Stellaris Dev Diary #325 - 3.10.3 "Pyxis" Released [d2aa] + Further Beta Plans

[sr999]

What’s After 3.10.4?
Tentatively scheduled for next Friday, we plan on putting up a longer open beta over the holidays that seeks to collect feedback regarding some potential balance changes to ship production, upkeep, and research in general.

Stellaris has undergone a significant amount of power creep over the years, and the speed at which we're able to burn through the entire technology tree is much higher than is healthy for the game. Due to the large number of stacking research speed modifiers, repeatable technologies are reached far too early in the game. Another power creep issue mentioned by many players, it's also become trivial to stack large numbers of ship build cost and ship upkeep reduction modifiers.

The Holiday Open Beta will be a feature branch that contains the following changes, which may or may not go into 3.11 (or 3.12, or any release at all for that matter). Similar to how we handled Industrial Districts several years ago, we're intentionally keeping these separated from core 3.11 development, isolating this in a parallel track.

We’ll have a feedback form set up to collect your thoughts, and the Open Beta will run until the middle of January.

Spoiler: Holiday Open Beta Preliminary Notes

• Research Speed Bonuses now usually come with increased Researcher Upkeep.
  • By changing these to throughput bonuses (cost + production), a technology focused empire will require more Consumer Goods or other resources depending on who they use to research. This puts a partial economic break on runaway technology.
• Reduction in most Research Speed bonus modifiers.
• The +20% Research Field technologies have been removed. In their place we have introduced new "Breakthrough Technologies". These technologies are required to reach the next tier of research.
  • Whether it be the transistor, the theory of relativity, or faster-than-light travel, occasionally there are technologies that redefine a field of science.
    • The intent of these breakthrough technologies is to slow down the front-runners a little bit, while still letting the slower empires get pulled along.
  • Breakthrough technologies start off more difficult than regular technologies but have a variant of tech spread - the more nations you have at least low Technological intel on who have already discovered them, the cheaper they are to research (even down to instant research once the theory is commonplace). This tech spread varies based on galaxy size.
    • Enigmatic Engineering prevents this tech spread.
  • Breakthrough technologies have animated borders to stand out.
• 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.
  • All other researchers, such as Necromancers, have been left alone for now
• 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.
• Replaced or removed most 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.
  • Chosen of the Eater of Worlds ship build cost reduction reduced to 5% from 15%, and no longer modifies ship upkeep.
  • 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.
  • Corporate Crusader Spirit 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

I find this deeply ironic. For years I've been playing Stellaris less and less, as it's a kind of trade-off between time and enjoyment (not helped by a slow rig). As a result, I rarely play beyond 2300, if that.

The latest Astral stuff has reversed this trend, because I've been able to start chomping into "mid-game" type stuff early enough to keep me interested.

Now to me this latest dev diary signals that the devs are going to "slow everything down". Don't get me wrong: I endorse this in theory, as it sounds very good in principle, indeed well thought out! However, for me the personal flip side is it may well push me further away from the game again.

All that said, I've just had a thought. I suppose I can "correct" all this just by messing with the settings, thus accelerating various things like research to "compensate" for this well-thought-out slowdown!

Never mind then, just my 2c...

 

[Foxosaur]

We do not have anywhere near enough information to make a qualified guess or mathematical estimate as to how long time it'll take to complete the tech tree.

All we have are some baseline numbers for tweaks to some of the systems involved coupled with vague descriptions without numbers of changes to other existing systems, and general information about the introduction of a new system that will shake up the tech game (breakthrough techs)

Couple this with the population curve in Stellaris making it difficult to extrapolate how substantial changes to existing systems will play out decades later, and I can assure you that anybody claiming that "the maths" points to any number in particular is making things up as he goes along, not making a qualified guess.

