• 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 #72: Crises & The Contingency

Hello everyone and welcome to another Stellaris development diary. We are now officially back from our break in communication and will be resuming weekly dev diaries and streams as usual. Today's dev diary is going to be about crises, and how we're changing them in the future, particularly in regards to the AI crisis. Before I dive into it, I also want to mention that we are still working to address the issues caused by 1.6 and get another bugfixing patch out, the process has just been somewhat complicated by the Bradbury multiplayer beta. See this post for details and discussion of Bradbury/1.6.2 and keep this thread focused on the topic at hand.

Crisis Improvements & AI
Some time back, when I was asked about issues with the crises and the AI crisis in particular, I said that I did not want to put a great deal of resources into improving the end-game when those resources could be put into the mid-game instead, and that these improvements and fixes would come when we felt the mid-game were in a good enough place to justify them. I now feel that we are in that place, and as such we are going to make a major push to improve, balance and rework the endgame crises for future updates.

Probably the most significant change to the non-AI crises is the addition of a Crisis Strength setting in game setup, replacing the old setting to turn endgame crises on or off. It also replaces the scaling to galaxy size and habitable worlds, and has a default setting for each of the galaxy sizes. This setting allows you to control the strength of crises, all the way down from 0.25x of their base power to a massive and likely unstoppable 5x power boost to their fleets. As before, you can also turn off crises entirely.
2017_06_01_2.png


Additionally, we've also spent a considerable amount of time improving the crisis AI, both in terms of how the crises themselves behave and how regular AI empires react to them. Crises should now expand in a more logical fashion and be better at defending and fortifying the space they have taken over. AI empires, in turn, should be far better at understanding when they are under mortal threat and react to a rapidly spreading crisis by banding together against it and coordinating their fleets to fight it.
2017_06_01_3.png


The Contingency
The old AI rebellion crisis suffers from a number of issues, mostly stemming from the fact that it's so different from the other crises. While the Extradimensionals and Scourge are large invasions that have to be fought with fleets, the AI rebellion is supposed to be primarily an internal crisis, with the dangers stemming from infiltration and subversion rather than outright warfare. The problem with this is twofold: The game mechanics do not support it, and it is inherently unsatisfying. Whereas huge fleets roaming around scourging the galaxy of life is an easily understood threat that can be fought by empires coming together and pooling their resources against the invaders, the AI crisis mostly ends up as a series of frustrating events affecting empires in isolation, or 'Spaceport Destruction Simulator' as it's been called.

In addition to the gameplay problems, there is also the narrative problems: Why exactly do rebelling synths pose a galaxy-wide threat? If sapient machines are so powerful, why are ascended synthetic empires not on the power level of an endgame crisis? Even if we were to simply boost the AI crisis by giving it massive fleets, this really doesn't make much sense that a handful of rebelling synths from a handful of regular empires were able to amass such fleet assets in the first place. It's for this reason that we decided to go back to the drawing board and remake the AI crisis in the mold of the other two endgame crises, while retaining as much as possible of the 'synth infiltration' flavor from the old crisis. Enter the Contingency.
2017_06_01_1.png


Without wishing to spoil too much, The Contingency is an ancient AI whose purpose appears to be to sterilize the galaxy of all higher biological life and control or destroy all other Synthetic life forms. At the start of the game, it is dormant, broadcasting a weak signal across the galaxy that affects Synthetics in unpredictable ways. The chance of the Contingency waking up is directly tied to the prevalence of Synthetic life in the galaxy, and should it wake, it will attempt to use its signal to control Synthetics and force them to aid it in its implacable task of galactic sterilization. Unlike the previous AI crisis, the Contingency has formidable fleet assets with which to carry out this task and has to be fought both in space and at home, as it makes use of subversion and infiltration to soften up its targets before the sterilization units arrive.
2017_06_01_4.png


Just as with the Extradimensionals and Scourge, there is additional events and hidden lore to be discovered regarding the Contingency, and synthetic empires will have special interactions and challenges related to it. The Contingency completely replaces the old AI uprising crisis, but we are currently looking at also implementing a new AI uprising, not as a galactic scale crisis but as a midgame event localized to one or a few empires. But more on that later!
 
