Stellaris Dev Diary #326 - 3.10.4 "Pyxis" Released [e9b6] + Upcoming Holiday Tech Beta

[SirBlackAxe]

i feel like scientists exploration always finishes far too early in the game..

A silver lining from the issue with AI not hiring enough scientists is that in my 3.10 game I've still got a pair of explorers surveying untouched systems and finding new anomalies after 2300, and I'm not even playing on max galaxy size. I'm going to miss that.

 

[amib]

Noo, why would you do this to me after I carefully calculated the numbers for you in the other thread when discussing science output modifiers: The realistic factor is something in the range of 1.17-1.32, depending on governor development and planetary ascension.

1.3 is definitely at the high end, for players that don't engage in planetary ascension of research ring worlds and don't have governors with good traits, but at least have highlevel governors.

Anyhow, it'll be interesting to see how this plays out.

And I really, really, hope that nobody goes away from your post dreading a factor 8 increase to research time because of the factor 8 increase to total research costs of your estimate. Would you please, pretty please, add just a few lines explaining how the research cost factor 8 does not in any way reflect a factor 8 to research time?

it already takes me til 2350-2375 to get mega engineering, and it's wella fter 2400 before i'm out of techs and burning repeatables. if they make it take longer i'll never play iron man mode again.

 

[blahmaster6k]

I'd like to point out something that no one is talking about in this thread, and that's that the planetary governor and sector governor effects for Erudite and Cyborg traits should be swapped for flavor reasons. Erudite has always been the trait that in terms of flavor is based around research and having faster research speeds than everyone else, but now Erudite has zero bonuses to research speed at all outside of its pop trait. Meanwhile, Cyborg gets +10% research from jobs on its governor traits when the reduced pop upkeep would fit well with its flavor that is seen elsewhere such as the ruler trait.

Erudite went from being better at specifically research compared to Cyborg and Synthetic to being the worst out of the three with this proposed change.

 

[amib]

I'd like to point out something that no one is talking about in this thread, and that's that the planetary governor and sector governor effects for Erudite and Cyborg traits should be swapped for flavor reasons. Erudite has always been the trait that in terms of flavor is based around research and having faster research speeds than everyone else, but now Erudite has zero bonuses to research speed at all outside of its pop trait. Meanwhile, Cyborg gets +10% research from jobs on its governor traits when the reduced pop upkeep would fit well with its flavor that is seen elsewhere such as the ruler trait.

Erudite went from being better at specifically research compared to Cyborg and Synthetic to being the worst out of the three with this proposed change.

that's because the erudite species trait, not the leader variant, gives the most research bonus of all the traits in the game.

 

[Millbot]

preferably they would make all techs useful.. ideal world i know.. but so many hanging fruits i feel in terms of tech that could use a buff.. some in terms of weapons that need looking at as they are useless now.. others are things like survey speed and colony speed maybe those could be condensed into 1 of each instead 2 or 3

While it's true that there are some techs that need a buff. There are other cases where someone might not want to research a tech because regardless of how good or bad it is, it doesn't really help them with the goal they are aiming for.

So even if we got an ideal setup where all techs were worthwhile, if one was shooting for the setup that makes good use of them. RNG could result in a really unpleasant experience where someone ends up either research breakthrough tech way earlier than they would like or researching a ton of techs that don't do them much good.

I'll concede that the idea of techs be useful dependent on build choice might not play super well with the RNG nature of Stellaris research, but I do feel the setup is actually a good idea. Done right it means people have more choices. I'd actually argue we probably should get some more leader traits, that are geared towards the idea of making the player want to lean into certain decisions (point in case, maybe have more admiral traits that make us lean into certain ship components). Granted, that would mean the devs would have to make the RNG nature of leader traits a little more consistent in some ways. Gonna be hard to sell people on the idea of an commander trait that increases energy weapon performance, if we're still deal with the issue of often having leaders stuck with traits that were never relevant to what we assigned them to do.

 

[Chill-Wind]

Thanks for delivering a steady version before Christmas!
The beta looks interesting. Currently I have to set tech cost to 2X even more to makes the tech improvement slower for realism. Do you have any suggestion for new setting benchmark?

