Again, with these posts I'm mostly interested in trying to discern any statistical trends in Clemson's play, and see if they are trending in the right or wrong way.
In the ACC season midseason update, I showed this chart plotting the division I rank of Clemson's adjusted offense and defense efficiencies as calculated at KenPom.com:
Aaah, the good old days when things were looking really good. Defense efficiency was more or less holding steady in the top 30, and you can't be unhappy with the offensive efficiency trend. Unfortunately, here's what the same plot looks like now at (post-)season's end:
I think there are two ways of interpreting these charts: 1) the first chart was a result of an extended statistical fluctuation that overvalued Clemson's defense, and the second chart is actually a closer approximation of Clemson's true talent level for the season. 2) Something happened midway through the ACC season, whether it be team chemistry-related or the league suddenly discovering how to break Clemson's defensive scheme, causing Clemson to play dramatically worse on defense the remainder of the season.
I tend to favor the latter interpretation for several reasons: 1) twenty-two games is probably enough data to give you a pretty good measure of the team's talent level, 2) I think there are obvious adjustments the rest of league made that we've pointed out on the blog before, and 3) subjectively I think we can all agree that Clemson just played like a different team in the second half of the ACC season, putting forth a much "sloppier" effort over the last 8 games and the postseason.
At any rate, to dig a little deeper into the defensive decline, I plotted some of the component defensive statistics over the years since Purnell has taken the helm. First, how the defense fared against the four factors of opposing offenses. The four factors (effective FG%, turnover%, offensive rebounding%, and FT attempts) were determined by Dean Oliver as the most important factors in predicting basketball success:
Right away we can can see where Oliver Purnell's defenses have made their mark: forcing turnovers. Clemson has consistently been amongst the best in division I at forcing turnovers since Purnell took over. The other difference between 2008 and 2009 that jumps out at me is the slight rise in effective FG% against rank. At first glance it might not seem like much, but what this chart fails to capture is the relative importance of each factor. As it turns out, FG% is by far the most important factor, generally twice as important in determining success as the turnover% and also about twice as important as offensive rebounding% and FT attempts combined. In other words, a rise in FG% against is twice or four times as bad as comparable increases in the other factors.
Looking deeper still, we can break the first two factors down a little more by looking at 3-point and 2-point FG% against, as well as steal% and block% against:
Here we again see some Oliver Purnell staples: steal% ranking in the top ten in division I and block% typically ranking in the top twenty or so. We can also possibly see a little more of the problem, the sudden jump in 3-point% against increased rapidly. This may have been negated somewhat by the drop in 2-point% against, but again, the 3-point% would probably have to be weighted as more important. Now, its true that the 2007-08 team was superhuman at guarding against the three, so we could have expected some regression coming into this year based on previous performances. But this is clearly the worst 3-point% against since Purnell took the helm.
One last chart to look at, this one breaks down how teams scored their points against Clemson into the percentage of points from the 3-point, the 2-point, and the FT shot:
Here we see the total percentage of points coming from 3-point shots taking a larger chunk of the points we gave up over the course of the year. This doesn't mean that we gave up more points, it just means that teams appear to have been looking more to the three point shot to put points on the board against our defense.
While none of the above three charts are particularly convincing on their own, I think they all add up to a pretty strong argument that Clemson's defensive struggles were directly related to a much poorer 3-point shooting defense. While this is worrying, it takes more than one point to indicate a trend. Definitely something to watch for next year, however, because over the course of an entire season that kind of 3-point defense could land us back in the NIT.
No comments:
Post a Comment