AI opening L-gates too early
- Thread starter Antsassin
- Start date
-
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.
Good talk everyone. I think what it comes down to is allowing all flavours in the game. I like a lot of diplomacy, espionage and niggly, sneaky wars. I don't mind a surprise or two at all and I very rarely go back to a save game. But I have played a myriad of Stellaris games and I'm always prepared for the grey tempest, except about 150 years later. It has only been the last few games since the update that I start getting caught with my pants down. After hours and hours of intrigue, it just takes the fun out of the game. I have whipped both the Khan and the Grey tempest when they come at a reasonable time. And that trick of having a 30k fleet 30 years into the game... I would like to know how to do that. In the end, I like SirBlackAxe mods and the no Grey Tempest before mid-year or a scalable Grey Tempest depending on the year of activation. Definitely, not easy strength but a little less than impossibly apocalyptic would be good. Anyway fixed for now although I can't get into the L Cluster after the drakes appeared... it says unknown destination... I have encountered that before but I was always the first one to activate the L Gate... any thoughts?
We should simply be able to set a start year for L-Gate Insights. What if no one was even able to get L-Gate Insights till 2350? That would make the L-Gates an end-game crisis, which is more my style. I really hate the tempest, but I would not mind it as much if it arrived in the end game.
Ok so what exactly counts as midgame for you and what settings you are playing on?
Also at which years did the AI opens the L-Gates for you? I haven't seen an AI open the Gates before 2300 and not a alot of players unless they reyll B-Lined in a MP game.
I have been lucky and have never seen the gates open before 2300'ish. But I also have not actually rolled the Tempest either. But I have, without making any particular effort towards insights, managed to acquire all 7 in the 2235-2240 range before. So a sub 2250 opening happening from time to time isn't super hard for me to believe.
I like the scaling idea SirBlackAxe brought up.
You are a player. Sure I have been lucky as well and belined the techs as well. Not as early as you but I think I got it around 50-60 years in.
I have the AI only open them around 2300 at at the time you can get the Khan as well so I disagree that its a early game event.
I also think there should be a possibility to turn the L-Cluster off and itr should be tied into Midgame start year for them to be able to be openend (AI and player).
Midgame is 2050 and my latest AI L Cluster that was opened was in 2299 (a year before the Endgame)
It was surprisingly open by a fanatic spiritualist, not the Meritocracy/Technocracy AI
I always like a long game and time to set myself up. So I set the date at 2500. And I agree to be able to turn the L-Gate off or maybe reset the odds so it is not always likely to get the Grey Tempest. Some of the other options mentioned I have never seen and I have played lots of Stellaris.
Apparently, this can happen much earlier, as anecdotal evidence from this thread suggests. A particularly (un-)lucky AI can manage it. Players as well, but they will generally hold off.. Settings where the galaxy is more empty (= more L-cluter insights per player to find in anomalies) could also facilitate this.
But you should also note that the player can also submit to the Great Khan temporarily. The Khan is a crisis that can be negotiated with, and it ends with the Khan's death. Unlike the Tempest, where an early occurance just screws you over.
If the L-gates were locked until midgame (e.g. 2300 in default settings), that would be fine-ish. The AI is just too foolhardy when opening them.
The point of my post was that I was not bee-lining at all. It was literally pure happenstance. Let me check when I had six I know it took a long time for me to find the seventh. Ok loading up some of the saves from that game. Somewhere between 2222 and 2225 I picked up insight number 6. All while only scanning my local area. The only thing I did that would be abnormal for the AI is build a lot of science ships. I always tend to build between 4-10 depending on galaxy size. Speaking of Galaxy size I think that was an 800 star system with default settings and I had scanned by a quick count 70-80'ish systems by 2225. Again probably more than your average AI would manage, but none of that is strictly bee-lining the L-Gate. I do the same thing pretty much every game and my 7 insight finding can vary from 2235-2240 to never unless I start buying them up. In the one game I had to start bee-lining it in the mid 2300's because I learned someone was very close to unlocking it and I wanted to be the first one in(so that I would be prepared if it went tail end up on me). Looking at the saves from that example it looks like at 2346 I had naturally found 4 of the 7. So yeah lots of variability in this event's timeline.
That said I wouldn't call the L-Gate 'early game content' just because it can potentially fire early, since it clearly isn't designed to be handled then. Rather I would refer to it as a poorly gated event(pun intended). If it scaled that would just be interesting since the Tempest could be a weak one fire in the mid 2200's or if it fired late game it could rival or surpass the endgame crisis in difficulty. But it doesn't scale so instead it just is frustrating for those rare games unlucky enough for the AI to get lucky with insights.