Spoiler: Consider this if you want boring reasoning:

• New base science costs for techs; These are details, that can be used in calculations
• New base researcher output; Likewise
• We have been told about the elimination of certain science output modifiers; Likewise
• We have been told about the introduction of throughput modifiers coupled with upkeep modifiers; We don't have any numbers, so we we cannot meaningfully use this information in calculations to extrapolate, yet this information is essential to making any sort of good estimate
• We have been told about the introduction of expensive breakthrough techs, that gatekeep technological advantage and provide something of a catch-up mechanic; We don't have any idea about the number of such gate-keeping techs, or costs if you are the prime mover researching it first, or the rapidity of decline in price as the tech spreads, and we don't know whether somebody who wants to be on the bleeding edge of tech will need to be prime mover on all those breaktroughs or whether it will make strategic sense to focus in some areas and letting others take the lead in other areas, so given that we actually know NOTHING about how it will impact tech gameplay, how can we possibly incorporate this information in our extrapolations with more confidence than "I made some assumptions, and I could theoretically be right"
• What does the current curve of research output divided by tech penalty (which is the value that matters, not science output, which is merely one factor in the equation) look like through a normal game (whatever that is), and how will the above changes - and other changes not specifically mentioned but falling under the general headers of change announced - affect this?
• From a certain point in current gameplay that curve will often trend towards the population growth curve for an empire; Do we assume that this will still be the case? If so, why? If not, why not? Given that we are talking about projecting things hundreds of years into the 3.11 game based on extrapolating from the current first hundred years or so, even if we had full information on anything else, which we don't, any mistaken assumptions here could easily end up making decades or hundreds of years of difference in the end

Too many interdependent systems, too little information. This really is a wait and see case.

Sometimes Peter, we’re liable to miss the wood for the trees. They are giving an indication, not everything must be finite or extrapolated fully to comprehend or interpret something, and that’s what forums allow us to do, speculate with what information we are given.

 

[Troyen]

"Multiple people", who exactly?

I'd appreciate it if you wouldn't repeatedly dismiss anyone that disagrees with you as a 25x GA player.

I've run a thousand hours of SP/MP games on captain/commodore with a group of people that are definitely not minmaxers. Roleplaying, handicap challenges (e.g. no using the market until the galactic market, or empire size restrictions, or no conquest runs, etc.), enabling automation after the 6th colony, etc. are all very not-meta. The last commodore game, players were getting their teeth kicked in, so no 25x crisis here. Haven't even beaten 5x yet. Galaxy is often in a state that we rarely even encounter end-game lag, and that's playing on the mac version which runs a full speed slower than the windows version.

Yet everyone in the group hits repeatables in the 2300s. Even with an accelerated 2425 end game date, that's enough for 20 levels of repeatables (with planetary automation running nuts and building most of those planets, mind you). I don't doubt there are players that like the current balance, but my experience is the second half of the tech tree blows by so fast people skip upgrading multiple tiers of weapon/ship components, or ignore mid-game weapons entirely. It's also a very common complaint I hear from new players (though I won't claim all new players share it) that the last 100-150 years of the game (on default settings) are boring. You can find it when they ask questions on this forum and get told to move the endgame year up. You can find the same questions and answers on reddit or discord or wherever. I think it's fair to claim that tech, especially later-game tech since your research speed grows exponentially, is too fast, and has been too fast for several years now.

When Victoria 3 came out, you could get all your law changes and research the entire tech tree and then sit around with nothing to do for the last few decades because the game wasn't balanced properly for the suggested defaults. They went back and among other changes slowed down the speed of tech (though it took a couple adjustments because it initially was overcorrected), and while I won't claim the game is perfect, there's certainly more to do now - enough that it doesn't feel like you're twiddling your thumbs for 40 years. Most nations can't quite finish the tech tree, but they can get close, and the result is you're making interesting decisions about what to prioritize all the way to the victory year. The number of complaints about having nothing to do at the end aren't gone, but they are much fewer in number.

