The effect of good rim protection

January 10, 2017 by J.H. Yeh

We previously discussed that the most efficient shots are the ones attempted underneath the basket. That is followed by corner 3s and on top of the key 3s. Therefore, it would make sense for teams to try and impede the opponents from taking shots right below the rim as that should be every team’s number one defensive assignment. With that in mind, blocks have become a metric that people use to assess a team’s defense near the rim.

Firstly, drawing from what we discovered in my earlier articles, teams that don’t have a good shooting percentage also don’t have a good 3-point percentage. Teams that struggle with shooting from both inside and outside of the 3-point line usually are not good at floor spacing. Moreover, by preventing the opponents from taking the most efficient shot at the rim, the teams’ effective field goal percentage should see a drop because all the range shots are lowering the overall field goal percentage. To sum up, we can conclude that opponents’ field goal percentage is correlated with the effective field goal percentage. A team can effectively reduce the opponents’ overall efficiency if it can successfully take away their most efficient shots near the rim. Remember that shooting 3s (despite the high returns) alone cannot win games. Even the most prolific 3-point shooting team attempts 40% of their shots from beyond the 3-point arc.




At an extremely high R-Squared at 0.899 (almost 0.90!) to go along with a slope higher than 1 and no obvious outliers, the strong positive correlation between opponents’ FG% and the opponents’ eFG% definitely suggests that very good rim protection can lead to overall inefficient shooting for the opponents. Evidently, this shows just how critical defense near the basket is. As expected, with teams featuring elite defenders in Rudy Gobert and former defender of the year Marc Gasol, both the Utah Jazz and the Memphis Grizzlies are very successful in holding down the opponents due to their size and sturdy interior defense.

Let’s narrow this category to opponents’ field goal percentage at the restricted area (RA), and see how it affects eFG%.


With a good R-Squared of  0.474, we can also conclude that good defense near the rim can lead to low opponent eFG%. The drop in R-Squared is due to narrowing down the FG% to just the restricted area and leaving out the defense in other regions such as the paint. In here,  defense in other regions means good lock down defense and effective defensive rotations that obstructs drives into the basket. Leaving all these other factors aside, this explains why teams vary along the regression line because it depicts different defensive strength in those teams. For example, teams positioned on the bottom such as the Los Angeles Clippers and the Houston Rockets are good at perimeter defense that significantly brings down the opponents’ eFG%. On the other hand, the Portland Trailblazers is only good in interior defense alone and very weak at stopping drives and challenging 3-point shots. With offense-only players in Damien Lillard and C.J. McCollum at the guard positions, the Trail Blazers give up 37.6% on the opponents’ 3s which is the fourth worst in the league. This is a major concern as teams today are all shooting 3s and driving to the basket, which explains why the Trail Blazers are struggling and underachieving this season.

We’ve gone over the relationship between defense at the restricted area and the opponents’ eFG%, now let’s move on to blocks as we take a close look at the correlation between blocks per game and opponents’ FG%. We have developed a habit of attributing good defense to individual players because blocks alone can tell us that Hassan Whiteside is a good defensive player and Zaza Pachulia is not. However, when looking at a team’s defense, it is better to look at the collective defensive performance of the entire team. And this also extends to defense near the rim as well!



The result is a bit messy with teams deviating from the line. Despite that, we can still draw a general trend that more blocks per game can lead to lower opponents’ FG% thanks to the acceptable R-Squared at 0.432. The downward sloping regression line explains the negative correlation between shots blocked per game and the opponents’ shooting efficiency reflected on their overall FG%. With notable outliers lurking at the bottom such as the Utah Jazz and the Toronto Raptors, this tells us that both teams are very good at reducing their opponents’ FG% without blocking too many shots because they have good defensive rotations and can timely anticipate and get into the optimal defensive position to challenge each shot. This phenomenon is one of the many evidences that illustrates blocks are overrated and are not a good metric to evaluate a team’s defense. On the other hand, the Minnesota Timberwolves, despite being coached by defensive-minded coach in Tom Thibodeau and blocking close to the league average at 4.6 blocks per game (league average is 4.9), the struggling franchise has a poor opponent FG% at 0.472. This is a direct reflection that shows the young and inexperienced lineup in the Timberwolves. The young players are athletic enough to get up and block the shots but are not experienced enough to skillfully and effectively prevent shots by good defensive rotation and playing smart defense by their feet. And that is okay, because young players are inefficient and it is only through time and experience for them to get better.

Similarly, now we compare the relationship between blocks per game and opponent’s eFG% as we take the opponents’ 3-point shooting into consideration.


Now the results are better with a R-Squared near 0.500. Again, we have the Clippers as one of the outliers at the bottom, further validating that they are skilled in playing good solid defense at the perimeter. The Jazz and the Grizzlies also made an appearance in that same region in large part because of the presence of Tony Allen and George Hill. Similarly, above the trend line features misfit or rebuilding teams, or a little bit of both. Teams such as the Lakers and the Philadelphia 76ers are clearly young and ineffective while the Dallas Mavericks are an overall misfit team with players who don’t complement each other too well on the defensive end.

