Projecting Defense: An Overview

Posted by on October 28, 2014

If you read the site last season, you might remember my look at North Dakota State and experience. To summarize, the Bison were one of the most experienced teams in the country last season and overwhelming favorites in the Summit League because of a strong defense in 2012-13. NDSU did in fact win the Summit by two full games in 2013-14, but it was a very strong offense that led them to the NCAA tournament. In fact, their defensive efficiency took a significant step backwards.
I started to discover last year that North Dakota State’s step backwards defensively was actually pretty normal. 13 of the 20 teams with the most roster continuity from the previous season had declined in defensive efficiency rank at the time of the post. Leading up to this season, I wanted to expand upon last year’s post to learn more about defensive expectations for experienced teams.
To take a look at how good defenses with most of the roster back from the previous season perform I had to define both of these aspects. There’s obviously a trade-off between precision and sample size with these definitions. I used returning MPG (for my methodology see the NDSU post) and AdjD. I then selected teams with both 30 minutes per lineup spot returning from the previous season and also a top 100 defense in the previous season. With data going back four seasons, this criteria returned 72 teams. (It should be noted stricter definitions of experience did return similar results albeit in a tiny sample).

On average, the 72 teams with top 100 defenses from the previous year and high roster continuity saw their defensive efficiency rank decline 25 spots. Just like NDSU, these teams did significantly improve on the offensive side of the ball.
I think these results are pretty surprising. They certainly go against what we know about the NBA. Rookies and young players in the NBA tend to consistently struggle defensively, usually requiring much needed experience to feel comfortable with complex NBA schemes. However, a few ideas come to mind on why experienced NCAA teams fail to make the defensive jump:
1. Regression. AdjD ranks for a given season are from a 30-something game sample and are subjected to variability. Because I am looking at good defensive teams it’s probably more likely that the variability or luck was favorable in year one and these teams are due for some regression to the mean in year two. One way to test this is to look at experienced teams with good offenses. I found 68 experienced teams with top 100 offenses the year before. These 68 teams saw their AdjO rank change from 53 in year one to 52 in year two. So while regression does probably play a role in both cases, there appears to definitely be something more to the severe decline on the defensive end.
2. Motivation. Off-season skill development largely focuses on offensive skills like ball handling and shooting. Weight training should certainly be expected to help a player’s defensive game, but it is probably safe to say defense takes a backseat to offense when it comes to individual training in the off-season. In North Dakota State’s case they were in some ways lacking incentive to get better on defense. I kind of expected them to be able to flip the switch on the defensive end later in the season during the conference and NCAA tournaments, and they did in fact have a very nice streak of games in February and March.
3. Coaching. While I haven’t looked into this rigorously (yet), I think coaching is a huge factor in defensive projections. For example, it was shocking to see such a talented Bill Self team struggles at times on defense last season. Self has proven he can produce elite defensive teams almost every season. Going back to the NDSU example one last time, the 2012-13 team was a pretty big outlier relative to other Saul Phillips defenses. It would be interesting to take a look at the experienced teams that did improve on defense in year two. I would not be surprised if the most common trait of those teams is a great defensive coach rather than any particular type of player.
Next up, I’ll breakdown the Wisconsin defense and what we should expect from the experienced Badgers this season.


  1.  Projecting Defense: Wisconsin | The Hoop Vision
  2.  Villanova’s Aggressive Defense | The Hoop Vision

You must be logged in to post a comment.