Stellaris is a different game. Would only finishing 95% of the tech tree feel awful here? I don't know. (Though unless they fix the severe physics vs engineering balance, the experience is going to vary dramatically across the three trees.). It's possible they overcompensated, and that's what the beta is designed to find out (previous open betas suggest they'll refine before going live, if the changes even ever go live). It's possible they overestimated the number of players who actually want science changes, and the beta feedback will find that out too.

But I do think it's an experiment worth running. And I say that as someone who knows a fair number of middle-skill and newer players.

 

[Foxosaur]

I'd appreciate it if you wouldn't repeatedly dismiss anyone that disagrees with you as a 25x GA player.

I've run a thousand hours of SP/MP games on captain/commodore with a group of people that are definitely not minmaxers. Roleplaying, handicap challenges (e.g. no using the market until the galactic market, or empire size restrictions, or no conquest runs, etc.), enabling automation after the 6th colony, etc. are all very not-meta. The last commodore game, players were getting their teeth kicked in, so no 25x crisis here. Haven't even beaten 5x yet. Galaxy is often in a state that we rarely even encounter end-game lag, and that's playing on the mac version which runs a full speed slower than the windows version.

Yet everyone in the group hits repeatables in the 2300s. Even with an accelerated 2425 end game date, that's enough for 20 levels of repeatables (with planetary automation running nuts and building most of those planets, mind you). I don't doubt there are players that like the current balance, but my experience is the second half of the tech tree blows by so fast people skip upgrading multiple tiers of weapon/ship components, or ignore mid-game weapons entirely. It's also a very common complaint I hear from new players (though I won't claim all new players share it) that the last 100-150 years of the game (on default settings) are boring. You can find it when they ask questions on this forum and get told to move the endgame year up. You can find the same questions and answers on reddit or discord or wherever. I think it's fair to claim that tech, especially later-game tech since your research speed grows exponentially, is too fast, and has been too fast for several years now.

When Victoria 3 came out, you could get all your law changes and research the entire tech tree and then sit around with nothing to do for the last few decades because the game wasn't balanced properly for the suggested defaults. They went back and among other changes slowed down the speed of tech (though it took a couple adjustments because it initially was overcorrected), and while I won't claim the game is perfect, there's certainly more to do now - enough that it doesn't feel like you're twiddling your thumbs for 40 years. Most nations can't quite finish the tech tree, but they can get close, and the result is you're making interesting decisions about what to prioritize all the way to the victory year. The number of complaints about having nothing to do at the end aren't gone, but they are much fewer in number.

Stellaris is a different game. Would only finishing 95% of the tech tree feel awful here? I don't know. (Though unless they fix the severe physics vs engineering balance, the experience is going to vary dramatically across the three trees.). It's possible they overcompensated, and that's what the beta is designed to find out (previous open betas suggest they'll refine before going live, if the changes even ever go live). It's possible they overestimated the number of players who actually want science changes, and the beta feedback will find that out too.

But I do think it's an experiment worth running. And I say that as someone who knows a fair number of middle-skill and newer players.

Where do I find such alluring games MP games?!

 

[Peter Ebbesen]

Sometimes Peter, we’re liable to miss the wood for the trees. They are giving an indication, not everything must be finite or extrapolated fully to comprehend or interpret something, and that’s what forums allow us to do, speculate with what information we are given.

I agree.

Speculation along the lines of, oh, the rate of tech advancement will overall slow down considerably, because that is the stated intention and the declared changes sure look like they support this, and I think it might work out thus, or might cause such and such problems, or change the game so and so for better or worse, is normal and healthy and I don't nitpick that, much.

But once somebody appends numbers to it and use it for predictions and arguments from authority, even if unintentionally, as in claims like "the math says.." we are no longer talking about indication or speculating, but extrapolation based on numbers, and it make sense to check the assumptions and the data behind such claims.

 

[currylambchop]

It might be a bit heavy handed with long dormant and well-balanced things like A Grand Fleet being noticed and nerfed now because of the whiplash from just how broken this currently is, but it's the right direction.

