Not every situation is the same, obviously, but the more data points you have, the more relevant the collective information becomes. The assumption being that while every situation is different, the larger the sample size the more variations of these situations are captured to give a general trend. You can dismiss the outliers, you can debate whether 5 years of records is representative, you can even debate who qualifies as a "legend" but I think the results of this exhaustive research is relevant.
That is not to say SU cannot execute an outlier succession when JB retires.
If only identical situations can be harvested to use for these types of discussions, then we don't have much to discuss and speculate. Heck we can't even say a 30-5 season is better than a 18-13 season because those are different seasons, different teams, different opponents, different venues, different weathers, different times, different referees, different substitutions, thus all different situations.
That might be true in some circumstances, but definitely not in this one. Here, irrelevant data points skew the result toward irrelevancy. Just look at it at a micro view:
Danny Kaspar was .713 at Stephen F. Austin. His successor, Brad Underwood was .864 = +148 points
Danny Kaspar, Brad Underwood, and Steven F. Austin have zero relevance to a HC job at Syracuse University. Zero. Expand this by adding a thousand more instances/datapoints like this and you see that it doesn't do anything toward making
our situation more predictable. Ironically, we can't even say it makes the predictability worse—because the aggregate might just balance out, coincidentally. It's just an increase in the noise to signal ratio when you want the opposite—more signal (more 'qualified' data). [I've been working with data people for a relatively short time and i'm already talking and thinking like them. I hate it!]
As noted above, there are just too many factors that determine whether a coach can or will be successful. Facilities, reputation, playing style, opposing conference talent level, geography, fanbase, legacy, NBA profile, age and energy, sanctions, academic standards, value of the degree, incumbent position within a conference, headspace within a conference, conference shifts/realignments, proximity to impact players, how willing you are to bend the rules or cheat, how much rope the AD/fans/media gives you, weather, coed hotness, uniforms, player idols and role models... There are scads more, even before you get into the individual coach's personality, skills, and character.
