general, I’m okay with Grand Admiral being significantly harder than it was in 3.14. but I’m interested to hear what you all strive for.

How much research and alloy production do you try to have 10 years, 30 years, 100 years, and when the end-game crisis comes calling? (Include your preferred difficulty settings and galaxy sizes as well if you could, as well as if you change any other important settings like tech costs.)

What’s Next?​

We’re going to continue with 4.0 post release support.

Since the next two weeks are both short weeks in Sweden, our next Stellaris Dev Diary will be June 12th. (You’ll be hearing from me in patch notes in the meantime though.)
 
  • 76Like
  • 16
  • 5
Reactions:
Maybe there needs to be some reactivity to the surrounding when deciding these economy plans. if the AI starts falling behind in tech, it should focus on picking up the pace of specializing into research worlds. if it starts falling behind in fleet size, it should start aggressively investing into industry worlds.

i play with difficulty scaling to mid game, and highest difficulty. but once the AI falls behind in tech, they can't seem to catch up anymore. making the whole galaxy feel uncompetetive.
 
  • 2
Reactions:
I will describe my experience of the game in version 4.0.13.

Administrative buildings cannot be improved, they need to be replaced, which makes me look for them.

Buildings that add a fixed number of jobs in mining districts force me to build one districs of each type on all planets. After all, in this case, I have 600 jobs from one districs (800 if I remove the automation building).

In the victory tab, I became the first about twice as fast as usual. I guess the AI is doing worse than usual in version 4.0. I like the new specializations, but you still have a lot of work to do.
 
To Improve the Ai, make it have a static goal until mid or endgame.
E.g. miltaristic country will always try to expand its alloy production and build its military up, also will go into research agreements and commercial pacts to catch up more willingly in exchange for favors.

Fix enslavement to make xenophobes the villains of the galaxy and be a threat. E.g. slaves will work all the worker jobs with no or marginal bonuses, but no upkeep. Maybe even let the slaves work in alloy foundries as workers?


I don't know how far the AI is using it, but generally let the Ai use all the systems there are and didn't use well until now like trade, diplomacy, planet specialization,
policies, different fleet builds,
espionage, galactic community, ascensions, research focus, starbases etc.
Tactics wise, let it change it fleet builds if it has enough info of the enemy fleet.

Make advanced empires focus on keeping it's dominance via vassalization/Federation, trading contacts to get the galactic community early.
Normal empires maybe even rush ascensions or diplomacy early.

Also make the player more of unknown factor and the Ai less likely to engage in diplomacy with the player.

TLDR: Let AI use all the systems the player has access to and focus on their strengths to assert dominance.
 
  • 1
Reactions:
.14 is coming? I got my Stellaris working, in the 4-13 area of updates. I've had some trouble with the screen size of the program vs the screen size of the screen. I found the trouble to be with the mods- and I did NOT start with that end of the list. Currently, I have switched off Resettle Colony, Noresettlementinfluence, Fleet transfer mod, and UI overhaul dynamic. There are other mods I left on.
After that was solved (not perfectly), the game shook down alright. There's still things almost off the edge of the screen, but half an X can still be clicked. The galaxy map button can be clicked again. I should take a second look at the ship creation screen. The empire economy is doing well from the start. No crippling shortages. The time is ticking well, no late game slowdown until the late game. I can only hope the new population setup will solve it.
Try to not to screw up window sizing. There's a lot of different monitor sizes out there.
 
I actually agree with you and ask the community, how do they defend this release given we have THREE HUNDRED AND SIXTY TWO bug fixes since release. Not one, not ten, not a hundred, but we are now over three hundred. This does not include balance changes, changes listed as improvements, AI, or UI/UX

For convenience I attached the link to the consolidate page of patch notes, its actually to large by any means to cut and paste just the bug fixes into a posting.

This weekend my friend and I sat down to play a multiplayer game on version 4.13. In two hours we encountered 2 desyncs and 3 errors. I don't even want to play the single player version.

There are so many errors that there are still many months of fixing them ahead. I won't even mention the balance, there's a lot of work to do there. I don't think they'll have time to fix all the errors and polish the balance before the next DLC comes out, which will give rise to a bunch of new ones.

Considering how they work - sometimes a short work week, sometimes another vacation, sometimes holidays. And the top management is inclined to release an unfinished product and, judging by the result, save on the number of developers.