I was suggesting it as an option really. You could still have the option of fighting them if you'd prefer that. It's like a choice between a short term "spend all our resources now to try to beat them with ships" or a long term "hope to beat them with research before it's too late (and gain other bonuses along the way)". You also have to remember that Stellaris is partially about generating an interesting story or series of events. Sending a big fleet to beat an end-game crisis isn't really a satisfying solution, but gathering whoever you can in an attempt to find a solution through research or a galaxy-wide project would be and there are lots of things you could have occur along the way. What i suggested would also be a mix between the two, where you're trying to hold them off and delay them somehow.

The things i think of with the end-game crisis stuff is the Tyranids from Warhammer 40,000 or the Ori from Stargate (or the Replicators). A powerful, galaxy changing force that is extremely difficult to overcome with conventional means. With Tyranids it's a case of temporarily pushing them back despite great losses, hoping that at some point a way to beat them will be found but until then you just have to do anything possible to slow them down. With the Ori in Stargate the fleets of multiple of several of the most advanced races in the galaxy banded together (even enemies) to try to stop them before they could gain a foothold, but that ended up being entirely useless and they were only beaten through a mix of technological advancements and undermining their very foundations (their religious beliefs). The Replicators were beaten by modifying a superweapon to wipe them all out at once. Those are the sort of interesting events that Stellaris needs more of, not more "Throw a giant fleet at them and win".

This actually kinda reminds me of the Prethoryn, to a degree. How you can take a queen, tame it, and get an immortal Prethoryn Admiral, and ships produced from the queen. That would be pretty cool if they added something similar to The Unbidden and The Contingency. The Unbidden give you disintegrators, sure, but it would be cool to have some research that gives you something more substantial in the long run.
 
So you give slaves, or cause a reduction in their shields or something? For the slaves thing, that's illogical as The Contingency wants to just eradicate all life anyway. Even if it didn't, what good would that do? You'd essentially willfully vassalize yourself to The Contingency and... Sit there and wait for someone else to throw a doom stack at blowing it up? It just seems impractical and, honestly, like less fun than blowing them all up for those examples you gave. Those things sound more like mid-game crises. For an end-game, I like fighting off the big bad(s). If this game had a culture spreading system, similar to Sins of a Solar Empire, then that could open up some other options. Otherwise... I don't want to sit and wait for x number of months or years for a fricking science ship to do research lol. Heck, how would you even protect the ships from enemy fleets if you don't have a doom stack anyway? And if you have one, why not just use it?

Im not suggesting we use these ideas im not saying they are the best they were literaly a spitball i typed as i was thinking them up. But the whole point i was making is there is multiple ways this can be handled that does not involve another doom stack. If i can come up with crap ideas in 5 minutes im sure a small team working 8-10 hours per day can come up with something.
 
A crisis that grows from an insignificant threat is one the player can and will immediately snuff out.

Not much of a crisis at all, in other words.

No this is disappointing. There were ways to make the AI Rebellion sturdier without abandoning the mechanic. You could have made the rebellion spawn on every planet with a robot/droid/synth, allowing smaller empires to crumble and for those successful uprising to link up, even if bigger empires were able to contain their uprisings early on.

You could have buried a technology rating for each planet, based not only on robotic-type pops, and robot/android armies, but also on how many tiles have been developed and at what level. This rating could have been used to determine the scale of an AI rebellion's severity on a given planet, even spiritualist planets could have a small percentage chance of rebellion if their tiles were highly developed.

You could explain the lore away by saying synthetic-ascended empires still have a basis in emotional biology and are less likely to operate with the ruthless efficiency of an AI rebellion. Even say Synthetic-Ascension empires aren't targeted by AI Rebellions. They could certainly use more pizazz/bonuses.

You could have better developed flags for rebellions that succeeded so that the AI attempted to consolidate towards each other.

And yes, you could have tweaked and buffed it in all of the above parameters until it wasn't so simple to snuff out.

You guys gave up. This 'fix' comes from a lack of imagination. I only speak for me personally, but this AI crisis is what sold the game for me, and it is bitterly disappointing to see it tossed aside a year later.
 
Im not suggesting we use these ideas im not saying they are the best they were literaly a spitball i typed as i was thinking them up. But the whole point i was making is there is multiple ways this can be handled that does not involve another doom stack. If i can come up with crap ideas in 5 minutes im sure a small team working 8-10 hours per day can come up with something.
Your ability to come up with crap ideas is not in any way related to their ability to come up with a good idea - especially one that would actually be fun or engaging.
 
