Points per Game can be often be misleading, but it is easy to understand: teams that play at a faster pace will allow more points. For this reason, Points per 100 Possessions is that preferred metric of most advanced basketball analysis. The Team Defense Ranks on KenPom.com (AdjD) are based on this metric.
Many teams are inconsistent – brilliant stretches of play are often interrupted by lapses of ineptitude. Because of this, I wanted to look at teams when they were at their best – and how often they were there. To do this, I created a new metric: Shutdown Stretches. A Shutdown Stretch is a period of 4+ game minutes without allowing the opponent to score. The teams with the most Shutdown Stretches can be seen below.
Of these teams, only Michigan is a surprise. The fact that they have 23 Shutdown Stretches is impressive, and being tied with Louisville and Florida is no small feat. Most impressive about that stat is that they didn’t rack all of those up against weaker competition. Michigan had two against Ohio State, two against Pittsburgh, two against Kansas State and one against Arkansas. The Wolverines have shown that they are dominant on Offense (#1 in AdjO by a significant Margin on KenPom) but this shows that they have the ability to be equally dominant on defense.
Also of note – several highly ranked teams were at the bottom of these stats.
NC State’s defensive struggles are well documented – they gave up 86 points (and lost) to Wake Forest Tuesday. Ole Miss has been solid defensively all year, but hasn’t shown the flashes of dominance that other top rated defenses have. North Carolina has been a model of inconsistency all year – it will be interesting to see if Roy Williams can put together some solid stretches and turn around the Tar Heels’ season.
Note: The rankings displayed in the charts are based on the SevenOvertimes system – a ranking system that was presented at the 2012 MIT Sloan Sports Analytics Conference. The key feature of the SevenOvertimes system is that it does not look at game outcomes or wins and losses when ranking teams – it examines the play-by-play data from each game to create a Cumulative Win Probability (CWP). The CWP from each game is a measure of how much of the game each team won. The CWP is then averaged across all of a team’s games to create a unique rating for each team.