Episode 5 | NBA + team statistics

Finally, a new sport. This is a timely video as the NBA playoffs have reached the conference finals, and we will soon have a champion. The question I sought to answer through my various analyses was: which stats (on the team level) impact winning across a season?

The question is simple, but as with all sports statistics, there is no such thing as simple. We began with counting stats, those that accrue over the course of a game like assists, rebounds, and turnovers. There were some surprises, such as assists not being significant (p = 0.41), and some glaringly obvious results (a turnover does not help the cause). The fact that Greg as a watcher and fan of basketball got the rankings so wrong, is an indication that basketball is complicated and misunderstood to some degree.

There are also some simple rate and sum statistics we went over such as stocks (blocks + steals) and 3-point percentages. Perhaps the least surprising of this group was the latter, where a percentage increase is worth 8 wins across the season. The surprise of this group was probably pace, which had a negative relationship to winning. Pace is often cited as a desirable trait (more possessions = more scoring, right?) but that does not appear to be the case. My speculation is that not all shots are equal and taking a few seconds off the clock to get a better one is more beneficial in the long run.

We also break down some of the playoff odds in relation to these stats and finally do a case study into bench points for a couple of teams. Bench points jumped out to me while entering data because of the cliff they seem to fall off between the regular season and the playoffs. To be expected but still some surprising figures. Hope you enjoy!