From a good and stable version 3.14 we got 4.xxx, which could still be unstable and frankly unbalanced for another year.

In fact, this is a complete failure that has undermined both trust in the company and the ability to comfortably play new versions of 4+ for a long time.
 
  • 1
  • 1
Reactions:
Been playing multiplayer, mostly vs AI with a lot of people recently and I'm starting to think the AI isn't that bad, at least, not at it's core functions.

Every single match have been the same, I mean, without GA bonuses it can't possibly hope to challenge a player in it's first war as it's very easy to simply take supremacy first then fill your expanded naval cap with corvettes and rush down any neighbor, including GA difficulty, but other than being far too slow with their opening moves they've been doing alright for the first ~100 years of the game under standard rules, they have big-ish fleets, they have a functional economy, they can research, it's not underperforming that much.

The game changes, radically, when the player gets access to these techs that start boosting flat pop production, such as +1 minerals per 100 miners or +1 research per 100 researchers, once the player starts stacking those the game goes nutts, economy goes insane, every resource starts getting capped at the max and every research ends within 6 months, they break the game, completely, and after a couple of decades every single AI, GA difficulty or not, will be pathetic compared to the player.

These bonuses ruin the game, now that pop numbers are no longer an issue, why do we even need them in the game? I mean, at all, why can't we simply delete them all?
 
  • 3
  • 3
  • 1
Reactions:
Hi everyone!

The 4.0.13 update released today with the following changes:

Stellaris 4.0.13 Patch​

Improvements​

  • Behemoth Fury is now available to Wilderness Empires.
  • Improved tooltips for the following civics:
    • Functional Architecture/Constructobot
    • Environmentalist
    • Astro-mining Drones
    • Maintenance Protocols
    • Ascensionists
    • Augmentation Bazaars
    • Brand Loyalty
    • Death Cults
    • Dimensional Worship

Balance​

  • Mutagenic Habitability now counts all planet types as ideal for upgrading Gaiaseeders
  • Dramatically increased the draw chance for the Mineral Purification, Global Energy Management and Food Processing technologies
  • Rebalanced the Pleasure Seeker civics to transform Civilians into Hedonists
  • Logistic Drones are now Complex Drones not Menial Drones

Bugfix​

  • Fixed invaded pre-ftls not becoming biotrophies
  • People once more die when they are put in the Lathe
  • Bio-Swarmer missiles can now be used by all biological ships with medium weapon slots (including defensive platforms)
  • Pops that are being pampered will now be forcibly switched to the correct living standard
  • Replacing a district specialisation no longer destroys CyberCreed buildings that should be kept
  • Corrected a tooltip bug where a planet would display itself as a possible migration target.
  • Fixed capitalisation for resources in trade policies
  • Updated assorted modifiers that still referred to Clerks
  • A Trade deficit now causes Job Efficiency and Empire Size issues
  • Fixed the tooltip for the Polymelic trait
  • Armies now protect 200 pops from raiding, not 2
  • Blocked the Federation Code technology for some empires, for example homicidals. To draw the tech, the empire is also required to be in contact with someone they can form a federation with.
  • Blocked the Development focus task Form a Federation for some empires, for example homicidals
  • Added swaps for some empires, for example homicidals, for the Development focus rewards Federation Code, Xeno Diplomacy, and Xeno Relations
  • Updated the Colony view tab mentioned in the hint of the focus task Enact a Planetary Decision to say Management
  • CyberCreed pops with Ritualistic Implants can now colonise planets
  • Fixed Recycled and Luxurious traits not applying to Roboticists
  • Catalytic Processing Civic now lists correct information regarding job swap
  • Cost for repairing orbital rings when you use bioships is now correctly calculated
  • Gale Speed trait gained from Defeat no longer causes errors
  • Fixed scope for LeaderShipSurvivalReason
  • Fixed scope bug for ruler in leader_election_weight
  • Fixed Worker Coop gaining Elite strata jobs in too many places and tidied up the civic tooltip
  • Updated tooltip for Warrior Culture civic
  • Added a pre-list colon to the Feudal Society civic's tooltip
  • The everychanging stone can no longer cause artisans to have negative mineral upkeep
  • Gave the Neural Chorus advanced authority the pop growth speed modifier that had accidentally been assigned to Memory Aggregator
  • The Planetary Supercomputer no longer has an empire cap of 1
  • The Research Institute/Planetary Supercomputer no longer give scientist capacity
  • Added dashes to Traits tooltips and list items
  • Fixed trigger logic for criminal syndicates and federations
  • Fixed Offspring Bioships not being visible in game
  • Fixed Offspring Bioships not being labelled as non-offspring ships in the ship designer
  • The Machine Uprising will no longer spawn 100 machine pops for every 1 missing housing. However the pop-rework seems to have handled 6 million machine pops okay.
  • Stopped removing occupation armies for bombarded and invaded planets on savegame load
  • Repairing ruined buildings in zones is now always possible.

Performance​

  • Flattened pop job modifier node into planet one
  • Made clearing modifiers a fire and forget job

Stability​

  • Fixed a possible OOS when a player leaves the game.
  • Fix CTD when generating a Cosmic Storm mesh.
  • Fixed a random freeze when loading save with stations containing multiple defence platforms.

We expect the 4.0.14 release will be next week (probably on Tuesday), and is expected to include some fixes to a few infinite loops and some select balance changes (like splitting up Enforcers and Telepaths again). It will be a short work week here in Sweden, so it’s likely to be the only update of the week.

As I mentioned last week, with multiplayer stability largely handled, AI is one of our next focuses. Today I want to talk about AI benchmarks, and have a discussion with you about how we should measure “success”.

What Makes a Good AI?​

The AI in Stellaris has always been designed as very reactive, and AI personality has a massive impact on their behavior. Our goal is for our AI empires to feel like actors in the galactic play - acting in a manner consistent with their Origins, Authorities, Civics, and Ethics rather than always picking the “meta” play.

They do still need to put up a bit of a challenge though, especially at higher difficulties.

The first economic goal we make for our AI is “please don’t collapse in an economic death spiral”, and it’s actually far better at that in 4.0 than it was in 3.x. The current AI does NOT meet the second “provide an adequate challenge” goal though.

One of the fundamental tools we have for our AIs are resource targets in their economic plans. They’ll strive to reach those targets, and many of these are set as “scaling” - if they meet the target, they’ll raise the target the next month. This attempts to ensure that they’ll keep thirsting for ever larger research and alloy numbers (or food if they use bioships!) as is appropriate. This is one of the tools we also use to make them exhibit their ethics - Materialists scale their Research targets faster than other empires, so they’ll inherently be more likely to build more Research specializations, while Spiritualists are more likely to have a lot of Unity specializations.

Ironically, improving AI tends to consume any benefits we carved out through performance improvements. The stronger the AI, the more stuff they have - fleets, colonies, and so on.

Benchmarking​

One way to decide whether or not the AI is performing up to expectations is through benchmarking - what kind of fleet power, alloy generation, and research generation should they have by 2230, 2250, 2300, and so on? Around what year should they hit 10k fleet power?

Then there come questions around whether the benchmarks should differ based on personality type. Should it be different if they’re Democratic Crusaders vs. Peaceful Traders? Or does differentiating them there make the friendlier empires too weak?

I’ve got my own set of benchmarks that come from running 3.14 and from the multiplayer community, and in general, I’m okay with Grand Admiral being significantly harder than it was in 3.14. but I’m interested to hear what you all strive for.

How much research and alloy production do you try to have 10 years, 30 years, 100 years, and when the end-game crisis comes calling? (Include your preferred difficulty settings and galaxy sizes as well if you could, as well as if you change any other important settings like tech costs.)

What’s Next?​

We’re going to continue with 4.0 post release support.

Since the next two weeks are both short weeks in Sweden, our next Stellaris Dev Diary will be June 12th. (You’ll be hearing from me in patch notes in the meantime though.)

One other thing I’d like to add is there are still several non- empire AI bugs in the game that are at best annoying and at worst capable or breaking a piece of content entirely. The main one off the top of my head is that FTL inhibitors can permanently disrupt marauder raids if there isn’t a valid path to the target colony that goes around the inhibitor but there are doubtlessly others with similar problems
 
  • 1
Reactions:
I had a look around my galaxy in 2400. and i've got this protectorate. it was originally a pre-ftl but when it hit "maturity" it got lucky and had space to expand. i don't remember if i acquired it peacefully or by force, but it was ages ago.

it now has 8 planets and 4 habitats, and not a single planet except their homeworld produces any research whatsoever. and since it's a swam i can't even help it along with megacorp research outposts.

and the rest of the galaxy doesn't fare much better. maybe they have some physics research from storm buildings. but rarely much else.
 
