Having gotten stuck in the project that I am working on at the moment, I decided to pass the time and heal my frustrations by playing a little bit with Roman Empire Data.
What I did was to match data for the duration of the Roman Empire with a few dummy variables on how each emperor rose to power. The data being what it is, what I found is purely descriptive. It does not really add anything that is particularly new, but at least it does not contradict common sense. Moreover, it is possible that by narrowing things down a little, it may hint at those things that matter most among those things we think matter.
It turns out that a lobbying mother, inheritance and practice make for longer rules while getting to power through the support of some general makes for shorter rules.