- 1
Just found something on the Stellaris wiki on this page:
stellaris.paradoxwikis.com
Specifically, under the heading "The following events take place after Mid-Game:" I found this gem:
"AI empires discover L-Gate clues faster"
So it appears there already is a mechanism for player control for L-Gates, but it's either bugged or requires some fine-tuning.
EDIT: Not to mention there's nothing PREVENTING AIs from gaining clues sooner.
Game settings - Stellaris Wiki
Specifically, under the heading "The following events take place after Mid-Game:" I found this gem:
"AI empires discover L-Gate clues faster"
So it appears there already is a mechanism for player control for L-Gates, but it's either bugged or requires some fine-tuning.
EDIT: Not to mention there's nothing PREVENTING AIs from gaining clues sooner.
- 1
Relax. The AI can not realistically open the L-gates before 2280-2300, at the very earliest. That the player can do this before is irrelevant.
By 2300 you can deal with a tempest fleet. There is no problem here, other than watching the rest of the galaxy burn if you, the player, choose to open the gates at 2260 or something like that.
And it is so sweet to watch the galaxy burn.
By 2300 you can deal with a tempest fleet. There is no problem here, other than watching the rest of the galaxy burn if you, the player, choose to open the gates at 2260 or something like that.
And it is so sweet to watch the galaxy burn.
- 1
- 1
The premise of this discussion is flawed.
It is not the AI that is opening the L-gates too early. It is the player who only has a couple of 5k fleets by the time an AI empire is able to open the gates, who is managing his empire very poorly. You are asking for disaster by neglecting your defenses.
- 4
- 1
If you're maintaining a fleet size that is consistent with the immediate threats around you (usually AI empires), that is not "neglecting your defenses." Why would someone build up and pay maintenance on a megafleet when it's clearly not needed? That's poor economic planning. The only reason the player would know to build one is because they've played the game before and already know this event might happen, which is effectively metagaming. I like to be able to play the game and enjoy it without metagaming. While knowledge of the Khan event and the FE awakening is also metagaming of a sort, it at least has clear controls in place so you can meet those challenges when your empire has naturally matured to the point where you could indeed build that megafleet as needed.
I get that things like the grey tempest are supposed to be a challenge. I don't want to eliminate challenge completely from the game, I just want it to not be so brutally challenging that it requires foreknowledge of the event to meet that challenge effectively.
- 5
I have asked for dates a couple of times and so far I have only Hunnuli saying its along 2300. No one else put a date down or provided a screenshot (thats what I would call evidence).
I was always under the impression the AI wouldn't open them before midgame because I never saw it happen earlier. This may have been due to my settings hence I am was always asking for peoples usual settings.
Having a game ruined by Tempest is surely annoying but I really doubt it's the standard. It's also definetly something you can prepare for.
Rather than using console commands, you can also directly check and even edit the save file. It's just two zipped text files; the wiki has an article on it: https://stellaris.paradoxwikis.com/Save-game_editing. The relevant file is the larger one (there will be no difficulty telling which it is, I promise) and it contains a section for global flags. I recommend using search rather than trying to read the file. The relevant items are:
- dragon_season - L-drakes
- gray_goo_empire_set - Dessanu empire
- gray_goo_crisis_set - Gray Tempest
- <none of the above> - Gray (nanite entity)
The random weights (rolled at game start) are 50 for the Tempest and 30 for each of the others, which means Tempest is 2/3 more likely than any single other outcome but still only happens 5/14s of the time (just over 1/3). If you wanted to mod this, the relevant event to override is distar.8000; the standard version is in events/distant_stars_events_3.txt in your Stellaris folder.
Just for the sake of completeness, I would like to introduce a mod that offers this dynamic conditions and many setting options.
I've counted the vanilla anomalies which gain L-Gate clues, 27! With repeatable tech, minor artifacts and Curator enclave (30) there is indeed no real obstacle for a too early L-Cluster opening.
steamcommunity.com
Includes:
I've counted the vanilla anomalies which gain L-Gate clues, 27! With repeatable tech, minor artifacts and Curator enclave (30) there is indeed no real obstacle for a too early L-Cluster opening.
Steam Workshop::A Deadly Tempest [3.7.☠]
Just higher requirements for the AI to open the L-Cluster added (only from the midgame and if the crisis is rolled).
You can also just separate the AI from opening.
The final project uses now dyn. 
).
AI_wait_days and requirement scientist leader skill = 3 (as this was already present in vanilla but commented out - maybe forgotten for test ).The AI_weigth for the repeatable clue tech is lowered and needs midgame
Last edited: