Stumped by the disappearing home-court advantage in the NBA?
So were we. We proposed some theories in this space here, but we decided to get our hands dirty and dig into the data.
Here's what we found. Let's tackle this FAQ style.
OK, so how much has home court really mattered?
Simply put: a lot. If a picture is worth a thousand words, then the following chart speaks volumes:
As you can see, the home-court edge has more or less been in straight decline over the past few decades, but it typically hovers just above 60 percent. However, this recent two-year drop has been enormous. So far this season, teams are winning just 53.7 percent of the time. That's down from 58 percent last season and 61.2 percent the season before that. Since 1975, the advantage has been an average of 61.2 percent. The peak? In 1988-89 it stood at 67.9 percent.
You might have heard that home-court advantage is worth at least 3.5 points per 100 possessions in the NBA. And that was true for a long time. But not this season (or last). This season, home-court advantage has stood at plus-2.2. Last season, it was plus-2.6. The season before that: plus-3.2. If we adjust for pace, the two-year decline is 36.7 percent, or more than one-third. In just two seasons. We've never seen such a dramatic two-year decline.
Is the decline in home-court advantage statistically significant?
It is. The large observed drop in home-court advantage since the start of the 2013-14 season is something that we would only expect to observe by random chance less than 1 in 1,000 times.
In other words, it's statistically significant. (For those interested in some gory details: We can calculate the odds of seeing a sample mean of any given value -- in this case, the mean home-court advantage over the sample of 1,897 games since the start of last season -- because the distribution of all possible sampled means is approximately normal with a standard deviation inversely proportional to the square root of the number of games in the sample. In a database of thousands of games dating back to 1996, we observed a per-game standard deviation in home-court advantage of 14.1, which yields a standard deviation of sample means equal to 14.1/(1897^0.5), which works out to a standard deviation of 0.324 points/100 possessions. So, the recent drop in home-court edge -- more than 1.0 point per 100 possessions -- is a shift of more than three standard deviations in magnitude, which is an event that occurs by random chance less than 1 in 1,000 times.)
Using a similar methodology, we also found statistically significant declines in the home-court edge in effective field goal percentage, offensive rebounding and free throw rate since the start of the 2013-14 season.
How many teams are worse at home?
One-third of the league. No joke.