A team's personnel can be measured across 5 different categories:
Height: total height of a team weighted by minutes played
Bench Minutes: the number of minutes a team's bench plays
Continuity: minutes played season-to-season
Experience: year in school weighted by minutes played
Defensive Fingerprint: identifies the style of a team's defense
2-Foul Participation: percentage of time a starter with 2 fouls plays in the first half
Assumptions are made for each of these components, and detailed explanations are below.
It does have a slight correlation to a team's offensive and defensive efficiency. And it makes sense, a taller team might defend well and score at ease in the paint versus a shorter team. But it's not the only metric that matters.
This is measured by gathering the average listed height of every player on a team, weighted by the minutes played.
It excludes players that play less than 10% of their team's minutes.
Syracuse led the country in average height at 80.1 inches in the 2017-2018 season.
The shortest player featured in over 10% of the team's minutes was 6'5 or 75 inches.
This can also be broken down into positional height, meaning by each of the 5 positions on the court.
It takes the minutes played for each team ordered by height. For example, 20% of minutes played by the tallest players are likely the center's minutes.
For example, Purdue's average height at the center position was +5.4 inches taller than the average of all teams.
Purdue's starting center, Isaac Haas, was 7'2 or 86 inches, and their backup center, Matt Haarms was 7'3 or 87 inches.
Effective height is the average of the center and power forward position. The same Purdue squad posted an effective height of 3 inches for these 2 positions.
Bench minutes, defined by KenPom, makes an assumption that the starters of a team are the 5 players that have played the most minutes on the year.
The remaining minutes are assumed to be the bench minutes. If a player has played less than 10% of their team's minutes, the player is excluded in the bench minutes.
Alabama's bench played in 36.6% of its minutes in the 2017-2018. That ranked 49th in the country. In 3 of its 5 positions, players shared a large percentage of minutes played.
In contrast, Kansas' bench played in only 20.9% of its minutes. That's 345th amongst all Division-I teams.
Whether a team relies on its bench is more of an attribute than a driving factor of its success. It can certainly be a factor if there is foul trouble, injuries, or a stretch of several games played in a short period.
Continuity is the percentage of a team’s minutes that are played by the same player or players from last season to this season.
It answers more than are the same players playing this season?
It uses a percentage of a players minutes played because of injuries and a player's increased or decreased minutes season to season.
KenPom takes each player on the current roster, uses the minimum minutes (on a percentage basis) he played comparing this season and last season, and then takes the sum for all players.
Let's use the 2015-16 North Carolina team as an example. Below are the minutes from the 2014-2015 season to the 2015-2016 team. The far column includes the minimum value.
Player | %Min 14-15 | %Min 15-16 | Min--------+---------------+-------------+-----+Paige | 16.5 | 13.4 | 13.4Johnson | 12.2 | 14.0 | 12.2Jackson | 13.3 | 14.1 | 13.3Hicks | 7.35 | 9.00 | 7.35Meeks | 11.2 | 8.48 | 8.48Berry | 5.20 | 15.4 | 5.20Pinson | 3.92 | 9.33 | 3.92Britt | 7.60 | 7.50 | 7.50James | 4.99 | 3.61 | 3.61Maye | 0.00 | 2.22 | 0.00Williams| 0.00 | 1.52 | 0.00White | 0.30 | 0.40 | 0.30Coleman | 0.24 | 0.21 | 0.21Coker | 0.00 | 0.20 | 0.00Dalton | 0.03 | 0.20 | 0.03Egbuna | 0.00 | 0.24 | 0.00--------+---------------+-------------+-----+
The sum of the last column is 75.7. North Carolina players played around 76% of the minutes from 2014-2015 to 2015-2016 season. That was good for 12th amongst all Division-I teams in the 2015-16 season.
In a landscape where the best teams often feature players headed to the NBA after one year, continuity is a fun metric to follow.
For example, below is Kentucky's continuity over the last 5 seasons:
Year | Continuity | Division-I Rank |--------+---------------+----------------+17-18 | 11.2 | 34816-17 | 30.2 | 32215-16 | 22.0 | 33514-15 | 49.6 | 17613-14 | 23.6 | 336--------+---------------+----------------+
Kentucky continuity is often low, but it doesn't mean they're not winning games.
Experience is another storytelling stat. KenPom makes the assumption that a player's eligibility class determines their experience.
Under this assumption, the following class has these levels of experience:
Year | Experience |-----------+------------+Freshmen | 0 |Sophomore | 1 |Junior | 2 |Senior | 3 |-----------+------------+
Experience uses minutes played similar to the average height calculation. Players that play in less than 10% of their team's minutes are excluded.
In the 2017-2018 season, Wichita State ranked 13th in the country with 2.36 years of experience. This squad featured 9 upperclassmen (seniors and/or juniors).
For a comparison, Kentucky ranked dead last in the country with 0.18 years of experience.
Does a team play man-to-man? Zone? Both?
The answer can be found in a team's defensive fingerprint. Kenpom uses a variety of stats to breakdown teams into 4 defensive categories:
Mostly man: team plays mostly man-to-man defense
Inconclusive: can't identify a specific defensive tendency
Some zone: team is in zone defense for a good amount of its defensive possessions
Mostly zone: team plays most its defensive possessions in a zone defense
KenPom describes defensive fingerprint as raw. It uses assist percentage, 3-point attempt percentage, free throw attempt percentage, turnover percentage, and defensive rebounding percentage.
It weights these stats to develop a profile, and place a team in a certain fingerprint category. For example, zone teams tend to have higher a defensive assist percentage and 3-point attempt percentage.
Syracuse is known for its 2-3 zone. It statistically is near the bottom of all teams in defensive assist percentage and 3-point attempt percentage.
In the 2017-2018 season, Syracuse ranked 351th in defensive A% at 73.8 and 338th in defensive 3PA% at a 44.5% clip.
When a player picks up his second foul in the first half, does he head to the bench?
2-foul participation attempt to answer this question and reveal more of a coaching philosophy around fouls. The measurement is the percentage of time that a starter with 2 fouls in the first half has been allowed to play as detailed in this blog post.
For example, starters for Virginia played 3.8% of minutes with 2 fouls in the first half in the 2017-18 season. Virginia, and coach Tony Bennett, ranked 331st in the country in this metric.
KenPom also includes a few other measurements:
total amount of time starters have 2 fouls
total time starters with 2 fouls are on the floor
adjusted 2-foul participation: accounts for backcourt players being allowed to play more
percentage of bench minutes