clock menu more-arrow no yes mobile

Filed under:

Here come the tempo-free stats: Our annual head's up on basketball coverage

Andy Lyons

With basketball season fast approaching, it's time to get my yearly warning/head's up out of the way: I use a lot of non-traditional stats in my coverage of NC State basketball and the ACC. None of them appear in a box score, but they're all built on top of that foundation.

They're called tempo-free stats, which I like to think of as necessary tweaks to the statistics everyone is used to seeing. The basic tenet behind their use is this: college basketball teams play at different speeds, which affects the number of shots they take, make, and rebound; it affects how many shots their opponents take, make, and rebound; it impacts turnover totals, foul totals, assist totals, steal totals, and on and on. This creates distortion in counting statistics, and over the course of a season, the level of distortion can be significant. Here's an example:

The difference between the fastest and slowest team in I-A last season was around 20 possessions per game. Northwestern State (No. 1) was runnin' its way to 79 possessions per game, while Miami (No. 351) was bogging games down to the tune of 59 possessions per game.

Over the course of the season, Northwestern State played about 500 more possessions of basketball than Miami! That's a lot of additional opportunities to score, rebound the ball, dish out assists, foul or be fouled, turn the ball over or force turnovers, etc. In an extreme case like this one, tempo-inclusive per-game statistics will obscure reality.

Let's say both teams averaged 15 turnovers per game last season. That statistic implies that these teams were equally adept at taking care of the basketball.  But one of them turns it over 15 times for every 79 offensive possessions, or 19% of the time; the other turns it over 15 times for every 59 offensive possessions, or roughly 25% of the time. By changing the denominator in the equation from games to possessions, we've taken tempo out of the mix, letting us compare the teams on equal footing. Once there, we see that Miami was far more prone to turnovers than Northwestern State (in this hypothetical case).

We can apply that line of thinking (with similar math) to a lot of statistical categories. In my game previews, I'll be using tables like this at the team level:

App State Offense 12-13
Four Factors Percent Nat'l Rank
eFG% 49.6 117
TO% 19.4 137
OR% 30.3 220
FT Rate
37.9 112

The "Four Factors" are the basic components of winning and losing: shooting (eFG%), ball security (turnover percentage), rebounding (offensive rebounding percentage), and the ability to generate free throw attempts (free throw rate). The game boils down to these outcomes, which together determine a team's overall level of efficiency at the offensive and defensive ends.

Effective Field Goal Percentage (eFG%): (FGM +(0.5 x 3FGM)) / FGA
That's just a quick adjustment to account for the added value of a three-pointer, which regular field goal percentage ignores.

Turnover Percentage (TO%): Turnovers / Possessions x 100

Offensive Rebounding Percentage (OR%): (OR / (OR + DR)) x 100
Defensive Rebounding Percentage (DR%): (DR / (OR+DR)) x 100

FT Rate (FTR): FTA / FGA

These factors don't carry the same weight--Dean Oliver, who was a pioneer of these concepts, estimated that shooting percentage was about twice as important as turnover or rebounding rates, and somewhere between three to five times more important than free throw rate.

In addition to the Four Factors, I'll also be using offensive efficiency (points per possession) as a replacement for points per game and defensive efficiency (points allowed per possession) as a replacement for points allowed per game.

These and other stats like them can be applied at the player level as well, but I won't get into all of that here.

If you have any questions now or during the season, don't be shy about asking, either on the site or via email. I'm always happy to help as best I can.