  • 2Like
Reactions:
Been playing multiplayer, mostly vs AI with a lot of people recently and I'm starting to think the AI isn't that bad, at least, not at it's core functions.

Every single match have been the same, I mean, without GA bonuses it can't possibly hope to challenge a player in it's first war as it's very easy to simply take supremacy first then fill your expanded naval cap with corvettes and rush down any neighbor, including GA difficulty, but other than being far too slow with their opening moves they've been doing alright for the first ~100 years of the game under standard rules, they have big-ish fleets, they have a functional economy, they can research, it's not underperforming that much.

The game changes, radically, when the player gets access to these techs that start boosting flat pop production, such as +1 minerals per 100 miners or +1 research per 100 researchers, once the player starts stacking those the game goes nutts, economy goes insane, every resource starts getting capped at the max and every research ends within 6 months, they break the game, completely, and after a couple of decades every single AI, GA difficulty or not, will be pathetic compared to the player.

These bonuses ruin the game, now that pop numbers are no longer an issue, why do we even need them in the game? I mean, at all, why can't we simply delete them all?
It allows you to progress with a tall playstyle. Without it all your planet development will look like it was in 3.14 - you slap 11 buildings (now districts + buildings) and you reach max production that planet can provide. So if you want more production - your only way is to get more planets. I hated to have extra 3-5 science/unity/fortress worlds as the only option to somehow increase production in 3.14.
 
  • 4
  • 1
Reactions:
The Chosen, GA, Mid game 2275, mid game scaling.
In 3.14 i saw how they opened their wormhole in ~2277 year and with "FREEEEEEEEEDOM!" attacked my vassal.
Yesterday on 4.0.13 at year 2307 i discovered their wormhole , war was declared, but it seems they didn't have a tech to use wormholes.
THIS is your benchmark. Isolated empire with static system and planets. They should research wormholes in 2275-2280.

In 2306 they have + 700 energy, +700 minerals, +600 food, +300 cg, +13 alloys (lol), +550 trade, +666 science, +900 unity.
Make them have +10-50 food and cg, and get much more alloys and science.
Here is their 10 district Science habitat with archive specialization, colonized in 2214 (91 year ago)
 

Attachments

  • 2306.04.05 Chosen no wormholes yet.sav
    6,8 MB · Views: 0
  • Screenshot_1354.png
    Screenshot_1354.png
    718,6 KB · Views: 0
  • 2Like
  • 1
Reactions:
It allows you to progress with a tall playstyle. Without it all your planet development will look like it was in 3.14 - you slap 11 buildings (now districts + buildings) and you reach max production that planet can provide. So if you want more production - your only way is to get more planets. I hated to have extra 3-5 science/unity/fortress worlds as the only option to somehow increase production in 3.14.
That's not progression, it's obliterating game balance to the point not even ashes remain.

And there's no reason why it doesn't apply to wider empires, as a giant empire with over 100 planets the economy just goes nutts exponentially once you get those techs to the point even a x25 crisis on GA starts to look puny.
 