Seems to me that the current best and easiest fix for the corvette swarm issue is to not build a corvette swarm.

No, the best and easiest fix for the corvette swarm issue is to make exceeding the fleet cap much, much more punishing. The marginal tech progression built into the game only makes sense if you need to get the most out of each unit of fleet. Since the penalty is so weak right now, that completely breaks down.

Telling someone not to use a dominant strategy doesn't mean the game is any better designed. Players should not have to intentionally gimp their strategy to have an enjoyable and challenging experience.
 
No, the best and easiest fix for the corvette swarm issue is to make exceeding the fleet cap much, much more punishing. The marginal tech progression built into the game only makes sense if you need to get the most out of each unit of fleet. Since the penalty is so weak right now, that completely breaks down.

Telling someone not to use a dominant strategy doesn't mean the game is any better designed. Players should not have to intentionally gimp their strategy to have an enjoyable and challenging experience.

If we are talking about an MP experience, and you can rule over other players with a exploitative strategy, I agree. Otherwise playing single player and complaining about this seems more like a self control problem.

And truthfully, we have people here who have argued that the strategy is unbeatable while others say this is not true.

I haven't even tried it. I saw the Youtube video that Aspec put out there where an all corvette fleet beat a couple of all BB fleets of equal firepower. *shrug* Okay. There's a situation you won't normally run into. The AI is not going to come at you with a 500 ship corvette fleet. At least I haven't seen it yet in almost 800 hours, but stranger things have happened I suppose. And I guarantee you that I'm not going to create a 500 strong corvette fleet anytime soon.

Hence, my lack of angst over this.

YMMV.
 
No this is disappointing. There were ways to make the AI Rebellion sturdier without abandoning the mechanic. You could have made the rebellion spawn on every planet with a robot/droid/synth, allowing smaller empires to crumble and for those successful uprising to link up, even if bigger empires were able to contain their uprisings early on.

You could have buried a technology rating for each planet, based not only on robotic-type pops, and robot/android armies, but also on how many tiles have been developed and at what level. This rating could have been used to determine the scale of an AI rebellion's severity on a given planet, even spiritualist planets could have a small percentage chance of rebellion if their tiles were highly developed.

You could explain the lore away by saying synthetic-ascended empires still have a basis in emotional biology and are less likely to operate with the ruthless efficiency of an AI rebellion. Even say Synthetic-Ascension empires aren't targeted by AI Rebellions. They could certainly use more pizazz/bonuses.

You could have better developed flags for rebellions that succeeded so that the AI attempted to consolidate towards each other.

And yes, you could have tweaked and buffed it in all of the above parameters until it wasn't so simple to snuff out.

You guys gave up. This 'fix' comes from a lack of imagination. I only speak for me personally, but this AI crisis is what sold the game for me, and it is bitterly disappointing to see it tossed aside a year later.

I completely agree with this. Wiz didn't just give up, he removed it because he doesn't like it personally. He didn't like the mechanics it used so chose to remove it from the game instead of fixing it. The AI rebellion was one of the better concepts in the game. After Wiz took over as game director he chose to ignore the AI rebellion, letting it be broken for months before removing it altogether. If I was one of the people who were originally working on the game I'd be pissed.
 
I completely agree with this. Wiz didn't just give up, he removed it because he doesn't like it personally. He didn't like the mechanics it used so chose to remove it from the game instead of fixing it. The AI rebellion was one of the better concepts in the game. After Wiz took over as game director he chose to ignore the AI rebellion, letting it be broken for months before removing it altogether. If I was one of the people who were originally working on the game I'd be pissed.

They said they're going to try to keep it as a mid-game crisis. I've never even seen it once in all the games I played, and even if I did I doubt it'd be a problem for me. I tend to play something that gets no robots, anyway, so it would really only soften up enemy empires for me to take over (if it ever did happen). Doesn't seem like it'd be so end-game to me lol
 
They said they're going to try to keep it as a mid-game crisis. I've never even seen it once in all the games I played, and even if I did I doubt it'd be a problem for me. I tend to play something that gets no robots, anyway, so it would really only soften up enemy empires for me to take over (if it ever did happen). Doesn't seem like it'd be so end-game to me lol

No, they said they want to add it back *later* as a mid game crisis, for now it's just gone. No mention as to when they would be creating the new mid game crisis. It could easily be another year before they decide to create any mid game crisis. Also, this sounds like something that would end up in a DLC expansion of the mid game from Paradox. I'm not going to buy DLC to get the crisis back that they removed from the game. If they do eventually add it back in (I'm not holding my breath) it better be a free patch.
 