 

[Miuramir]

- archeological sites these in real life are never always done.. there is always more sites to be found over time as tech improves and people find more tombs etc.. this should be a semi constant thing too do.. maybe the results could vary much more so u not getting too much resources off it? .. or give us more unique and luxury resources from them we could trade or sell or research maybe another function scientists could do..
- astral rifts .. these could also take longer too research in the game..

I think there's an idea in here that you just missed: what if you could assign scientists to already-completed archaeology sites and astral rifts to generate a small amount of useful stuff? And have the useful stuff depend not only on scientist level, but on any bonuses they have to the relevant type of thing (skill, speed, etc.).

This gives late-game scientists focused on archaeology something to do once the sites are all explored (and similarly for astral rifts), and also makes the sites slightly valuable as things to capture as they are a potential small but steady source of resources. Some archaeology sites already leave a 1 or 2 minor artifacts deposit when finished, so this would be an extension of that. There might also be a "commercial tourism" policy that reduced the science / artifact / etc. output in exchange for some energy and consumer goods.

 

[Shark7]

Okay, I actually did end up running some extremely approximate but still reasonable calculations on this. The net result is that under the new system, you'd spend about 600 years to do as much research as you currently do in 150. (This is assuming linear population growth over time, which seems rather optimistic, but also makes the math a lot more tractable.) So given my playstyle and abilities, I'd expect to finish the tech tree in about 2800. In my current games, I generally feel like I'm starting to get into late-game tech (e.g. Citadels, X-slotted battleships, etc.) between 2270 and 2300. New ETA would be between 2420 and 2550.

In short, actual research times generally seem likely to be about 3-4x what they currently are. Again, pretty rough model so the error bars are pretty big, but it also seems fairly unsurprising. If anyone wants me to describe my math in more detail here or via DM, I'm happy to, so let me know. It's pretty nerdy though.

While I personally love the game to last a long time, this might be just a bit excessive. I don't think the average player wants to spend weeks on one play-through like I do. Doubling the research time would seem sufficient to me.

I wonder, do these calculations you have done also take into account the effect of empire size? Or would this be based on not exceeding the limit where the penalties are applied?

 

[blahmaster6k]

that's because the erudite species trait, not the leader variant, gives the most research bonus of all the traits in the game.

Not true unless you isolate only to that one trait. Erudite specifically cannot be stacked with intelligent, so bio ascension pops cap out on +25% research from jobs if you have both +20% from Erudite and the +5% from Robust.

Meanwhile:

• Cybernetics can stack Intelligent with Logic Engines and Efficient Processors for the same +25% from traits. Add +10% resources from cybernetic pops from the Cybernetics tradition tree and +10% research from a Cyborg governor for a total of +45% research from jobs.
• Synthetic benefits from Logic Engines' +10% and efficient processors' +5% for a total of 15% from pop traits, but Synthetic also has +25% robot output that also applies to research for a total of +40%. (5% from governor, +10% from synthetics tradition, +10% from synthetics tech).
• Psionic ascension can stack Intelligent with Psionic for a total of +20% research output. Then add another +20% resources from Psionic pops from telepaths thanks to the Psi Corps building, and you have a total of +40% research output from jobs for Psionic as well.

So basically, biological ascension actually has the slowest research output per pop of all four ascension paths. Bio's entire power budget is in species traits. The Genetics tradition tree itself offers no pop output bonuses, and it doesn't uniquely benefit from any technologies that improve output either. All three other ascension paths have buffs to resource output and some specific research buffs, but Bio ascension has none of that and must compete based solely on its species traits for output bonuses.

At this point the only thing that keeps bio ascension competitive with the others is the ability to have the most pop growth and assimilate hive minds, because its resource output per pop is lower than all the others unless you're specifically micromanaging a unique species template for every job in your empire. Its research output is always the lowest.

That's a far cry from the old days where bio could compete with synthetic on research output because the erudite trait granted research bonuses at all levels of leader (ruler, governor, scientist).

 

[Andy_Dandy]

When is the Technology Open Beta out today, lads?

 

[Eladrin]

When is the Technology Open Beta out today, lads?

It's now live.

