In a thread at FeartheSword commenter S.C. complains that FTs late in a Boston game can cause net rating to get "jacked". Is this plausible?
In the game under discussion the Celtics had a 1 pt lead with about 1 min to play and ended up winning by 9 in part thanks to desperation fouling. How much could this effect net rtg over the course of a season? Well 8 pts over 82 games is worth less than 0.1 pts in final differential, so for a team to experience a substantial increase in their net rtg due to circumstances like these they would need 10+ events worth 8 pts to give them a fake improvement of a full point. Of course this assumes that they are never on the other end of the scenario, so they need to be on the winning side 10 more times than the losing to gain a full point. Even if they were very good/lucky and had a 2:1 ratio in their favor they would need to play 30 games a year that ended with desperation fouling. Not particularly plausible.
How about the 8 pt swing? Is that plausible? Well, no. Only 4 of those 8 pts were due to desperation fouling, the Celtics and Pacers had "normal" possessions until the Pacers missed a shot with 29 seconds left and started fouling. Realistically for a team to jack their net rating by +1 due to these games they would have to play them 60 times a year, and be on the winning side at a 2:1 ratio.
There are also lots of other potentially odd situations late in games that "effect" net rating, large leads where one team pulls its starters before the other, and teams dribbling out the clock and not working for a shot generally push back in the other direction. Anyway you look at it though it is very hard for these things to substantially alter a full seasons play because they are not common enough nor are they perpetually one sided.