PER is a pretty good statistical measure. No, it doesn't track everything, and it doesn't isolate good man to man defense, but it does evaluate lots of little things.
I heard a scout once say that the reason some guys get praised for "little things" is because they aren't good at the big things. No, everyone won't be a big time scorer, but Garrett hasn't just been a poor shooter. He's been a poor overall offensive player. As a freshman, his assist to turnover ratio was 1.46. As a sophomore, it was 1.5. He is a solid perimeter rebounder, but not exceptional. His steal numbers are also solid, but not exceptional.
Garrett is roughly average as a player - he's a net plus on defense because of his length, rebounding and good man to man defense (he doesn't block many shots or cause tons of turnovers, so he doesn't rate exceptionally on that end), but he's a net minus on offense because he is a very poor shooter that isn't an exceptional passer or offensive rebounder.
Now, if Garrett made himself into a slightly below average shooter, he would be a net plus overall because he is good defensively, so you can afford a bit of a downgrade on the other end. Last night's performance was a good sign, although we need to see consistency.
But PER captures Garrett's value pretty well. He's a good defensive player and a bad offensive player. Put that together and he's roughly average. If he becomes only a below average offensive player (rather than bad) he likely becomes a pretty good player overall. If he becomes an average offensive player, he probably becomes one of the better players in the Big 12.