My main problem is how it seems they went overboard. They weren’t a problem until the addition of the council, and it would make more sense to remove ship build cost from the council rather than touching balanced things like a grand fleet or supremacy (admittedly, the latter is quite powerful but it wasn’t meta defining).

It doesn’t seem like much thought went into actually solving the problem. Other ways existed, for example changing it to ‘ship build cost multiplier’ which stacks multiplicatively rather than additively. And they wouldn’t have removed fun.

 

[Foxosaur]

I agree.

Speculation along the lines of, oh, the rate of tech advancement will overall slow down considerably, because that is the stated intention and the declared changes sure look like they support this, and I think it might work out thus, or might cause such and such problems, or change the game so and so for better or worse, is normal and healthy and I don't nitpick that, much.

But once somebody appends numbers to it and use it for predictions and arguments from authority, even if unintentionally, as in claims like "the math says.." we are no longer talking about indication or speculating, but extrapolation based on numbers, and it make sense to check the assumptions and the data behind such claims.

I think we’ll have to disagree, we don’t all have the time or inclination to look at numbers and neither should speculation and extrapolation be reserved for those that do.

As for authority, if that was aimed at me, Im no different to anyone else and am invested into this game, I’m allowed to interpret relative accuracy as well as definitive which others may prefer to stick to.

TLDR, you do you and I’ll do me

 

[Dr Pippy]

We do not have anywhere near enough information to make a qualified guess or mathematical estimate as to how long time it'll take to complete the tech tree.

All we have are some baseline numbers for tweaks to some of the systems involved coupled with vague descriptions without numbers of changes to other existing systems, and general information about the introduction of a new system that will shake up the tech game (breakthrough techs)

Couple this with the population curve in Stellaris making it difficult to extrapolate how substantial changes to existing systems will play out decades later, and I can assure you that anybody claiming that "the maths" points to any number in particular is making things up as he goes along, not making a qualified guess.

We've been given the numbers for tech generation; we know roughly the amount of research points required to exhaust the tree. It's a simple exercise in proportional reasoning. The other effects you mention seem geared to slow down the tech curve even more; the fact that we don't know, e.g., the exact effect of the gatekeeping techs (which I think are actually an interesting idea) is a pretty weak reasoning for dismissing the rest of the calculations.

Just to be transparent about the calculations and assumptions:

• 4->3 researcher output: 1.333x research time. This one is cut and dried.
• Loss of +20% techs: Roughly (2 to 2.2)/(1.4 to 1.6)~1.4x research time. Based on a late-game jobs bonus of ~100-120% from scientists. This one is a little more handwavy: (1) you really only get +40% in the early game; the last +20% is a late-game tech; (2) the other bonuses (e.g., from capitol productivity boost) also come online later. These uncertainties have opposite effects, so (1.3x to 1.5x) is likely a fairly accurate result.
• 1000*2^n->500X(2^n+3^n): ~3x research time. This one is much more questionable: I just based it on adding up the total time to research a single tech at each level from T1 to T6 tech, under each system. This is an accurate calculation if and only if the techs are evenly weighted between the tiers, which they probably aren't. Without carefully checking, it seems like there are more high-tier techs than low tier, which means the 3x is almost certainly a substantial underestimate. But I'd rather err on the side of being cautious with the calculations.
• 1.333x1.4x3=5.6: the proposed changes will increase total time to fill the tech tree by a factor of 5.6.

If you think the uncertainties that you've mentioned are likely to be similar effect sizes, or if you want to question any of my assumptions (which are in places very questionable!), that's a much more interesting argument than appealing to second-order unknowns to dismiss estimates of a zeroth-order effect.