 

[Peter Ebbesen]

Okay, I actually did end up running some extremely approximate but still reasonable calculations on this. The net result is that under the new system, you'd spend about 600 years to do as much research as you currently do in 150. (This is assuming linear population growth over time, which seems rather optimistic, but also makes the math a lot more tractable.) So given my playstyle and abilities, I'd expect to finish the tech tree in about 2800. In my current games, I generally feel like I'm starting to get into late-game tech (e.g. Citadels, X-slotted battleships, etc.) between 2270 and 2300. New ETA would be between 2420 and 2550.

In short, actual research times generally seem likely to be about 3-4x what they currently are. Again, pretty rough model so the error bars are pretty big, but it also seems fairly unsurprising. If anyone wants me to describe my math in more detail here or via DM, I'm happy to, so let me know. It's pretty nerdy though.

I don't mind nerdy!

I am curious how you went about making assumptions to create reasonable calculations, since, to me, growth that is linear or sublinear is something that happens when you are pretty much done with the game and are playing on fast-forward - not something that can be considered representative of most of the game.

Spoiler: Argument for why this is so hidden for those who don't care

I mean, at which point in your empire's development do you say, "ok, from this point forwards, after everything that has gone before, the empire's POP growth from all sources will be at best linear, but more likely sublinear?" Or, "looking at this from a distant enough vantage, the empire's growth curve from start to finish can pretty much be interpolated with a line that we can extend for predictive purposes?"

You obviously don't say it at the start of the game, since linear population growth based on the native growth of one planet is so very, very, far from what the empire's growth curve will be like even if you play a one-planet challenge. E.g. an empire that starts out with growing one POP every 20 months or so is likely to gain a LOT more than 30 POPs in the first half of the 23rd century, and a lot more than that during the second half of the 23rd century, and a lot more than that half century's growth during the first half of the 24th century, which is why an empire with 118 POPs in 2350 is, assuming default galaxy settings, considered tiny.

Of course, nobody says the contant growth factor for linear growth needs to match that particular factor of the earliest game growth from one planet, but how would one go about making a good estimate that is useful over an arbitrary time frame for an arbitrary empire?

At some time expansion and the wild superlinear POP growth slows down, then gradually approaches linear, then becomes sublinear as you have exhausted most sources of rapid POP growth and you are fighting the native growth function, but when that happens is heavily play-style dependent, so I simply cannot see how you can create reasonable assumptions allowing you to make reasonable calculations based on linear growth with any degree of confidence.

I mean, just consider the potential sources of growth: How would one even go about making a good general assumption about what any empire's growth curve would look like over a timespan of several centuries, without knowing the empire's build and the plan for how the empire would be run and which of the following opportunities it would avail itself of, and which not?

• Native growth on planets
• Immigration treaties aiding native growth
• Native assembly on planets
• Colonization of existing planets adding new sources of POP growth
• Creation of new planets adding adding new sources of POP growth
• Refugees
• Slave market, if there are any slaving empires around
• Raiding for POPs
• Conquest
• Vassal integration

With regards to your own tech - congratulations, you are pretty much following the intended tech curve, or at least not wildly violating it, and I can say that with some degree of confidence - at least if we define intended as "the game files say so".

This is because you are getting into Citadels in between 2270 and 2300, which is completely natural, because that is when the Citadel tech is explicitly gated.

Spoiler: Argument for why this is so hidden for those who don't care

The odds of getting it before 2270 are, everything else being equal, very remote due to the time gated (x0.1 factor) unless you are both Unyielding (x1.25) and have the Unyielding's Agenda Finisher active (x5), and have 3 or more Star Fortresses (x5), and even then the end result (x3.125) makes it very unlikely compared to most other techs available at the time unless you in addition have a hyperteching neighbour that already has the tech (x10).

Then it gets increasingly likely until 2300, when it is practically guaranteed, with a base factor before considering unyielding, neighbours, or star fortresses of (x24), making it very likely to be drawn. Simply the existence of a neighbour with citadel tech gives it a whopping (x240) factor on the tech draw at that time, and if you also have built three or more Star Fortresses, that is a (x1200 factor).