  • 4
Reactions:
It allows you to progress with a tall playstyle. Without it all your planet development will look like it was in 3.14 - you slap 11 buildings (now districts + buildings) and you reach max production that planet can provide. So if you want more production - your only way is to get more planets. I hated to have extra 3-5 science/unity/fortress worlds as the only option to somehow increase production in 3.14.
I'd actually argue that you have this a little backwards, at least in 4.0:
Wide playstyles have difficulty filling planets because you don't get the extra population per planet from the base 3 population growth anymore. You have a lot of space for pops, but what is valuable is getting as much productivity out of each worker that you have. This means that buildings which increase output without adding jobs are very valuable. After all, if you need more jobs for people, you have the space to build them!
By contrast, a taller empire will have more filled-up planets, so getting as many jobs out of your planet is much more important. You probably still don't have an excess of workers without a lot of automation, but only a rather mature planet will have all its districts built. At that point, you're looking for any marginal productivity you can find, and that could favor jobs over efficiency. ... except the efficiency buildings are powerful enough that they're quite possibly the right choice on a filled-up planet, too, especially if you have some things that buff civilian output.
For researchers, it's a little complicated, so I'm going to do the math out. The +2 to a specific researcher type adds 33% to that researcher's base output while also giving 200 of the relevant job. If you build all three of them, that's 600 jobs and all research jobs are 1/3 more productive than they would have been. (And when you go wide, strategic resources are currently basically free once you can build refineries, because you're not especially space-limited.) Or, you could build three Advanced Research Complexes and get 1620 jobs, which is sort of like getting an extra 5 double-specialized urban districts of researchers, because your baseline is 600 jobs. (If you use the mixed one, it's a bit more than 11 urban districts, but you could double it up. Single-research ones are 11% more jobs.) But if you have around 15 double-specialized urban districts built, then you wouldn't get extra research from it, because the flat bonus (which is handy here because it multiplies with the other modifiers) buffs stuff up enough to make up for it. 15 is actually a lot of urban districts for a planet, but also each advanced research building (built in another slot, for example) counts for an effective three double-specialized mixed research urban districts. The +2 to all researchers one is even more effective in this respect, despite adding zero jobs, because it is a third as many buildings for the same effect (aside from the multiplier and extra jobs), which means it would be worth it at closer to 10 districts, which is very plausible. (With the two types together, you're looking at +25% for whichever you consider "second", at which point a the single building break-even is at around 12 equivalent districts, but three of those are also providing jobs equivalent to a district, so it ends up in about the same place.) On the other hand, the building that isn't worth it when your planet is filled up and you have a lot of tech, ironically, is the +15% output building, because 180 of each job type will add as much as 15% of 21(!) urban districts.

4.0 may have successfully killed "more planets is better because of the fixed population growth bonuses," but it's still the case that more planets means you don't even need to consider whether marginal jobs are better than output bonuses. Assuming tall empires need to get the most productivity out of each of their planets rather than out of each of their pops, it turns out that the efficiency buildings mostly end up benefiting wide empires over tall ones.
 
I generally like this--except for the idea of having AI-buffs that trigger off the player getting something---then that would make getting some events that are nominally good a *BAD* thing because it buffs everyone else.

"oh man, I got the <blank> event which gives <blank> which is not very good--I guess I won't finish this chain"---> which is neither fun, nor lore/role-playing friendly.
If the modifier is generic enough (equivalent to a repeatable technology for example), only given when the player gets meaningful modifiers of at least that strength, and the rate low enough (like in the example given) then statistically it never be a bad thing to get the modifier while still letting the player scale. Infact, it diversify the AI slightly because each AI might roll a different modifier (if it rolls one at all).

Another more complex option would be to give each AI a situation like Evolutionary Predators that scales slowly with time, such that it only gain the first modifier in a decade. Then have it gain progress with each survey and colony, akin to the player getting more opportunities by doing so, as well as each time the player gets a modifier granting progress and difficulty multiplying progress. Then, each modifier takes longer and longer to get and the player will outpace the AI as long as they grow steadily, while still scaling somewhat based on player fortune so as to not over/underwhelm too easily.
 
Hi everyone!

The 4.0.13 update released today with the following changes:

Stellaris 4.0.13 Patch​

Improvements​

  • Behemoth Fury is now available to Wilderness Empires.
  • Improved tooltips for the following civics:
    • Functional Architecture/Constructobot
    • Environmentalist
    • Astro-mining Drones
    • Maintenance Protocols
    • Ascensionists
    • Augmentation Bazaars
    • Brand Loyalty
    • Death Cults
    • Dimensional Worship

Balance​

  • Mutagenic Habitability now counts all planet types as ideal for upgrading Gaiaseeders
  • Dramatically increased the draw chance for the Mineral Purification, Global Energy Management and Food Processing technologies
  • Rebalanced the Pleasure Seeker civics to transform Civilians into Hedonists
  • Logistic Drones are now Complex Drones not Menial Drones