 

[Ajacoba]

The rubber band mechanic will probably shorten the research times significantly, so once the galactic community exists and you get some contacts in your list you and the ai will all work together (at least indirectly), so you only have to pay the big cost if you're tech rushing, if you chill a bit the ai will prepare the way for you and you can then research it at a lowered cost

Unironically enough might make espionage better because you get to steal tech on top of rubberbanding, even if AI is behind you in tech there's probably stuff you skipped which you can then steal

 

[hart30]

The only issue I have with breakthrough techs is...the RNG of Stellaris's tech tree. I feel like breakthrough techs might end up like how ascension used to be...you'd sit around for ages waiting for say gene modding or psionics or whatever, because the RNG would play against you. I'd rather it be, a 'once you research x number of techs in this tier it gives you a permanent option for the breakthrough tech', because we all know it'll happen, you get your breakthrough tech at the same time something like chemical motes or some other 'need to have' tech rolls and then next thing you know you're stuck for ten years in tier one because the tech doesn't re-roll.

Just dont research the next breakthrough tech before the tech u want pops up. So if its a tier 2 tech u dont research breakthrough for tier 3. In my opinion this allows u to even goal for a specific tech or even rare tech faster as it forces the tech pool to dry up.

 

[Ajacoba]

I mean, the breakthrough will just replace the current tier techs, no? The progression should stay just as rng as it is currently

 

[Tannhäuser Cake]

It's worth noting that, compared to other specialist jobs, research is the only one without a throughput-boost building (nanoplants etc) and also the one with by far the most output-from-tech free buffs. They may not require changes, as they already use the exact same overall system as tech will in this beta.

I am also thinking about the non-resource-specific bonuses to output from jobs, such as from Stability, Civics, Traditions, World Types, Edicts, Telepaths, and so on. These sources could also be considered for replacing "generic" output bonuses with "generic" throughput bonuses.

(Perhaps even Worker job resource bonuses, if they - hypothetically - were to start using Energy Credits as their input resource.)

Here's some food for thought. If 1x tech cost in the beta is 8x the research cost compared to before, players who like the old research times won't even be able to use the slider to get it back, because even the lowest setting on the slider (0.25x tech cost) will be twice as slow as the current 1x tech cost.

As someone that enjoys 0.75x tech cost with early mid and endgame years because condensing the game into fewer real-life hours and avoiding endgame lag is just more fun, I won't even be able to play the game nearly the same way as before. The sliders won't help me. I think a lot of people might even quit Stellaris altogether if it becomes that bad.

That seems very easy to address - just increase the granularity of the slider, either in the base game or via mods.

 

[Arkarnia]

It's interesting because civilizations that only got into space in 2800 don't look so bad because all the others have already overtaken empires that have almost fallen.

It remains important to fix the Marauders. After half a century the Marauders are getting overwhelmed and there are no more punts.

These huge stacks have to disappear from the fleets. The size of the fleet is too utopian exaggerated no matter what century.

The most important thing remains to change the pop system to minimal calculations. The pops should not be calculated individually but as a sum and should be a kind of pool.

 

[blahmaster6k]

That seems very easy to address - just increase the granularity of the slider, either in the base game or via mods.

If they did that in the base game, I would be fine with that. As someone who enjoys achievement hunting though, I don't want to have to resort to mods in order to have the same experience that was previously available in the base game. My main hope is that they will test out these more extreme changes, see how they play out, and then rein them in and put out a more polished set of changes later.

That's why my initial post in this thread was asking for a Dev to talk about what they actually want out of these nerfs and what their goal is for an "appropriate" time to reach technology milestones. Without knowing what the devs want other than "general slowdown of tech progression" it's impossible to really judge whether the changes are meeting the intended goals or not.

 

[Peter Ebbesen]

We've been given the numbers for tech generation; we know roughly the amount of research points required to exhaust the tree. It's a simple exercise in proportional reasoning. The other effects you mention seem geared to slow down the tech curve even more; the fact that we don't know, e.g., the exact effect of the gatekeeping techs (which I think are actually an interesting idea) is a pretty weak reasoning for dismissing the rest of the calculations.

