Like it was ever NOT one, Lestat?
Remember that Draconi discovered before he left UO, that the RNG being used for UO since Beta or earlier was a public domain one that was commonly used as a college programming assignment. It's also commonly found in WAYYY too many Facebook games.
Said RNG is "fair" over a huge number of samples, but if you look at very short-term samples (over seconds, or a few minutes), it tends to streak because too many of the seeding components are the same for a given source. Granted, the streaks happen at ALL points on the curve, but that takes thousands of samples to balance out.
Examples most of us have experienced are things like failing 6-8 times in a row on a 95%-99% smelting chance, a string of misses on a target you shouldn't miss once at (or getting repeatedly hit by something that should have a miniscule change of hitting), getting the exact same fish 4-8 times in a row, pulling 18 BODs and having 6+ of them being the same normal weapon type (and most of them the same count), etc.
He was supposedly working on a new one, but I never heard for certain if it was ever implemented (right around the time he was supposed to be implementing it, was when he was let go in that huge round of layoffs)