Bugfix​

  • Fixed invaded pre-ftls not becoming biotrophies
  • People once more die when they are put in the Lathe
  • Bio-Swarmer missiles can now be used by all biological ships with medium weapon slots (including defensive platforms)
  • Pops that are being pampered will now be forcibly switched to the correct living standard
  • Replacing a district specialisation no longer destroys CyberCreed buildings that should be kept
  • Corrected a tooltip bug where a planet would display itself as a possible migration target.
  • Fixed capitalisation for resources in trade policies
  • Updated assorted modifiers that still referred to Clerks
  • A Trade deficit now causes Job Efficiency and Empire Size issues
  • Fixed the tooltip for the Polymelic trait
  • Armies now protect 200 pops from raiding, not 2
  • Blocked the Federation Code technology for some empires, for example homicidals. To draw the tech, the empire is also required to be in contact with someone they can form a federation with.
  • Blocked the Development focus task Form a Federation for some empires, for example homicidals
  • Added swaps for some empires, for example homicidals, for the Development focus rewards Federation Code, Xeno Diplomacy, and Xeno Relations
  • Updated the Colony view tab mentioned in the hint of the focus task Enact a Planetary Decision to say Management
  • CyberCreed pops with Ritualistic Implants can now colonise planets
  • Fixed Recycled and Luxurious traits not applying to Roboticists
  • Catalytic Processing Civic now lists correct information regarding job swap
  • Cost for repairing orbital rings when you use bioships is now correctly calculated
  • Gale Speed trait gained from Defeat no longer causes errors
  • Fixed scope for LeaderShipSurvivalReason
  • Fixed scope bug for ruler in leader_election_weight
  • Fixed Worker Coop gaining Elite strata jobs in too many places and tidied up the civic tooltip
  • Updated tooltip for Warrior Culture civic
  • Added a pre-list colon to the Feudal Society civic's tooltip
  • The everychanging stone can no longer cause artisans to have negative mineral upkeep
  • Gave the Neural Chorus advanced authority the pop growth speed modifier that had accidentally been assigned to Memory Aggregator
  • The Planetary Supercomputer no longer has an empire cap of 1
  • The Research Institute/Planetary Supercomputer no longer give scientist capacity
  • Added dashes to Traits tooltips and list items
  • Fixed trigger logic for criminal syndicates and federations
  • Fixed Offspring Bioships not being visible in game
  • Fixed Offspring Bioships not being labelled as non-offspring ships in the ship designer
  • The Machine Uprising will no longer spawn 100 machine pops for every 1 missing housing. However the pop-rework seems to have handled 6 million machine pops okay.
  • Stopped removing occupation armies for bombarded and invaded planets on savegame load
  • Repairing ruined buildings in zones is now always possible.

Performance​

  • Flattened pop job modifier node into planet one
  • Made clearing modifiers a fire and forget job

Stability​

  • Fixed a possible OOS when a player leaves the game.
  • Fix CTD when generating a Cosmic Storm mesh.
  • Fixed a random freeze when loading save with stations containing multiple defence platforms.

We expect the 4.0.14 release will be next week (probably on Tuesday), and is expected to include some fixes to a few infinite loops and some select balance changes (like splitting up Enforcers and Telepaths again). It will be a short work week here in Sweden, so it’s likely to be the only update of the week.

As I mentioned last week, with multiplayer stability largely handled, AI is one of our next focuses. Today I want to talk about AI benchmarks, and have a discussion with you about how we should measure “success”.

What Makes a Good AI?​

The AI in Stellaris has always been designed as very reactive, and AI personality has a massive impact on their behavior. Our goal is for our AI empires to feel like actors in the galactic play - acting in a manner consistent with their Origins, Authorities, Civics, and Ethics rather than always picking the “meta” play.

They do still need to put up a bit of a challenge though, especially at higher difficulties.

The first economic goal we make for our AI is “please don’t collapse in an economic death spiral”, and it’s actually far better at that in 4.0 than it was in 3.x. The current AI does NOT meet the second “provide an adequate challenge” goal though.

One of the fundamental tools we have for our AIs are resource targets in their economic plans. They’ll strive to reach those targets, and many of these are set as “scaling” - if they meet the target, they’ll raise the target the next month. This attempts to ensure that they’ll keep thirsting for ever larger research and alloy numbers (or food if they use bioships!) as is appropriate. This is one of the tools we also use to make them exhibit their ethics - Materialists scale their Research targets faster than other empires, so they’ll inherently be more likely to build more Research specializations, while Spiritualists are more likely to have a lot of Unity specializations.

Ironically, improving AI tends to consume any benefits we carved out through performance improvements. The stronger the AI, the more stuff they have - fleets, colonies, and so on.

Benchmarking​

One way to decide whether or not the AI is performing up to expectations is through benchmarking - what kind of fleet power, alloy generation, and research generation should they have by 2230, 2250, 2300, and so on? Around what year should they hit 10k fleet power?

Then there come questions around whether the benchmarks should differ based on personality type. Should it be different if they’re Democratic Crusaders vs. Peaceful Traders? Or does differentiating them there make the friendlier empires too weak?

I’ve got my own set of benchmarks that come from running 3.14 and from the multiplayer community, and in general, I’m okay with Grand Admiral being significantly harder than it was in 3.14. but I’m interested to hear what you all strive for.

How much research and alloy production do you try to have 10 years, 30 years, 100 years, and when the end-game crisis comes calling? (Include your preferred difficulty settings and galaxy sizes as well if you could, as well as if you change any other important settings like tech costs.)

What’s Next?​

We’re going to continue with 4.0 post release support.

Since the next two weeks are both short weeks in Sweden, our next Stellaris Dev Diary will be June 12th. (You’ll be hearing from me in patch notes in the meantime though.)
for year 30 benchmark marks it depends on what exactly I’m trying to do if I’m tech rushing I’m looking for 380-500 tech with maybe 40-100 alloys trying to keep my other resources above 50 a month.
If I’m trying to be aggressive I’m looking for 180+ alloys and 150+ tech around year 30. 4 planets for tech rush 5 with aloy rush. You should be able to start making a fleet with a
Fleet power around 12k to 30k depending on how well you hit rng.

year 10 bench marks 20 energy 50 minerals 0 food 100-120 science 30-60unity 20-30 alloys 3 planets
If playing aggressive 2-4K fleet power or if playing fruitful partner ship 8-15k if really lucky 20k.
Note I play a lot of gestalts and I use to play pvp pretty often haven’t played in a bit but for what I can remember these would be about what I was getting to in my games depending on rng.
 
I hate "unfair bonuses to AI as a difficulty scale" with passion. That is like increasing difficulty of chess by replacing some of AI's pawns with queens.

I want the difficult AI to outsmart me, by superiour micromanaging, efficient economy, tricks with jump drive or cloaking - stuff I could then learn myself. Not "giving the control of my country to AI would instantly make the country stronger due to bonuses AI gets".

Instead, now we have AI that is just absurdly dumb, for instance: AI's capital/homeworld was completely depopulated, but they still had two other colonies in the home system left intact. The AI did built a colony ship, but even a century later, it would just stay in space - so the AI is not even smart enough to recolonize their homeworld/former capital.

That's just riddiculous for a game which's patch number starts with 4.
 
  • 4
  • 1
Reactions:
I hate "unfair bonuses to AI as a difficulty scale" with passion. That is like increasing difficulty of chess by replacing some of AI's pawns with queens.

I want the difficult AI to outsmart me, by superiour micromanaging, efficient economy, tricks with jump drive or cloaking - stuff I could then learn myself. Not "giving the control of my country to AI would instantly make the country stronger due to bonuses AI gets".

Instead, now we have AI that is just absurdly dumb, for instance: AI's capital/homeworld was completely depopulated, but they still had two other colonies in the home system left intact. The AI did built a colony ship, but even a century later, it would just stay in space - so the AI is not even smart enough to recolonize their homeworld/former capital.

That's just riddiculous for a game which's patch number starts with 4.
For truly superior performance they need to re accommodate the code to support neural networks for fleet decision making. Right now much of it is embedded on non public game code so that no modder can interfere.
 
  • 1
Reactions:
I hate "unfair bonuses to AI as a difficulty scale" with passion.
The funny thing is that compared to, say, Firaxis or Triumph, Paradox's in-house studios are really restrained when it comes to the use of AI bonuses as a difficulty scale...
 
I know I'm late to this topic, but i just wanted to say my 2 cents.

Can you balance the a.i around the ensign rank (where the ai and players have no inherent bonuses)? If the ai is stronger there it should make higher difficulties more challenging to right?
 
  • 3Like
  • 1
Reactions: