Defensive Rating specifically measures how the defense of a team performs. This can be at the team level - the Duke Blue Devils had a 97.1 Defensive Rating in the 2018/2019 NCAA season - or for each lineup, player, or situation within a team. This is accomplished by choosing some criteria (i.e. when James Harden is on the floor in the 4th Quarter) and dividing the points scored by the opponent by the opponent's total possessions. This provides a repeatable method for comparing and understanding how an offense performed or, if the sample is large enough, will perform in certain situations.
DefRtg = 100* [ (Opp Points) / (Opp POSS) ]
However, Defensive Rating is the surface level of defensive performance. How the opponent actually scored those points can vary significantly and can tell you more about the other team.
Shot types, avoiding fouls, forcing turnovers, and grabbing defensive rebounds all play significant factors in how a team allows points. As you may have noticed, those are the ever popular Four Factors and every defensive rating has a unique and easily understood signature.
For example, at this moment, Cassius Winston(Michigan State) and Tre Jones(Duke) have very similar defensive ratings, both around 96.6, but when we look at the four factors we see how different the paths can be to get there.
The Defensive or Opponent stats are in orange and it's quite clear that Duke(when Jones is on the floor) allows opponents to get to the foul line less(19.9 < 20.6) and forces more turnovers(18.5 > 14.6). Meanwhile Michigan State(when Winston is on the floor) forces a more inefficient shot profile(42.9 < 46.7) and cleans up the defensive glass at a slightly higher clip(26.6 < 28.0).
Offensive Rating in basketball is very similar to Defensive Rating in its creation and usefulness. You can read more about that here.
At Pivot, our focus is on understanding how a player or set of player impact the entire team's performance rather than his or her individual statistical prowess. Whether in our single player on/off data, the player combination matrix or the lineup explorer, Pivot Analysis provides players and coaches the ability to measure the effects of each individual and player combination on the whole team's performance. In the coming months, we will be adding functionality enabling the user to measure and analyze rebounds, turnovers, and shooting percentages (four factors and more) in the same fashion. The Pivot Analysis application will tell you when Player X is on the floor how many points a team scores, how often they turn the ball over, how well they shoot threes, how often they shoot free throws, and how well they crash the offensive boards.