Hello to readers and hopefully the developers of Crusader Kings 3,
This is posted both in the forums and in the subreddit to spread awareness of this unfortunate bug.
In Crusader Kings 3, in the blood legacy track for dynastic legacies, the 4th legacy is Architected Ancestry which says : "Select a Congenital trait to become more common in the Dynasty". You can either select "good" traits (comely, quick, hale, fecund) or "bad" traits (albino, giant, dwarf, scaly). The benefits of picking a good trait are obvious and the legacy works as intended for good traits.
However the intent of picking bad traits is to make your dynasty a certain type of "weird" appearance which is a great gameplay idea. Picking giant (which is going to be the example of my experiment) for Architected Ancestry is to make your dynasty a dynasty of giants.
The issue with using Architected Ancestry for making a dynasty of giants is that it comes after the 3rd blood legacy : Resilient Bloodline, which gives -30% chance of inheriting bad congenital traits. With the following experiment I've proven that the effect of Architected Ancestry for bad traits is hard-countered by Resilient Bloodline which is a requirement of Architected Ancestry. Before reading the details of my experiment giving numbers, if you're only interested for my personal recommendations, please go the last paragraph directly.
Here is the following experiment I've done in the game. One man and one woman are both giants and a couple. They are of the same dynasty but they are 9th cousins with no common ancestor except their ancestor 10 generations before. No ancestors are giants. They are unrelated in ck3 mechanics (no common great-great-great-grandparents).This makes this couple the "ideal" couple to test blood legacies on: same dynasty but unrelated and no noise from ancestors.
From this ideal couple of giants, I have conducted three tests. Each test was a hundred independent pregnancies and I mesured in each test how many of the children were giants.
Test 1 : the first test has the ideal couple with no dynasty modifiers and no dynasty legacies. This makes it a perfect result in a vacuum from two giant characters having children.
Result of Test 1 : 87 children out of 100 were giants.
Test 2 : The first three blood legacies (including Resilient Bloodline) are activated. The expectation of the test is to have much less giant children due to the Resilient Bloodline applying from the mother and father.
Result of Test 2 : 29 children out of 100 were giants.
Test 3 : Architected Ancestry with giant was activated.
Result of Test 3 : 39 children out of 100 were giants.The standard deviation of these three tests is roughly at 3.5 which means that every test result is distant from at least 2 standard deviations. This proves the impact of both Resilient Bloodline (between test 1 and test 2) and Architected Ancestry (between test 2 and test 3).
The results of these three tests show that if you want to play a giant dynasty, aiming to get Architected Ancestry is detrimental to that objective because Resilient Bloodline comes before and has a stronger impact than Architected Ancestry does. The intended objective of both the developers and the players are not achieved.
My personal recommendation would be to protect the Architected Ancestry trait from being applied a -X% chance of inheriting bad congenital traits. If the ck3 engine can't handle this, another idea would be to create a good version of those traits (Giant but good, Albino but good etc) and have the good version only available to dynasty members that have the valid Architected Ancestry (at birth, a dynast with a good or bad giant would get changed to good giant, and a non-dynast with good or bad giant would get changed to bad giant). This would ultimately fix the issue.