This is part of a series of posts examining whether offense or defense has more control of various aspects of a typical college basketball game. The introduction is here. A description of the methodology is here.
Three-point percentage is next on the pecking order of offensive control, checking in at an average of 83% over the past ten seasons. There’s been enough chit-chat around here about long-range shooting (to wit) to understand that the offense has majority control over the chances of a three-point attempt being successful. And while three-point defense exists on some level, it’s difficult to distinguish how much is sneaky close-out ability and how much is opposing offenses choosing to take low-percentage 3’s over low-percentage 2’s.
But given that defenses have more influence on opposing two-point percentage, one can imagine that there’s at least a component here that involves offenses electing to take more 3’s in an environment where getting easy 2’s is not possible. Against a good two-point defense, invariably shots will be challenged.
But you can move back far enough to take an unchallenged three against most defenses. That doesn’t mean those will be high-percentage shots, even accounting for the higher payoff. But one can understand why good three-point defenses and good two-point defenses are often found together.
One thing that has been ignored in this analysis to date is the influence of luck or random variance, call it what you will. On a game level, there’s a lot of random variance in three-point shooting. So this result is not to say that a coach should pile up good shooters expecting to make a bulletproof offense. It’s simply that a good three-point offense will beat a good three-point defense over the long term. The offense controls most of what can be controlled, but randomness is a huge specter that looms over three-point shooting on a game level. Controlling for attempts, free throw shooting is most predictable, followed by two-point shooting, which is followed by three-point shooting.
Playing great three-point defense is admirable and for teams like Kentucky and Arizona and Baylor, who showed the consistent ability to hold opponents below their season average last season, that trait contributed to their defensive superiority. (In Baylor’s case it’s relative. The Bears finished 38th in adjusted defensive efficiency, but that was the best figure of the Scott Drew era. And having the eighth-lowest opponent’s three-point percentage helped.)
But there’s always the possibility that Sam Dekker or R.J. Hunter will make shots at the end of games against even the best perimeter defenses. The three-point shot is worth more than any other shot and with the exception of free-throw accuracy, it’s the box-score event least influenced by the defense. So it figures that the trifecta is a good starting point for explaining the unpredictability of college basketball. The reduction in the shot clock may reduce the parity in the game in the extreme margins but as long as the three-point shot exists at such a makeable range, there will be plenty of unexpected outcomes.
The method used here spits out a large variation of percentages from season to season, with values ranging from 76 to 97%. Which gives you an clue that this method is not spitting out answers with absolute precision. Last season’s value of 84% is in line with the long-term average, though. Teams at home can expect a bump of 0.7% in their shooting compared to a neutral-site game.
Year %Offense HCA 2015 84 0.7% 2014 81 0.5 2013 76 0.8 2012 78 1.0 2011 90 0.7 2010 83 0.7 2009 97 0.6 2008 81 0.9 2007 81 0.9 2006 82 0.5 AVG 83 0.7%
And here’s a summary of the findings to date.
Offensive Spectrum – Ordered by pct of offensive “control”
FT% 98% (HCA=0.5%) APL 86% (HCA=-0.1s) 3P% 83% (HCA=0.7%) ??? 73% ??? 72% ??? 71% ??? 64% ??? 59% ??? 50% ??? 49% ??? 36% ??? 30% ??? 15%