Cheers for the DD Wiz :). The game is defintely (imo, but it's the only o I have :) ) at the stage where refining the crises is a good next step. The Contingency sound interesting, but at some point it'd be nice to have a crisis that interacted with other gameplay elements (the new factions, for example). Not suggesting you're not already thinking of this, and not suggesting it should be done before refining existing crises, but at the moment it seems the sensible way to play to deal with crises is to snowball as much as possible and get a great big fleet. There's nothing wrong with this, but it'd add depth and replayability, say, if taking that approach sometimes lead to bad outcomes instead of good (say, for example, an internal factions crisis in a very large empire could lead to the fleet splintering - you could even have cross-empire factions that splinter at a similar time, combining their fleets - such that a lot of fighting the crisis is trying to stop the faction rebelling in the first place).

This'd go double given the somewhat limited-dimensional nature of Stellaris' combat at the moment (there's no logistics, trade or movement of strategic resources to interfere with, for example - things that are both things (even if simple) in EU4 and HoI4 - losing a planet is just losing that planet, rather than a galactic trading hub, and there's no way of interdicting a strategic resource, say). There's some interesting stuff happening within fleet battles (although I'd argue it could be surfaced better through the UI), but war is just fleet battles and capturing/losing planets.

Not having a go - am a big fan of the game - just throwing ideas around in case helpful, ignore if not :).

A standard 1x crisis fleet is about 60-80k fleet power currently.

Any thought of including information in a tooltip over the slider to give some indication to a player? You may well be well on top of this, not suggesting you aren't :).
 
If we are talking about an MP experience, and you can rule over other players with a exploitative strategy, I agree. Otherwise playing single player and complaining about this seems more like a self control problem.
"Exceptionally effective strategy is simple, boring, and ignores huge swathes of the game's feature set" is a serious problem in any computer strategy game, whether single-player or multi-player.
 
That's neat, but when will the game be changed so that it's ever worth building anything but naked corvettes?

It was many updates ago. To clarify, If you're talking about the power of corvettes in relation to other ships, bigger ships will always win in an equal battle (equal in terms of naval capacity)

However, If you're talking about the OPness of Corvette maintenance fee in relation to fleet capacity, I'm sure they will fix that soon.
 
Last edited:
So let me get this straight. Because a video was posted that seemingly showed a ridiculously huge navy of nothing but "naked corvettes" defeating a fleet of, was it BB's?, with an "rounded" equal fleet strength number, there are a large number of players who will never play any strategy other than the naked corvette strategy ever again in a single-player game.
 
So let me get this straight. Because a video was posted that seemingly showed a ridiculously huge navy of nothing but "naked corvettes" defeating a fleet of, was it BB's?, with an "rounded" equal fleet strength number, there are a large number of players who will never play any strategy other than the naked corvette strategy ever again in a single-player game.

Not an equal-sized fleet in terms of naval capacity. If 500 cruisers went up against 2000 corvettes the cruisers win, etc. The same applies to all sizes. It's a naval capacity/maintenance exploit and has nothing to do with the strength of a corvette.

The strength of naked corvettes is so low that equal fleet powers would be around 500 cruisers to 4000 corvettes. In that case the corvettes would certainly win.

I wonder how much more effective this strategy could be with battleships? Someone should try it and report back; I bet the results would be twice as high. Although, the fleet maintenance exploit may not work with battleships.

In other words, corvettes are still weak. However, spamming naked corvettes (the way you have them when you start the game) is an incredibly OP fleet design. They each run at about 0.25 compared to a fully optimal cruiser's 4-6. That's 16-24 corvettes per cruiser. The corvettes will of course win. Since their cost is so low exceeding naval capacity isn't an issue either. So, as a result, this strategy is being spammed in MP and everyone is mad.
 
Last edited: