It was not just the 33 no-leg run, but also a 40 no-leg run before that, both with mixing.
But if the experimentation and the statistics are so air tight then guess I was just really unlucky lately. And I concede the point. My sample size was always small even though thought provoking. But there is also a chance things no longer work as they did, but I am sure someone with more data than me will bring that out if it is the case.
I also got very shit luck with Standard packs so that further proves those have no pity timer.