Now let’s further breakdown the regions. We will start with the relationship between blocks and 2-point field goal attempts, made, and percentage.


The result is pretty poor as teams deviate all over the place. What’s more befuddling is that the pattern suggests more blocks are leading to more shots taken within the 3-point line. Overall, the result is inconclusive.

Next up is the relationship between blocks per game and 2-point field goal made.


There is virtually no relations here as the slope is near horizontal. The Golden State Warriors lead the league in blocks per game at 6.2 while the Lakers are blocking the least at 3.5 blocks per game. In spite of the huge difference in blocks, this doesn’t discourage opponents from making shots within the 2-point line as opponents make an average of 30.7 2-point field goals against the Warriors and 29.5 when playing against the Lakers. However, it is noteworthy that the Memphis Grizzlies are the sole outlier as opponents only make an average of 24.3 shots inside the 3-point line. This may be attributed to the combination of Gasol’s excellent interior defense and the big lineup that clog around the basket which leaves little room to drive and post up.

Now moving on to the  2-point field goal percentage.


This is much better. Even with a mediocre R-Squared at 0.305, it’s a far cry from R-Squared of  0.0007! Or just nothing at all.

In this regression chart, it shows a nice relationship between blocks and 2-point field goal percentage as more blocks can lead to lower opponents’ made baskets in the 2-point territory. Again, as expected, young teams like the Timberwolves and the Lakers’ block numbers are over-estimate their actual defensive ability, which further confirms what we went over earlier that blocks alone are a misleading metric. And of course, the Jazz are the outlier because blocks alone under-estimate their actual defensive ability.

Now we will move on from 2-point field goals to field goals inside the paint area. Again, we will go through each one of them with the same order: attempts, made, and percentage.


This is actually worse than the relationship between blocks and 2-point field goals attempted, with a negligible R-Squared of 0.0001!


And the relationship between blocks and field goals made in the paint is also equally bad. At least we have an extremely slight improvement of 0.0012 in R-Squared.  Nevertheless, a good explanation why the relationships between blocks and field goals attempted and made inside the paint is so poor compared to inside the 3-point line is because substantially more shots are attempted near the basket and fewer shots are attempted inside the paint. With that in mind, we can also safely assume that the relationships between blocks and mid-range shots (between paint and 3-pointer) will also see a very similar pattern, if not worse.


Also featuring an almost zero in R-Squared, this result surprises me because I expect a better correlation between the two. However, the chart suggests otherwise. This may very well due to the fact that shots attempted in the paint are too far for the blocks to have a palpable effect on their percentage and floor spacing plays a big role in that there are more highly contested shots occur near the rim in the restricted area than in the paint.

And speaking of the restricted area, this leads to our final region to look into. Again, we will go through this with the same order in shots attempted, made, and the percentage.


We have a very low R-Squared. Interestingly, more blocks per game does not lead to fewer attempts near the basket. Instead, it leads to more shots taken at the rim.


And once again, we have a near horizontal trend line with teams vary wildly on the plot. Like the previous chart, the line is sloping upwards which suggests more blocks lead to more made shots near the rim. Now let’s observe our final chart in this post.


The regression line reflects the overall defensive prowess of each team. With a good R-Squared in 0.433, the scatter plot reflects the defensive ability of the teams as we see the Jazz are once again the outlier as their number of blocks does not reflect their true defensive ability. Additionally, a good explanation why blocks do not discourage teams from attempting shots is because teams want to score from the most efficient regions by powering their way and taking as many shots close to the basket as possible. Even with robust defense near the rim, teams are still not afraid to challenge the defense, which explains why shots attempted and made in the paint and the restricted area see near horizontal slopes that trend upwards.

Moreover, the 0.305  R-Squared featured on the correlation between blocks per game and 2-point field goal % is an obvious result that shows significantly more shots are taken near the rim as the effect of opponent shooting percentage near the rim is overshadowing percentage at the paint and the mid-range.

Conclusively, the charts affirm the general understanding that blocks can lead to lower shooting percentage at the rim, and this is more apparent at the restricted area where teams strive to take the most efficient shots there. Furthermore, despite the negative correlation between shots blocked and the shooting percentage, little to no relationship exists between blocks and field goal attempts and made as good defense can actually lead to the opponents wanting to take more shots to compensate for their shortcoming on offense. As manifested on the charts, blocks are not an accurate metric to measure a team’s defense. This is more obvious when evaluating a young team’s defensive performance as blocks can over-estimate their defensive capability since young players are athletic but ineffective. And the similar can be applied to some playoff contending teams as their blocks under-estimate their overall defense.

— J.H. Yeh

(all graphs are created by me, and all sources of stats are courtesy of and as of January 9, 2017)

%d bloggers like this: