"Isaiah is not only going to block a lot of shots, he's going to alter shots and make people think twice about their shots." -- Baylor Coach Scott Drew, on his freshman center Isaiah Austin (ESPN.com, May 23, 2012).
It is commonly accepted that great shot blockers do more than just block shots. They also alter shots that they do not block, which leads to additional misses. This bit of conventional wisdom makes sense, but it is frightfully difficult to measure. Just how big is this effect, exactly? Is it big enough to matter?
It is hard to reach a completely satisfactory answer to this question, but basketball play-by-play data can be used to get some sense of the magnitude of this effect. Using data from the hoop-math.com database, it is possible to look at the fate of unblocked shots taken against each team in Division I basketball. While not every shot attempted at the NCAA’s highest level is captured in this database, many are. Through December 31, 2012, the database contains around 200,000 shot attempts taken this season.
It turns out that teams that block a high percentage of shots at the rim also see a higher percentage of opponent misses on unblocked shots from in close. The effect is real, and appears to be big enough to worry about. For shots attempted further from the rim, the altering effect of shot blockers is small enough that it cannot be detected by the methods used in this article.
Shots attempted at the rim are the most likely shots to be blocked. Across Division I, around 10 percent of all shot attempts taken at the rim are swated. This shot blocking percentage varies significantly from team to team; Kansas has turned away 28 percent of opponent shot attempts at the rim so far this season. In order to determine how big the altering effect of a shot blocker is on unblocked shots, we need to look at how shooting percentages on unblocked shots change as a function of block percentage. The graph below plots opponent field goal percentage at the rim on unblocked shots as a function of shot block percentage at the rim for each team with 200 or more shot attempts in the hoop-math database.
There is a fair bit of scatter in the data in that figure, but there does appear to be a trend of decreasing shooting percentages on unblocked shots as shot block percentage increases; this trend is shown with the best fit line in the plot. On average, for every five percentage point increase in shot blocking percentage at the rim, there is a two percentage point decrease in the field goal percentage on unblocked shots at the rim.
This effect is large. When we combine it with the effect of shot blocking on the blocked attempts at the rim, a five percentage point increase in shot block percentage at the rim reduces opponent shooting percentage at the rim by about five percentage points. To better put this into context, a team like Duke that blocks 10 percent of opponent shots at the rim gives up an NCAA D-I average field goal percentage of 61 percent on these attempts from in close. The Texas Longhorns, who have blocked 16 percent of opponent attempts at the rim so far this season, are allowing opponents to connect on 55 percent of their attempts from in close, which is exactly what you would expect based on the graph above. Blue Devil opponents have shot 68 percent on unblocked attempts at the rim, while the Longhorn’s foes have made 65 percent of these unblocked shots. Those three percentage points matter.
While there are teams like Texas and Duke that closely follow the trend line in the plot, there is still a lot of spread in the data. Many teams do not come close to the average result predicted by the best fit line. Some teams, such as the Cincinnati Bearcats, have seen opponents shoot for a much lower percentage on unblocked attempts at the rim than what an average team with the same shot block percentage would expect. Cincinnati opponents have only made 56 percent of their unblocked layups and dunks, which is substantially lower than the 65 percent predicted from the best fit line. At the other end of the spectrum is Syracuse. The Orange have tossed 21 percent of opponent attempts at the rim, but opponents have made 72 percent of their unblocked attempts from in close.
While shot blockers on average appear to affect unblocked shots at the rim, they don’t appear to do much to shots taken away from the rim. There is essentially no meaningful correlation between shot block percentage on two point jump shots and the shooting percentage on unblocked jumpers.
But at the rim, shot blockers do affect unblocked shots enough to measurably lower shooting percentages.