Tempo-free stats: Reviewing the "Four Factors"

Points per game is dead. So is rebound margin. These old-school statistics may be pervasive in college basketball discourse, but advanced statistics better explain what we are seeing on the court. The good news is that these aren't actually THAT "advanced." You can understand and use them too!

The place to start is with the "Four Factors," as identified by Dean Oliver. (Here's his full paper explaining his methodology.) Success in these four areas (on offense and defense) is the recipe to winning basketball games: Shooting, turnovers, rebounding, and free throws.

Shooting:

Factor 1 is the team-wide effective field goal percentage.

In basketball, making shots (and keeping opponents from making shots) is king. Rather than measuring shooting success with shooting percentage, using effective field goal percentage (eFG%) is an easy and more telling tool.

There are two different kinds of shots, two-pointers and three-pointers. The latter are worth 50% more. So why group them together? If Mike Tobey shoots 4-8 and Evan Nolte shoots 3-8, who did better? If all of Tobey's shots are two-pointers, while Nolte's are threes, then Nolte scored more points on the same number of shots.

eFG% is calculated as:  (Two-point field goals made + 1.5*three-point field goals made ) / field-goals attempted. This gives proper credit to those long-range makes.

Turnovers:

Factor 2 is turnover percentage.

If Virginia turns the ball over 11 times, while North Carolina lost the ball 13 times down the road in Chapel Hill, which team did a better job of taking care of the ball? If UVA had 60 possessions, while UNC had 75, then the Tar Heels did - they turned the ball over on 17% of possessions to UVA's 18%. Because teams play at different paces, turnovers is not a useful counting stat.

Turnover percentage (TO%) is calculated as turnovers / total possessions.

Rebounding:

Factor 3 is offensive rebound percentage.

There are two different types of rebounds, offensive and defensive. Defensive rebounds are much easier; over two-thirds of rebounds are earned by the defense. Therefore, a team that forces a high number of missed shots is likely to gain a high number of rebounds. Teams that allow a high shooting percentage (or force many turnovers or commit many fouls, thus ending possessions before a missed shot), will have fewer defensive rebound opportunities.

Offensive rebound percentage (oReb%) eliminates these external factors and actually tells us which team rebounded better. OReb% = offensive rebounds / (offensive rebounds + opponent defensive rebounds). In other words: what percentage of offensive rebound opportunities does a team convert? Defensive rebound percentage tells us the same thing on the opposite side of the court.

Getting to the free-throw line:

Factor 4 is free-throw rate.

Here, we aren't talking about free-throw percentage. Rather, this factor measures how often a team got to the free-throw line (and earned the chance to shoot those valuable free-throws).

Free-throw rate (FTR) is free-throws attempted / field goals attempted.

What do the Four Factors lead to? Points-per-possession:

A team's combined success in the Four Factors will determine its ultimate results on both offense and defense. Looking at scoring on a "points per game" basis adds another variable, tempo. Different teams play at different speeds, giving them varying opportunities to score. The best measure of a team's total offensive and defensive output is how well the team performed, on average, in each one of its opportunities.

Points per possession (PPP) is calculated as total points scored (or allowed, on defense) / total possessions. A possession is ended when the ball changes hands, whether by a basket, turnover, defensive rebound, foul, or some other violation. (Offensive rebounds do not result in a new possession).

Last season, the NCAA-wide average was 1.02 PPP. Generally, a team that scores over 1.1 PPP or allows under .9 PPP is doing a very good job.  We'll often report season-wide averages in Ken Pomeroy's "adjusted" offensive/defensive efficiency, which modifies PPP based on strength of opponents.

Where does UVA stand?

2015 Virginia Cavaliers statistics - Four Factors

Virginia's defense, ranked #1 in Ken Pomeroy's ratings of adjusted efficiency last season, is predicated on forcing opponents to attempt difficult shots, then cleaning up the rebounds. The team's low turnover percentage is an effect of the pack-line defense. Rather than take risks and potentially allow easy baskets, the unit is satisfied keeping its TO% low as long as the opponent's eFG% follows.

Offensively, UVA finished the year 24th in KenPom. The team did a great job taking care of the ball. Offensive rebound % has reflected an interesting philosophy shift over Tony Bennett's tenure, as he directed his teams to be more aggressive crashing the boards on offense. From 2010-2013 (and back to his WSU days), Bennett's teams brought in 30% of offensive rebounds or less, never ranking better than 268th in the nation. But the past two seasons, the team's OReb% was around 34%, and in the top-100. To improve its offense, Virginia must not only do a better job of making shots, but also get to the line more.

We'll be following UVA's performance in the Four Factors all season long, and using them to better understand the team's performance in each game.