March 23, 2012 - by Jeff Haley
Kentucky has one of the best defenses in all of college basketball. They are currently ranked #10 overall on defense in the Pomeroy rankings and are the number one ranked team in opponent’s effective field goal percentage, at 41.7%. Anchoring their defense is menacing shot blocker Anthony Davis, who blocks an estimated 14% of opponent two point attempts while on the court. Kentucky is very difficult to score on.
Much of Kentucky’s defensive value comes from blocking shots. As a team, Kentucky blocks about 21% of their opponent’s two point attempts. If we were to assume that these shots went unblocked and were made at the same rate as opponent’s unblocked two point shots, Kentucky’s opponents would have an effective field goal percentage of 49.1%.
Based on this idea, I started out with the hypothesis that the best way to attack the Kentucky defense would be to minimize the opportunity for Kentucky to block shots. My first thought was that the best way to do this would be to attack Kentucky in transition. While watching Kentucky play North Carolina earlier in the season, I noticed that the Tar Heels had a very difficult time scoring in half-court sets, but seemed to get many good opportunities against Kentucky in transition.
Play-by-play data provide a way for us to look at how teams have fared against Kentucky in transition. Using a play-by-play file that contains every game of the Kentucky season (other than their initial contest vs. Marist), I determined how long it takes after an opponent gets the ball before they take their first shot of the possession. I logged the type and result of each shot. I have not included possessions where Kentucky fouls their opponent before they can shoot.
31% of the initial shots in a possession against Kentucky occurred in the first 10 seconds of the possession, and the effective field goal percentage on these shots was 52%. If teams don’t shoot within the first 10 seconds of a possession against Kentucky, things become much more difficult. Initial shots that occur between 11-25 seconds after the start of the possession make up 55% of the initial shots taken against Kentucky, and these shots have an effective field goal percentage of 35%. Initial shots that occur after 25 seconds compose the remaining 14% of the shots in this study, and the effective field goal percentage on these shots is 39%.
As always when we look at numbers like this, we need to give them context.
Using my entire play-by-play database, which includes a majority of the games played in division one college basketball, the effective field goal percentage on initial shots taken in the first 10 seconds of a possession is 54%. Kentucky’s field goal percentage defense in the first 10 seconds of a possession is only a little bit better than average for division one college basketball.
On average, Initial shots taken between 11 and 25 seconds have an effective field goal percentage of 48%, and initial shots taken after 25 seconds have an effective field goal percentage of 43%. Kentucky’s defense becomes significantly better than average after the first 10 seconds of a possession.
My initial hypothesis was that Kentucky would have a lower shot block percentage in the first 10 seconds of a possession. Kentucky blocks 21% of their opponent’s initial two point shots in the first 10 seconds of a possession, 22% of their opponent’s initial two point shots in the second 15 seconds of a possession, and 26% of their opponent’s initial two point shots taken after the first 25 seconds of a possession.
While the shot blocking rates seem to increase a bit as the time to the first shot increases, the difference is not large enough to account for the effective field goal percentage differences. Kentucky’s shot block percentage in the first 10 seconds of a possession is still very high.
The table below shows what happens on the first shot of a possession against the Kentucky defense.
Several things are different for shots taken in the first 10 seconds of a possession. In the first 10 seconds of a possession, Kentucky’s opponents are shooting a higher percentage of their total shots from three point range and are making these three point shots at a higher percentage. Teams are also shooting significantly better on initial shots from two point range in the first 10 seconds of a possession against Kentucky than they shoot during the rest of the possession.
This was surprising, given the small differences in shot block percentage. But the answer to why this is lies in the types of shots that are attempted. Against Kentucky teams shoot more than 60% of their initial two point shot attempts at the rim in the first 10 seconds of the possession. After that, teams attempt less than 40% of their initial two point shot attempts at the rim.
It is very hard to get to the rim against Kentucky once they are set up in their half-court defense, but in transition you can get there. Shots at the rim generally go in at almost twice the rate of two point jump shots.
Shots in the first 10 seconds of a possession include things like uncontested layups that come off of steals. We would expect teams to shoot better on uncontested layups against Kentucky than on the rest of their shots. But this information is not of very much practical value. Nearly every team tries to score quickly off of steals, and does so pretty effectively. In division one basketball, after a steal teams take 74% of their initial shots within the first 10 seconds of a possession, and the effective field goal percentage of these shots is 62%. Against Kentucky, 71% of the initial shots off of steals came within 10 seconds of the steal, and the effective field goal percentage on these shots was 61%.
I think it is far more interesting if we remove possessions that start from steals, and look instead at how teams fared against Kentucky after taking possession on a missed shot by Kentucky, or after inbounding the ball on a made basket. Why is this? These are the two situations where some teams will choose to push the tempo, and others will not.
The table below shows that even once we remove the effect of shots that come off of steals, initial shots that come in the first 10 seconds of a possession against Kentucky have a higher effective field goal percentage than initial shots that occur after the first 10 seconds of a possession.
In the first 10 seconds, opponents are much more likely to get and make three point shots, and are much more likely to get to the rim. Note that the result on initial shots that occur 25 seconds after a Kentucky miss suffer from a small sample size, as only 42 of these shots were in the database. There is no other evidence anywhere in the data that Kentucky’s three point defense lets up as the shot clock winds down. This high three point percentage came on a total of 16 shots.
Again, it is important to compare these results with the averages for division one basketball. After an opponent a miss, the average effective field goal percentage on initial shots that come within the first 10 seconds of a possession is 53%. After a made basket by an opponent, the average effective field goal percentage on initial shots that come within the first 10 seconds of a possession is 49%. After 10 seconds the average effective field goal percentage for initial shots for possessions starting with either a miss or a made basket by the opponent is 47%.
Comparing to those NCAA-wide rates, Kentucky is about average defensively in the first 10 seconds of a possession after either a missed basket or a made basket, and fantastic after that.
I don’t know if this means that a team should attempt to engage Kentucky in an up tempo game. The Kentucky offense is pretty good in the first 10 seconds of a possession, with an effective field goal percentage on initial shots of 60%.
But I do think that getting out in transition is a way to score on Kentucky, and could be a good strategy for a team that likes to run and can get back on defense.
Printed from TeamRankings.com - © 2005-2018 Team Rankings, LLC. All Rights Reserved.