It there a glitch in the RNG programming ? Maybe the rand() seed value is having refresh issues ? Last 4 primary engrams became Questing Beast Last 8 heavy became Dilluvian.._. If the seed value was based on something non random like my IP, then a mistake in code was refreshing it before every rand() call we might see that kind of pattern Alternatively the RNG math library is badly coded and there are excessive symmetries in the distribution Can you have one of your Coders dig into it?