The other thing to keep in mind with KenPom and other metric-based rankings is that they do try to adjust for quality of opponent, but it's very likely somewhat imperfect and subject to a lot of game-to-game variance. Players/Teams/Coaches aren't AI SimRobots. Yet.
So let's look at the recent data for one of our recent quality opponents - Gonzaga. Here's their current KenPom rating:
Looking great, with the #6 offense, #19 defense. Meanwhile, we're currently "horrible" at #134 defense and have been in a freefall. Glancing at the KP "Strength of schedule" data for Gonzaga, it looks like "on average", they've played pretty quality comp overall...all average metrics for their comp are in the upper 20's for SoS (Net, O, and D).
But let's look at the
ORtg for the actual teams that they and we have played since our showdown in MSG...a great battle between 2 obviously quality teams who displayed very good offense and defense during that game [in general]:
For Gonzaga:
Nicholls - 239
Bucknell - 344
UCLA - 89
Pepperidge Farms - 211
Portland - 298
Loyola Marymount - 203
All of these teams they've played are also near the bottom quartile in 3pt% too (with the exception of UCLA, who is 103rd) so they're less likely to have a "hot" shooting game. As
@RuffRuff likes to say, Gonzaga has played a bunch of tomato cans since our game.
For UConn:
Xavier - 46
Butler - 86
DePaul -123
Providence - 107
Xavier, Butler, and DePaul are high-level 3pt shooting teams, much more likely to get "hot" no matter what the D does.
Rhetorical Question:
Which team do you think is more likely to keep their KenPom DRtg and/or improve it over that stretch of games? Which one is more likely to drop spots?