Will the 30-second shot clock benefit defenses more than offenses?
On June 8, the NCAA officially announced the arrival of the 30-second shot clock for the 2015-16 season. I had my hopes up for a triumphant announcement video, maybe something that opened with footage from the Nov. 29, 2014, slowgrind where Virginia beat Rutgers, 45-26, only to have a giant, CGI'd version of men's basketball rules committee chair Rick Byrd emerge, stop play, tear the center-hanging scoreboard out of the arena ceiling, smash it onto the court, set it afire, then look directly into the camera as it zooms in on his face and he proclaims: ENOUGH IS ENOUGH!
The NCAA thought a Tweet would be more appropriate.
What matters, I guess, is whether the rule change yields results. The objective of shaving five seconds off the shot clock is to increase pace and scoring in a sport that has been trending downward in both of those categories. The early returns, based on Ken Pomeroy's study of the 2015 NIT, CBI and CIT tournaments, which used the 30-second shot clock on an experimental basis, showed some positive effect: pace was roughly two possessions per team higher than expected under normal rules; and scoring was 2.4 points per team higher than expected under normal rules.
I'd love these scoring gains continue into 2015-16 and beyond, and the sport enters an era where walking the ball up goes out of style, offenses are initiated more quickly, and teams hunt for more transition shot attempts.
But the pessimist in me wonders if slowball is really on its deathbed, and if a new era of offense is truly upon us. I worry that an unintended consequence of the 30-second shot clock is that grind-it-out defenses will become even more popular—and successful—than they already are. The shot clock will be expiring sooner, and there's a distinct advantage in sending as many possessions as possible into panicked, end-of-shot-clock territory.
According to my study of Synergy Sports Technology's logs, the shot clock ran down to under four seconds on just 4.0% of Division I possessions in 2014-15. In the NBA this season, the 24-second clock ran down to under four seconds on 11.3% of possessions. As the college shot clock moves closer the NBA's, I would expect the college volume of end-of-shot-clock possessions to shift closer to the NBA's as well.
There's a reason this is a big deal: Getting jammed up against the shot-clock deadline is a highly inefficient situation for an offense, and thus an attractive goal for a defense.
The data supporting this is strong, starting with my D-I-wide calculations of Synergy splits for PPP allowed on possessions that end in transition (those IDed by Synergy's human loggers as fastbreaks and secondary fastbreaks), versus halfcourt with more than four seconds on the shot clock, versus halfcourt with less than four seconds on the shot clock (this includes expirations):
College offenses are horrible—scoring just 0.703 PPP!—when up against the shot-clock deadline. They're horrible in comparison to other college scoring situations, and bad even in comparison to the NBA's under-four-second efficiency this season, which was 0.767 PPP. (NBA players, as you'd expect, are better at late-clock, one-on-one shot creation.)
Data from hoop-math.com further reinforces this, by allowing us to gauge the quality of D-I-wide field-goal attempts at different stages of possessions:
Under the 35-second shot clock in 2014-15, according to hoop-math, 10.7% of FGAs were taken 30 seconds or more into a possession. The effective field goal percentage on those shots was an ugly 42.1%, well below the eFG% on all other shots, which was 49.5%. Even when against-the-deadline offenses didn't turn the ball over or let the clock expire, they were stuck taking bad shots.
These late-clock jam-ups are now a more attainable goal for defenses, as it should be easier to maintain intensity and avoid breakdowns for 30 seconds than it was for 35. For any coach who's aware of the expected (low) efficiency in those end-clock situations, a defensive scheme that eats up time is going to be attractive.
That brings us to the (obvious) next question: What defense scheme is best suited for grinding out the shot clock? Conventional wisdom seems to be that zone defenses take longer to score against:
Common thought among college coaches: teams that rely on zone defense will thrive with 30-sec shot clock. Harder to probe zone w less time.— Jon Rothstein (@JonRothstein) June 1, 2015
I suspect that this is because Syracuse is the program most associated with zoning, and Syracuse's 2-3 zone requires opponents to work at length to find scoring opportunities. But across all of D-I, does playing more zone necessarily result in longer defensive possessions?
