AI characters can try to achieve Golconda depending on their personality and such. They can fail and if they do, they become wights

Could you shed some insight into the system? I did some more testing after I posted this to see if I could figure out how it works. By briefly swapping to Catubodua before she turns into a wight, she doesn't appear to have an active Golconda journey. If you skip the date, she doesn't turn but it will trigger again in about 5-10 years. I saw her become a wight on 3 separate triggers (after swapping to her, switching back, and waiting for a new trigger).

With reloads included, she had a 100% failure rate for that playthrough for over a dozen attempts. I'm curious if all npcs use the same Golconda system, or if Catubodua is unique because she starts with the seeker trait and a set amount of xp.

The wight trigger does seem to be related at least, because if you give her the Golconda trait she won't become a wight.

Thanks for your time and insights!

AI can't use journeys so it's a separate system that makes them try for Golconda. Catu is likely to rush before she is ready and thus fail (and become a wight) because of her traits, which are the main modifier for AI success or failure in this.

If it is that important to you then maybe change her characer traits via dominate. Maybe give her calm, forgiving, patient, generous?

Thanks for the responses and insights. I'm more curious about how the system works, since seeing that the date and outcome seem to be seeded far in advance is unexpected. In the first case, the outcome on Jan 7 1246 was consistent, even testing from a reload 12+ years ago. I'm guessing the AI chose to roll the Golconda trigger over 12 years ago, and the time spent for it to resolve uses some calculation?

Because in another case of testing where I skipped the Jan 7 trigger, the next one only took about 5 years (also a failure). Skipping that one too, the next trigger was about 10 years later (also a fail).

I would have expected smaller ranges for dates and the outcome to be rolled closer to the triggering day. Of course, this is from a sample size of slightly over a dozen, so maybe it's not a 100% fail chance in reality. Just seemed unusual given how CK3 usually rolls from my observations.

Edit: I did some more testing on a fresh save. As near as I can tell the outcome and date for Catubodua seems to be determined as soon as you start a game before any time passes. On my new save the date is June 20 1239 and she succeeds. To test, I made a save as soon as I started my character, played until an outcome, noted the date, and then restarted from my day 1 save and let the game run. Sure enough, June 20 1239 the same Golconda trait triggered. She also starts at rank 3 Golconda with 75 xp and jumps to completion if she succeeds the trigger, so I'm guessing she's unique unless that's standard for npcs.

AI golconda seekers get a hidden maintenance event that triggers every 10-25 years. it gives them 3-7 golconda seeker xp, and, if they feel "ready", they will attempt the suspire

feeling "ready" is based on their golconda seeker xp vs their ai_boldness stat (the same one that determines whether your dread will intimidate or terrify them). maximum boldness will make them attempt the suspire at >= 50 xp, minimum boldness will make them wait until 100 xp

they will NOT attempt the suspire if they're in a coterie with a non-AI character, so that's the way to stop them

how RNG works in ck3 is unclear but the seed seems to be based on the game's checksum, so things like that get rerolled every time the game updates. in my current playthrough on the dev build, catubodua has succeeded