So in practice, unless you have pretty much exhausted the tech tree by hyperteching, you shouldn't expect to get Citadels before 2270, and unless you are behind the intended curve for a default setup galaxy, you should expect it to start turning up during the 2270-2300 timespan, being practically guaranteed that it'll show up near the end of it if you qualify.

(Battleships are gated 2250-2270, with the earliest and weakest XL weapon requiring both Battleships and X-Ray lasers; So XL weapons will typically show up a bit earlier than Citadels if you have focused on laser research)

tech_starbase_5 = {
	cost = @tier4cost2
	area = engineering
	category = { voidcraft }
	tier = 4
	prerequisites = { "tech_starbase_4" }
	weight = @tier4weight2

	# unlocks Citadel
	weight_modifier = {
		modifier = {
			factor = 0.1
			NOT = { years_passed > 70 }
		}
		modifier = {
			factor = 10
			any_neighbor_country = {
				has_technology = tech_starbase_5
			}
		}
		modifier = {
			factor = 2
			years_passed > 90
		}
		modifier = {
			factor = 3
			years_passed > 95
		}
		modifier = {
			factor = 4
			years_passed > 100
		}
		modifier = {
			factor = 5
			count_starbase_sizes = {
				starbase_size = starbase_starfortress
				count >= 3
			}
		}
		modifier = {
			factor = 1.25
			has_tradition = tr_unyielding_adopt
		}
		modifier = {
			factor = 5
			has_modifier = agenda_impenetrable_border_finish
		}
	}

	prereqfor_desc = {
		ship = {
			title = "TECH_UNLOCK_CITADEL_CONSTRUCTION_TITLE"
			desc = "TECH_UNLOCK_CITADEL_CONSTRUCTION_DESC"
		}
	}

	ai_weight = {
		factor = @ai_starbase_types_factor

	}
}

and battleships:

tech_battleships = {
	cost = @tier4cost1
	area = engineering
	tier = 4
	category = { voidcraft }
	prerequisites = { "tech_cruisers" }
	weight = @tier4weight1

	## unlock battleships
	prereqfor_desc = {
		ship = {
			title = "TECH_UNLOCK_BATTLESHIP_CONSTRUCTION_TITLE"
			desc = "TECH_UNLOCK_BATTLESHIP_CONSTRUCTION_DESC"
		}
	}

	modifier = {
		command_limit_add = 10
	}

	weight_modifier = {
		modifier = {
			factor = 0.1
			NOT = { years_passed > 50 }
		}
		modifier = {
			factor = 10
			any_neighbor_country = {
				has_technology = tech_battleships
			}
		}
		modifier = {
			factor = 2
			years_passed > 60
		}
		modifier = {
			factor = 3
			years_passed > 65
		}
		modifier = {
			factor = 4
			years_passed > 70
		}
		modifier = {
			factor = 1.25
			has_tradition = tr_supremacy_adopt
		}
	}

	ai_weight = {
		factor = @ai_ship_types_factor

	}
}

 

[Peter Ebbesen]

Not true unless you isolate only to that one trait. Erudite specifically cannot be stacked with intelligent, so bio ascension pops cap out on +25% research from jobs if you have both +20% from Erudite and the +5% from Robust.

Meanwhile:

• Cybernetics can stack Intelligent with Logic Engines and Efficient Processors for the same +25% from traits. Add +10% resources from cybernetic pops from the Cybernetics tradition tree and +10% research from a Cyborg governor for a total of +45% research from jobs.
• Synthetic benefits from Logic Engines' +10% and efficient processors' +5% for a total of 15% from pop traits, but Synthetic also has +25% robot output that also applies to research for a total of +40%. (5% from governor, +10% from synthetics tradition, +10% from synthetics tech).
• Psionic ascension can stack Intelligent with Psionic for a total of +20% research output. Then add another +20% resources from Psionic pops from telepaths thanks to the Psi Corps building, and you have a total of +40% research output from jobs for Psionic as well.

Then add another 20-26% from telepaths to Psionic once your empire has a Divine Conduit that improves their function assuming you have a highlevel governor and some effective level increases, and another 12.5% or 25% if the sector/planetary governor is a Shroud Preacher or Truth Seeker, and you end up with, on average, around +80% for Psionics on planets with telepaths in the late game, assuming the player gets the Divine Sovereign event and accepts, and makes an effort leveling governors, which admittedly is a very, very time consuming affair these days.

Psionics may have awful growth, but once its gets up and running, and assuming that the player doubles down on assimilating everybody to psionic, the POP efficiency for all jobs is much, much, better than anybody else's - and this is desparately needed.

 

[Alfray Stryke]

Psionics may have awful growth, but ... the POP efficiency for all jobs is much, much, better than anybody else's

This is 100% intended, when we rebalanced the ascension paths for Orion last year one of the goals was to ensure that the more pop growth/assembly an ascension path gave, the less effective their pops were, compared to other ascension paths.

 

[Peter Ebbesen]

This is 100% intended, when we rebalanced the ascension paths for Orion last year one of the goals was to ensure that the more pop growth/assembly an ascension path gave, the less effective their pops were, compared to other ascension paths.

And it has been much appreciated.

The goal of the Psionics ascension as communicated to players from Utopia onwards had been that Psionics compensated for less growth by being better/more efficient and the privilege of participating in the Shroud lottery, but until Orion that was never the case in practice. Apart from the exciting! participation in the Shroud Lottery.

Not that it was ever in as bad shape as the general forum wisdom had it, because once a thing gets tagged as "bad" that perception tends to self-reinforce, but it definitely didn't feel as if it accomplished its goals.

But I would have to say that it was the combination of Orion's ascension path rebalance and Orion's Holy Covenant, Ascensionists, and Harmony changes that did did the trick to make it feel right. You got both a baseline Psionic "less numerous, but more efficient" effect going for all Psionic empires and tied it loosely to the planetary designation system for those who wanted to focus even more on POP efficiency, by making the options for improving planetary designation be natural choices for Psionic empires, thus making "going Psionic, all out" be a whole-empire experience and not merely the choice of ascension path - and one centered around Unity, at that, as distinct from the very science heavy focus of nearly everything else.

Even if we forget efficiency for a moment, the experience of playing Psionic in Stellaris is just more rewarding these days because you are assembling many elements into a whole greater than the sum of its parts.

We still lack a repeatable psionic tech to make experience complete. And, of course, the PSI-fighters strike craft component that I've been asking for since Utopia. If Luke Starwalker could do it, you can do it!

 

[amib]

Hello everyone!

The 3.10.4 "Pyxis" update has been released today. Of particular interest in this week’s update are a fix that lets AI be more willing to recruit Scientists for exploration, a fix to certain modifiers doubling up, and some adjustments to negative leader traits.

Spoiler: 3.10.4 "Pyxis" Patch Notes

Balance

• Significantly reduced the yearly chance for leaders to gain negative traits.
• Due to player feedback, the Micromanager negative trait for Commanders now increases fleet upkeep instead of reducing command limit.
• The Lethargic negative trait for Commanders now also reduces fleet upkeep.
• The Nervous negative trait for Commanders now also increases disengagement chance.
• Having the Antagonistic Diplomatic Stance will now appease your factions asking for you to have a Strong Diplomatic Stance.

Bugfixes

• Added some failsafes to generate council positions and gestalt nodes in various edge cases where they weren't created or disappeared
• Clarified reformed tooltip for Dark Consortium civics
• Declining to Hunt for the Hyacinth should now correctly remove the event chain.
• Fix Astral Harvesting not available for heavily conquering empires
• Fixed a typo in the pre-FTL Provide Technology Tooltip.
• Fixed Civics not swapping to their alternative when switching Governments
• Fixed missing loc for the Fear of the Dark Admiral trait
• Fixed the Accelerated Time Astral Planes Modifier having a description for a title.
• Removed a redundant tooltip from the Synthetic Evolution event
• Ruler clothing selection in the empire designer is now respected
• Some planet modifiers that were getting incorrectly applied twice should now only be applied once.
• Updated the All Crisis Strength tooltip to accurately state that it doubles the strength of each subsequent crisis.
• Your ice miners will no longer make your Mercenaries homeless if they happened to have built their base above an ice asteroid.

AI

• AI Empires will now be more willing to hire leaders for their science ships as long as they believe they can afford the Unity upkeep.
• UI
• Fixed XP not being shown in XP bar tooltip for gestalt rulers and councilors.

Modding

• collapsable_leader_container should now calculate the correct height no matter the amount of horizontal slots.
• Fixed lock_country scope change incorrectly stating that the output scope was a bypass, when it is actually a country

Barring unexpected developments, we consider the 3.10 “Pyxis” release stable at this point, and the 3.11 “Eridanus” update is the next expected release - currently planned for 2024Q1. As mentioned in an earlier dev diary, 3.11 “Eridanus” will be focused primarily on general bugfixing and stability.

Technology Open Beta

We’ll be putting a Technology Open Beta up tomorrow to gather feedback and data on some possible changes to research. While some changes planned for 3.11 “Eridanus” have snuck into this release (like some improvements to Galactic Doorstep), the overall technology rebalance is considered experimental and may or may not end up being released. It is likely to undergo some changes if it does go live.

The Technology Open Beta will run from December 15th through January 15th. We’ll post a feedback form tomorrow to help gather your impressions and thoughts.

Spoiler: Technology Open Beta Release Notes

Features

• Replaced basic research technologies such as Quantum Theory with Breakthrough Technologies. Breakthrough Technologies will only appear once you have researched enough techs of your current tier and are required to research to reach the next tier.
• Breakthrough Technologies become easier to research based on the number of other empires you have low Tech intel on that have already researched them.
• Enigmatic Engineering blocks your Breakthrough Technologies from spreading.
• Events that give progress towards random technologies can grant progress towards Breakthrough Technologies, but will not give as much as they would a regular technology.

Improvements

• Technology and Tradition costs are now distinct sliders in galaxy setup.
• Added notification message when new pop settles in zeya (Gaia planet in azilash)

Balance

• Adjustments made to the Galactic Doorstep origin:
  • Added Gateway Cost and Megastructure Build Speed modifiers to the Galactic Doorstep origin.
  • Galactic Doorstep event chain now directly rewards the Gateway Activation technology and gives far more progress on the Gateway Construction technology
• Increased the effects of Empire Size on Technology to match its effect on Traditions.
• Rebalanced research speed modifiers. Most sources of Research Speed now have a corresponding increase in Researcher job upkeep
• Removed or adjusted many sources of Ship Cost and Upkeep reductions from the game.
• Military Buildup Agenda now improves ship build speed and reduces claim costs. (It still reduces War Exhaustion on completion.)
• Naval Procurement Officer councilor now improves ship build speed.
• Crusader Spirit civic now improves ship build speed.
• Psionic Supremacy (Eater of Worlds) finisher no longer reduces ship build costs.
• Vyctor's Improved Fleet Logistics trait now reduces ship build costs by 10% instead of 20%.
• Progress Oriented modifier no longer reduces ship build costs.
• Match tradition in the Enmity tree bonus to ship build costs reduced to 5% instead of 10%.
• Master Shipwrights tradition in the Supremacy tree no longer reduces ship build costs.
• Military Pioneer trait now reduces starbase upgrade costs instead of ship build costs.
• Shipwright trait no longer reduces ship build costs.
• Reduced penalty the Irenic trait applies to ship build costs.
• Sanctum of the Eater ship upkeep reduction reduced from 10% to 5%.
• Mark of the Instrument ship component no longer reduces ship upkeep.
• Grand Fleet ambition now increases power projection instead of reducing ship upkeep.
• Fleet Supremacy edict no longer reduces ship upkeep.
• Letters of Marque now reduces ship upkeep by 5% instead of 10%.
• Bulwark ship upkeep reductions reduced by 50%.
• Logistic Understanding, Armada Logistician, and Gunboat Diplomat traits now reduces ship upkeep while docked.
• Increased technology costs, especially those of higher tier technologies.
• The Technology curve has been changed from 1000 × 2^n to 500 × (2^n + 3^n), making the difference between an early and late-game tech more distinct.
• Reduced output of researcher jobs
• Researchers and their gestalt equivalents now produce 3 of each research instead of 4
• Head Researchers now produce 4 of each research instead of 6
• The effectiveness of Ministry of Science has been halved
• Astral Researchers now produce 5 physics and 1 of each other research instead of 5 physics and 2 of the other researches.
• Knights of the Toxic God balancing:
  • Slightly reduced the research output for Knight and Lord Commander jobs
  • Refactored how the output scaling for Knights from Squires functions, these now behave as normal additive modifiers instead of multiplicative modifiers
  • Knights now correctly inherit production modifiers from researchers and administrators
  • Slightly reduced the unity and research output for Knight and Lord Commander jobs
  • Knight output modifiers now only apply to resources, like other job output modifiers
  • The Fortress Habitat Designation for Knights no longer provides +1 Defensive Army per pop on the habitat, instead each Squire job provides +1 Defensive Army
  • Squires now increase the resource output of Knights by 2.5% not 2% per Squire.
  • The Luminous Blades modifier from the Knight's Quest now removes the alloy upkeep of knights and gives +25% Army damage instead of an empire-wide +1.5% alloy production modifier per knight
• Delegate GalCom focus traits now have a small chance to give favors.
• Reworked and rebalanced the Erudite, Cyborg, Synthetic, Psionic, Chosen One, and Chosen of X traits to include new leader assignments.