Just to be transparent about the calculations and assumptions:

• 4->3 researcher output: 1.333x research time. This one is cut and dried.
• Loss of +20% techs: Roughly (2 to 2.2)/(1.4 to 1.6)~1.4x research time. Based on a late-game jobs bonus of ~100-120% from scientists. This one is a little more handwavy: (1) you really only get +40% in the early game; the last +20% is a late-game tech; (2) the other bonuses (e.g., from capitol productivity boost) also come online later. These uncertainties have opposite effects, so (1.3x to 1.5x) is likely a fairly accurate result.
• 1000*2^n->500X(2^n+3^n): ~3x research time. This one is much more questionable: I just based it on adding up the total time to research a single tech at each level from T1 to T6 tech, under each system. This is an accurate calculation if and only if the techs are evenly weighted between the tiers, which they probably aren't. Without carefully checking, it seems like there are more high-tier techs than low tier, which means the 3x is almost certainly a substantial underestimate. But I'd rather err on the side of being cautious with the calculations.
• 1.333x1.4x3=5.6: the proposed changes will increase total time to fill the tech tree by a factor of 5.6.

If you think the uncertainties that you've mentioned are likely to be similar effect sizes, or if you want to question any of my assumptions (which are in places very questionable!), that's a much more interesting argument than appealing to second-order unknowns to dismiss estimates of a zeroth-order effect.

Roger, in that case let me question your second and third assumptions. Let me start with the third, research time, as it is the easiest to deal with, and it ties directly into my earlier somewhat undiplomatic statement, that Couple this [the unknown new systems] with the population curve in Stellaris making it difficult to extrapolate how substantial changes to existing systems will play out decades later, and I can assure you that anybody claiming that "the maths" points to any number in particular is making things up as he goes along, not making a qualified guess.

#3: You have a much more fundamental problem than the one you mention. Your calculation does not tally research time; it tallies research output cost.

Spoiler: Reasoning

That a tech requires thrice the total research output cost to research does not mean it takes thrice the time to research. It might, but it is much more likely that it'll take either longer or shorter depending on changes in the empire occurring while it is being researched.

The cost is only equivalent to the time, if the ratio of research output/(1+empire size penalty) is constant over time, where research output = science output * (1 + science output modifiers) * (1 + techspeed modifiers)

This is rarely the case, since - if nothing else - POPs keep growing, even when you've reached a state where the techspeed might be constant for the rest of the game.

So long as you invest the POP growth such that the research output/(1 + empire size penalty) ratio is increasing, the projected research time for each remaining tech is decreasing, which means that the remaining time to completion decreases faster than an 1:1 correspondence with the passage of time, and it keeps speeding up every year you manage to keep the ratio increasing.

Likewise, if the ratio is decreasing, the time to completion slows down more than the 1:1 passage of time.

In short, an argument based on "total cost of all current research is X and takes Y years, so if the total cost is nX, it'll take nY years, and I can use nY multiplied by other factors affecting time completion for projecting research completion" is, absent an argument for why the ratio can be assumed constant, fundamentally flawed. If the ratio is increasing, it will take shorter time, if decreasing, longer, and the larger n the greater the difference between actual time to completion and factor n completion is likely to be in both cases.

Currently, after techspeed becomes constant, and sources of science output increases have been obtained, that ratio is still increasing over time in the late game, accelerating research, when all that is left is to employ ever more researchers in ascended research environments supported by ascended consumer goods and unity producers, and none of the announced changes suggests that this will not remain the case.

EDIT: Further information for the non-mathematicians. I didn't include this originally because I was sure DrPippy would follow the argument without it and I was addressing him, but for the enlightenment of others.

It is certainly possible that the ratio will reach a plateau or even start decreasing, but if it does, will that be before or after the tech tree has been completed? And how will this differ for different builds? At best the calculation can be used to establish an upper boundary for completion time if one makes the assumption that it will only reach a plateau after completion, which might be wrong, but even if the assumption holds true, that wouldn't give us any idea whether actual completion time would be anywhere close to that upper boundary, or decades, even hundreds, of years earlier.