To find that answer, I combined Synergy's data on percentage of possessions teams spent playing zone with kenpom.com's team-by-team data on average defensive possession length (adjusted, based on Synergy's transition-allowed% stat, to attempt to remove the impact of fastbreaks). The chart below plots individual defenses' Average Non-Transition Possession Length vs. Zone-Played% over the past three seasons:
While the trend line shows a slight increase in possession length for zone-heavy teams, the data is all over the place, and the overall correlation between possession length and Zone% is weak (R^2=0.051). Playing zone does not guarantee you'll eat up time—all zones are not Syracuse's—and many of the longest-possession creators played majority man-to-man.
Examining this on an individual-team basis led to more interesting conclusions. These were the top 10 teams in average halfcourt defensive possession length over the past three seasons:
Syracuse and its all-zone clone, Eastern Michigan, coached by former Orange assistant Rob Murphy, are Nos. 2 and 5, respectively. But the rest of the top 10 was not majority zone. Florida was No. 1 with only minimal zone usage, and San Diego State, which is almost exclusively man-to-man and forced the longest possessions of any defense in 2014-15, was No. 3.
What Florida and San Diego State had in common is not just excellent coaches that teach players to prevent transition, contain penetration and maintain defensive focus in man-to-man. While those things contributed to long possessions, another shared attribute led to the longest defensive possessions in D-I: full-court pressure. Here's the percentage of time, via Synergy, that the top 10 possession-length teams spent pressing:
When West Virginia famously morphed into "Press Virginia" this past season, it was by using full-court pressure that created chaos, high turnover volume, constant transition for both teams, and short possessions. A scheme like San Diego State's, with early pressure that falls back into containment man-to-man—I call it press-n-grind—has a different effect. It creates the occasional turnover but doesn't gamble enough to give up transition buckets.
The Aztecs' main goals seem to be to use an early trap to get the ball out of the opposing point guard's hands, force fewer ballhandlers to get involved in bringing the ball across halfcourt ... and kill so many seconds that, by the time the opponent gets situated in its halfcourt offense and starts running its first action, there are only 25 seconds, max, left on the shot clock. And I'm talking about the old shot clock.
Watch this possession from San Diego State's Nov. 18, 2014, win over Utah, and you'll grasp the impact of the time-killing press, which gets applied after most made shots:
In 2014-15, the Aztecs forced a nation-high 7.9% of their defensive possessions into the final four seconds of the shot clock, according to Synergy. It was no coincidence that they ranked fourth nationally in adjusted defensive efficiency.
Florida was awful this past season but, in 2013-14, ranked second nationally in adjusted defensive efficiency, bolstered by its own time-sucking press. This example, taken from the Gators' 2014 NCAA tournament win over UCLA, is more helter-skelter than San Diego State's press, but has a similar result. The Bruins' point guard, Bryce Alford, is prevented from easily bringing up the ball, the offense isn't initiated until 25 seconds are left on the shot clock, and they run two dead-end actions before getting stuck in a low-percentage isolation situation with the clock running out:
Under the new shot clock, presses like San Diego State's (or VCU's, or Louisville's, or Florida's, if it continues post-Billy Donovan) could prevent opponents from initiating offense until there's only 20 seconds left to shoot. That's enough time to run a couple of actions; if they don't yield good scoring looks, the dreaded, low-efficiency, up-against-the-deadline situation arises. And how many college teams have a killer isolation player who can bail them out of trouble? The 30-second shot-clock could trigger the spike in scoring that college hoops needs, but it could just as likely bring rise to a formidable army of press-n-grinders and Syracuse copycats, eating up time and forcing panicked heaves before the buzzer. It's far too early to concede this war to the offenses.