View attachment 1062609
View attachment 1062610​

Bugfix

• Repairing The Black Crown should no longer fire generic gateway repaired events.

Modding

• Added `last_resolution_category_changed` trigger

Meet the Devs Video

Game Systems Designer @Gruntsatwork did a video interview about what it’s like to work on the Custodian and Crisis teams.

No, you can’t implement trash. - E​

Next Week

The nights are getting cold and long, so for the last dev diary of the year it’s time for a look at the year in review.

See you then!

after playing on the beta, i really wish you guys would state the actual intended tech curve at default settings. i'm still on third tier techs in 2400 and getting my butt kicked by the crisis. feels horrible, and it seems like you're over-correcting a problem that never existed in the first place.

please stop trying to tune the game for the 1% meta build "rush" players. the rest of us don't do that crap.

 

[Dr Pippy]

I don't mind nerdy!

I am curious how you went about making assumptions to create reasonable calculations, since, to me, growth that is linear or sublinear is something that happens when you are pretty much done with the game and are playing on fast-forward - not something that can be considered representative of most of the game.

Probably overkill, but I've attached a png of a pdf with my calculations. (Wasn't going be able to type all this intelligently in a regular forum post.) It's obviously a very approximate answer, but given appropriately wide error bars (say, +/-100 to 150 years) seems pretty robust to different assumptions.

Pop growth seems pretty heavily playstyle dependent; I tend not to go on conquering sprees, so once I've got my borders secured and my habitats built, linear or sublinear growth is what I usually experience. But I did include some extra calculations for superlinear growth; the short version is that it makes a slight difference but doesn't change the overall story that I'd expect tech times 3-4x longer than what's on live.

I don't necessarily expect that anyone else would want to (or be interested) in this kind of modelling, but at the very least it was a good intellectual workout for me, so thanks for prodding me in that direction!

 

[Alfray Stryke]

Probably overkill, but I've attached a png of a pdf with my calculations. (Wasn't going be able to type all this intelligently in a regular forum post.) It's obviously a very approximate answer, but given appropriately wide error bars (say, +/-100 to 150 years) seems pretty robust to different assumptions.

Pop growth seems pretty heavily playstyle dependent; I tend not to go on conquering sprees, so once I've got my borders secured and my habitats built, linear or sublinear growth is what I usually experience. But I did include some extra calculations for superlinear growth; the short version is that it makes a slight difference but doesn't change the overall story that I'd expect tech times 3-4x longer than what's on live.

I don't necessarily expect that anyone else would want to (or be interested) in this kind of modelling, but at the very least it was a good intellectual workout for me, so thanks for prodding me in that direction!

Did you really write a document in LaTeX to detail this? I’m impressed.

 

[Peter Ebbesen]

Probably overkill, but I've attached a png of a pdf with my calculations. (Wasn't going be able to type all this intelligently in a regular forum post.) It's obviously a very approximate answer, but given appropriately wide error bars (say, +/-100 to 150 years) seems pretty robust to different assumptions.

