Originally Posted by
Des Esseintes
Here's the thing, CDu. You're saying Vermont has a garbage resume for a top-100 team and must therefore be overrated by the machines. But have you looked at what a top-100 resume actually consists of? Here's Louisiana-Lafayette, ranked one kenpom spot higher than Vermont. You mentioned that Vermont has ugly losses to Bradley #149 and Wagner #217, with little of import accomplished against good teams except Duke.
Among their six losses, Louisiana-Lafayette has lost to:
#224 South Alabama
#227 Louisiana-Monroe
#292 Jackson St.
Aside from several non-DI wins, Louisiana-Lafayette has beaten:
#119 Houston
#35 La. Tech (on the road! Go Cajuns!)
#306 McNeese St
#161 Oakland
#259 Coastal Carolina
#228 Northwestern St.
#194 UT-Arlington
#250 Texas St.
#245 Troy
How is this better than the Vermont resume? The Ragin' Cajuns have one good win over La. Tech, which I would argue is *almost* as impressive as a 1-point loss in Cameron. Otherwise, a wasteland. Not a single top-100 victory, the majority 200-level and lower. This is what a top-100 resume looks like. It's ugly. Vermont's ranking is not an indictment of kenpom. Moreover, your mistake is exactly why we let computers do this kind of work. It is impossible for people to eyeball schedule quality once we get into the scrum of teams north of about 70. For those of us conditioned to following an every-day-is-Christmas program such as Duke, at least. We have a skewed sense of what constitutes reasonable quality at that level. We might think this looks like nothing to speak of, but the fact remains there are 351 Division I teams, and this sort of record puts you in the top third. Maybe you will say Louisiana-Lafayette benefits from the Wisconsin effect, too. For my part, when the numbers keep saying something different from my preconceived notions, it's time to reexamine my preconceived notions.