And this fundamental problem, that the ratio changes over time both depending on player's actions, that are themselves affected by how the tech changes play out, and which even if the player takes no actions is nevertheless subject to POP growth over time, is why I stated confidently that anybody claiming that "the maths" points to any number in particular for how long it would take to complete the tech tree wasn't making a qualified guess, but making it up - because we simply don't have the necessary information for anything but, at best, taking a stab at establishing an upper boundary that might be way too high to be of practical use, and we cannot say with any confidence whether it is or not.

Now, on to:

#2: Loss of +20% techs: 1.4x research time, based on a late-game jobs bonus of ~100-120% from scientists. I think a factor 1.17 to 1.32 based on actual late game job-bonuses of 145%-300% (or a bit more) from scientists is a much better reflection of reality

Spoiler: Reasoning

Does a late game of 100-120% increased science output reflect reality in the current game? Well, it certainly might in a given game, but it probably doesn't.

Common late game science output modifiers:
65% tech (3x20% + 5% from synthetic thought patterns)
30% stability
30% capital building
10% research subsidies
10% or 20% from a level 10 governor
----
145% total for a research planet that isn't sector capital and doesn't have its own dedicated governor, +155% if it is.

Situational late game science output modifiers:
52.5% - 78.25% ascended research ringworld modifier; Not everybody uses these late game, but it is certainly an option; With the proposed changes this becomes available much later,but there are existing ringworlds out there to be conquered or found in many games, and you can even start with one, so a determined player will get his hands on some by origin, conquest, or beelining the required techs
x% scholarium vassal(s)
x% governor traits
x% POP traits
x% ascension bonuses
x% traditions, civics, or ethics, if any of the science boosting ones (usually small 5% or 10% bonuses)
x% origin related
x% anomaly, archaeological sites, astral rifts (if there are any that provide science output rather than speed... My initial thought is that this is the case, but right now I can't think of any, so perhaps there aren't. EDIT: Oh, right. The Memorex and the Evermore. So there are at least two items in this category!)
x% precursor bonus for some disciplines from some precursors
x% from non-of-the-above (such as Void Loops from studying Void clouds)


So taking advantage of the bare minimum available to everybody in the late game, and nothing more than that, we hit +145% or +155%. Perhaps you don't have a level 10 governor, or perhaps you don't have the highest level of capital unlocked yet, or possibly the empire has chronic instability, so it is possible for that part of the equation to end up in the +100-120% range - but it isn't likely.

Getting none of the situational bonuses, however, that requires real work. If nothing else then because every ascension path provides some, and it is hard not to have some researchers with additional research increasing traits if the player wants it.

In practice, that means that a numbers of +100% to +120% science output are often seen mid-game, not late-game.

For the highest current modifiers, we must go to Psionics with a Divine Sovereign and a Divine Conduit on the Council (+40-46%), in an ascensionists, harmonious, Holy Covenant, with a level 10 Shroud Preacher or Truth Seeker governor on the planet, who has further traits boosting science production, with POPs that have Psionics and optionally a common science trait as well, on an fully ascended research ringworld with a Psi-Corps (but no Sanctum; Since you only have one Sanctum, it would be unfair to count it for the general research ring world), and there might be a few extra bonuses from anomalies, or Void Loop research for 10% physics, or.. well, there are currently a lot of small modifiers available in the game.

For such an empire, it is normal to end up with 16-17 research of each discipline/researcher, i.e. a total of 300%+ increased science output, and we are talking something like 4/3.4 ~ 1.17 factor, with a slightly lower factor if you are pushing the boundaries by min-maxing.

That's the extreme case - my hard-and-fast rule is that non-psionic empires typically end up with around 80% lower increased output than psionic ones in the late-game, but of course every game differs and it isn't the same for all types of resources. What is clear is that most empires that don't have some special research boosting mechanic from origin will fall somewhere in between the +145% and the +300% increased science output in the endgame if it uses the tools available.

---

Stray thought: Unless some unannounced changes are made to the planetary ascension/designation system, I expect one welcome outcome of these changes will be to finally convince those who haven't already realized how valuable a resource unity is, that both unity and science are of high value in building a high tech empire, either by making using of ascended research ringworlds for increased science output or ascended research non-ringworlds for consumer goods savings, something that may become more important with the throughput changes to research speed.

Then again, it just might make them double down on raw science production.

 

[CocoCincinnati]

If research happening too quickly was a concern, maybe it wasn't the best idea to allow scientists to assist research on entire sectors instead of just one planet like they used to.

 

[Abdulijubjub]

If research happening too quickly was a concern, maybe it wasn't the best idea to allow scientists to assist research on entire sectors instead of just one planet like they used to.

The effect of scientists assisting research is still weaker than it was before the rework. You used to be able to get a governor and a scientist assisting research on literally every planet, if you wanted. That would be +50-70% research output (depending on level and traits), vs. +40-60% to the sector capital/+20-30% to every other planet, now.

Brain Slugs+Intellectual+assisting research meant +47% baseline, with another +4% per level (of the governor and scientist) beyond 1. Add in Erudite, and you could have started with 52 baseline and gone up to 88% (though that would require a level 10 governor and scientist assisting research, which isn't likely to happen).

Research speed went bananas with the Paragons rework, though. The council is incredibly powerful.

 

[Starfury]

If research happening too quickly was a concern, maybe it wasn't the best idea to allow scientists to assist research on entire sectors instead of just one planet like they used to.

It fits their (and presumably a large/loud part of the playerbase's) idea of how sector governors are supposed to work

Also it was quite possibly the very least of the factors which lead to the current research snowball.

 

[Millbot]

The developer only have so much time to run a beta and it sounds like this one will be about a month. You cannot run a worthwhile beta, if you thought is "we must do painstakingly measured changes every step of the way," because you're going to run out of time well before you manage to do anything of note. What you want to do is come up with your ideas for changes, then make some guesstimates on how far you can push things before they start to break. That usually the point that you do want to start you beta. If the outcome isn't even close to what you had estimated, that can be a good indication that your changes aren't going to work be it that you broke the game in ways you didn't expect or that the changes do practically nothing to impact the game in the desired manner. If it does pan out close enough to expectations, that let's you know that the changes are more or less doing what they are intended to do and you can start to zero in on the numbers that are needed.

As a player, you shouldn't go into the start of beta with the expectation of the game being anywhere near balanced and should expect things to be broken in ways that might make the beta unplayable. If one or both of those things easily tilt you, then you are the kind of player that should stay out of beta tests because you clearly aren't going to into them with a mindset that is going to result in you giving useful feedback. Trust me, there are plenty of players that can go into a beta that is absolutely busted and give feedback that doesn't boil down to a rant of accusing the devs of being hacks or having no idea on what they are doing. I've noticed that players that often do get titled by the early beta experience often accuse developers of both those things without bothering to provide any feedback of substance that covers the actual state of things in the beta. If you're a player that doesn't like the beta experience at all, but feel you have to play it somewhat so that you can try to give feedback, that will prevent the devs from implementing changes you really don't like. Well your best bet is to wait until their is 1-2 weeks left in the beta. Where the game is likely to be more stable and where the devs are more likely to be trying to fine tune things that they are strongly considering for the live game experience.

Anyways, on the engineering tech bloat. I'd suggest that maybe some of the consumer good techs should go into society research, for similar reason that the repeatables on fighter tech ended up there. Might even be worth considering moving fighter tech completely into that tree. Then maybe move some of the robotic, cybernetic and synthetic stuff into physics, since a fair bit of it does technically deal with computing.

 

[ASGeek2012]

As are your concerns about prolonging the game and dying by lag.

This is the elephant in the room that I'm surprised is not being talked about as much. As someone who routinely plays with 1.5x tech cost specifically because I feel the tech tree goes by too fast, I often hit that abominable lag by the time I do get to the end techs.

The devs need to realize that their tech slowdown is not going to work unless it is paired with significant performance improvements in the late game.