Pop growth seems pretty heavily playstyle dependent; I tend not to go on conquering sprees, so once I've got my borders secured and my habitats built, linear or sublinear growth is what I usually experience. But I did include some extra calculations for superlinear growth; the short version is that it makes a slight difference but doesn't change the overall story that I'd expect tech times 3-4x longer than what's on live.

I don't necessarily expect that anyone else would want to (or be interested) in this kind of modelling, but at the very least it was a good intellectual workout for me, so thanks for prodding me in that direction!

Thanks!

That's not overkill - it is great - and I appreciate the amount of work you put into it. AND that you presented it in LaTeX to make it easy to follow. I can see how you reach your conclusions, and it'll be interesting to see whether your estimates are anything like your experiences in the beta test.

Spoiler: Comments on model hidden to void distracting from the thread's topic

One obvious question based on a trivial observation is, why would you play the way you are doing now when the environment changes, rather than adapting to the changes in the environment? But if we assume for the sake of argument that mostly your playing style won't change, a before and after comparison could be interesting.

As for the model itself, I like your simplification for empire size in terms of population as a substitute for the various sources and that you don't sweat the small details lsuch as the 100 offset because you are trying to look at the big picture. On the other hand, I must note that this approximation becomes ever worse as players attempt to reduce their empire size or the empire size effect.

As an example, even as an approximation, linear empire size penalty scaling with POPs is not how the total ES penalty behaves at all in ascensionist gameplay. Typically ES keeps rising in the early game, then stagnates as you start ascending planets, perhaps even dropping if you are focusing more on ascension than expanding. It will then start rising again, and using an approximation of scaling with POPs is (the special Guardianship example excluded) probably a decent approximation at that point - but it will be a much, much, less steep curve than that observed in the early game. So a better approximation would be piecewise linear function, that for completely non-ascensionist gameplay can be considered a linear function for simplicity.

And planetary ascension is not gated by tech, but by traditions.

As for using a power function to model POP growth, I can't think of anything better either, if you are to make a workable model, but that too gets more problematic the further from relying on native-POP growth the empire is.

For your entertainment, I looked at my most recent game from two months ago (so 3.9, with Champions of the Empire distorting the picture, and with extra unity from clearing blockers etc.), before I mostly ran out of time to play games. It is ascensionist, wide, vassal oriented, with a few chunks of conquest, one at the end of the early game (see if you can spot it ), and then killing off fallen empires at the end before I abandonend the game. Vanilla, huge galaxy, default settings apart from difficulty. (And turning off xeno-compatibility for the sake of sanity.)

It is not min-maxed for ES as I was testing other mechanics, so it has quite high ES for an ascensionist game and it doesn't have a period after ascending planets really gets going where the ES drops like a stone - but it is still an example where your approximations would have trouble coping.

It is also part of the reason that even when I try putting myself in your shoes, the memory of POP and ES numbers that behave so strikingly different from your approximations makes it hard to accept them as generally applicable, despite your playstyle probably being a lot closer to many players' way of playing the game than this, which is pretty much my default, when I'm not performing a particularly wide or tall test.

Year Col.  POPs   ES  ES/POP
2200    1    28   50    1.79
2210    6    46  129    2.80
2220   12    91  149    1.64
2230   15   134  205    1.53
2240   33   219  283    1.29
2250   35   339  359    1.06 Traditions completed late 2240s (3.9 unity blocker clearing, yummy!)
2260   36   454  392    0.86
2270   40   621  395    0.64
2280   41   730  391    0.54 Tech tree for all disciplines finally exhausted early 2270s; Highest repeatable 2280 is shield 15.
2290   55  1016  525    0.52
2300   89  1376  727    0.53
2310  115  1882  942    0.50 Highest repeatable is shield 42
GAME ABANDONED. At this point Champions of Empire provided a -142 ES distortion.

 

[blahmaster6k]

I appreciate the effort that went into modeling that, but I hope I won't have any more calculus flashbacks when all I want is to talk about Stellaris lol. The college classes were enough for me for one lifespan.