SI Vault
LUKE WINN (with additional research by David Hess)
November 14, 2011
For a sport awash in stats, accurate measurements of defensive prowess are hard to come by. So SI did its own possession-by-possession study of five title contenders, graded every player and every play—and came to some surprising conclusions
Decrease font Decrease font
Enlarge font Enlarge font
November 14, 2011

The Case For The Defense

For a sport awash in stats, accurate measurements of defensive prowess are hard to come by. So SI did its own possession-by-possession study of five title contenders, graded every player and every play—and came to some surprising conclusions

View CoverRead All Articles
BERNARD JAMES, F 23.1 18.8 64.6 27.3 13.6 24.8 83.1
CHRIS SINGLETON*, F 28.0 22.6 62.6 31.2 20.2 36.8 83.5
JOHN KREFT, C 9.6 21.4 60.9 27.4 15.9 69.8 83.6
OKARO WHITE, F 17.0 17.4 59.0 37.4 22.5 51.7 84.9
DEIVIDAS DULKYS, G 23.9 14.7 60.6 36.3 23.3 36.3 85.0
DERWIN KITCHEN*, G 31.9 14.0 59.3 45.5 21.8 39.0 85.2
MICHAEL SNAER, G 29.0 13.6 53.5 39.2 19.9 39.6 87.1
LUKE LOUCKS, G 14.4 14.5 52.7 42.9 14.0 71.4 87.5
TERRANCE SHANNON, F 14.1 20.2 50.1 40.5 13.3 56.5 88.6
IAN MILLER, G 13.7 16.5 47.1 48.0 14.1 53.3 89.4

What does the information age have against defense? In 2011 you can choose any prominent college basketball player and use recruiting Websites to see his high school rankings by class and position, his scholarship offers, and his commitments to and de-commitments from colleges. From his college box scores you can track his shots, assists, rebounds and turnovers. On tempo-free stat sites you can find his per-game and -season offensive efficiency rates and the percentage of offensive possessions he uses (that is, how often the possession ends with his taking a shot). Through advanced video scouting services you can study his offensive tendencies—such as the number of times he shoots over his right or left shoulder in the post and how many points per possession he averages from each of those angles. With minimal effort, you can overwhelm yourself with information.

But there is no site that provides data on how effective that player is on defense. You can learn how efficient his team is, and try to extrapolate, using the frequency at which he grabs defensive rebounds, blocks shots or steals the ball. But on any given possession it's not clear whether he forced a nonblocked miss, harried his man into throwing a pass out-of-bounds or tipped a dribble that fell into the hands of a teammate who was credited with a cheap steal. You don't even know the percentage of defensive possessions in which he was directly involved.

A box score offers only a partial strand of defensive DNA and allows players to earn reputations as stoppers based on anecdotes rather than hard data. In his 2003 book, Basketball on Paper, statistician Dean Oliver—the godfather of the tempo-free movement—proposed formulas for individual offensive and defensive efficiency ratings. His offensive rating (ORating) is now widely used, because it can be determined from play-by-play data. But his defensive rating (DRating), which tracks all the factors leading to a stop, is rarely calculated because it requires charting film of every game and reviewing each possession multiple times to properly assess every defender's contributions and shortcomings.

Using Oliver's methods as a guide, SI conducted the most comprehensive study of individual defense ever done in college basketball. After a national title game in which the champion, UConn, held Butler to a record-low 18.8% shooting from the field, and previewing a season in which a dazzling amount of NBA-level talent is returning, breaking down the defenses of five 2011--12 championship contenders seemed a worthy exercise. (Just who is going to stop all those future pros?) SI analyzed every defensive possession from each team's final 20 games, including the postseason, to the micro level of assigning fractions of blame or credit. The subjects comprised the nation's most efficient defensive team from '10--11 (Florida State); the two preseason title favorites with the majority of their rotations returning (Ohio State and North Carolina); the defending champion Huskies; and, for contrast, one offensive powerhouse that was undone by its struggles on D (Vanderbilt).

The charts in this story include SI's individual DRatings, derived from Oliver's formula and then adjusted for competition. A defensive rating is essentially the answer to the question, If Player X were on the floor for 100 defensive possessions, engaging in his normal rate of plays, how many points would an opposing team score? (Thus, an elite defender will have a rating significantly lower than his team's.) And of those 100 possessions, if Player X is involved in 20, then 20% (DPoss%) of the formula is based on his individual stop percentage (Stop%), and the other 80% is based on the team's average stop percentage. That's why the range within a team's chart is so narrow, why differences of even a few points between players are noteworthy and why DRatings are best used to assess a single team, rather than to compare one team with another.

The stats used to compile the rating are revealing on their own. Field goal percentage against (OppFG%) isn't just a measure of the shots taken in a player's face; it's a ratio of the misses he forced against the baskets for which he was responsible. Turnover percentage (OppTO%) reveals the percentage of a player's defensive possessions in which he forces any kind of turnover. Free throw rate (FTRate) assesses the impact of a player's fouls, by ratio of his free throw attempts allowed versus field goal attempts allowed. (The lower the number, the better.) Defensive rebounding rates—the percentage of possible rebounds a player gets—are referenced when relevant.

The final step was to discuss the results with the subjects, putting the data in the context of the teams' defensive philosophies, personnel and proprietary methods of evaluating defenders. If college basketball analysis has defined eras, box scores represent 1.0, tempo-free stats are 2.0 and defensive enlightenment—when it fully arrives—will be 3.0. These are the first glimmers of light.


For 10 lonely years Seminoles coach Leonard Hamilton has been a dog lover without a dog. Until 2001 he always had a boxer, four in a row, all named Chopper. After Chopper IV passed away while Hamilton was coaching the NBA's Wizards, he could not persuade his family to welcome a fifth. He eventually found a way to fill the void.

Hamilton has long insisted that his teams play defense with a junkyard-dog mentality, which means, "If you hear some trouble around the fence, all the dogs run there to protect it," he says. "We don't want you coming in and stealing parts off of our cars." With the Seminoles vying to lead the nation in defensive efficiency for the third straight season, Hamilton expanded on that metaphor by commissioning a special logo—a snarling pit bull he named Stopper—and having it printed on T-shirts that were unveiled last month during Midnight Madness.

Continue Story
1 2 3 4 5 6 7