Whether it’s College Football, Major League Baseball, or NBA Basketball, I put a lot of emphasis on looking at margin of victory as another way of measuring just how good a team and its players are. It doesn’t take a stats nerd to figure out that averaging more points than the other team will probably lead to more wins, but just how strong is the relationship in the NBA?
This season, margin of victory has been able to explain more than 93% of the fluctuations in team win-loss percentages (R-squared value). In other words, it is an excellent-but-not-perfect predictor. It is common to see this strong a relationship in team sports, and the chart below summarizes how many wins you can expect from teams at season’s end with the given margins of victory, give or take about 2 wins.
Obviously not every team follows this model perfectly. 7% of the fluctuation in win-loss records went unexplained, and there are a few teams that have strayed significantly from the model. Before we chalk it up to dumb luck, I thought I’d have a little fun with the numbers.
Logical traits of teams that over-perform their margins of victory would include teams that are experienced, teams that can control pace and limit turnovers, and teams that can defend. I used average team age (at the beginning of the season) to measure experience, opponent’s true shooting percentage and blocked shots to measure defense, and turnovers and possessions per game to measure pace. This new model suggests that each of this traits is at least a little important.
Specifically this season, San Antonio is out-performing its expected winning percentages by 10 percent, or about 7 wins. The Spurs are the oldest team in the NBA. In terms of defense, the Spurs are the 6th best at reducing opponents’ true shooting percentages. Conversely, the Minnesota Timberwolves are—if you can believe it—doing even worse than their margin of victory would suggest. Age? Youngest in the NBA. Turnover rate? Worst in the NBA. Defense? Seventh worst in the NBA. We might be onto something here.
These two teams are just the extreme examples, and the traits mentioned aren’t incredibly strong indicators, but likely play a measurable role in each team’s season (beyond just looking at margin of victory). There’s still room for luck to sneak in, and end-of-game coaching in blowouts probably plays a role, but there’s reason to believe that teams have some control over out-performing their margins of victory. So using margin of victory together with a little extra information can help us make even better predictions.
Good news for Blazer fans, nothing suggests we’re going in the opposite direction anytime soon. And Wallace seems to be getting used to us, if 40 points on 16-28 from the floor and no turnovers means anything…