Teams take advantage of playing on their home court. Players wake up in their own beds and warm-up in their own gyms. If you’re an Arizona State University (ASU) player, you have current Bellator Light Heavyweight Champion Ryan Bader appear at your game. He wears only his championship belt and aggressively brandishes a shake weight just past the baseline. He distracts your opponent during free throw attempts while two Pikachu dance next to him.
The Curtain of Distraction
It is called the curtain of distraction. ASU began using it during their 2013-2014 basketball season. This technique recently captivated the “#random” thread on our company Slack after the Bleacher Report shared a video of the curtain in action.
The question on everyone’s mind: “Does it actually work?”
We collected ASU’s opponents’ free throw percentages from the Fox Sports’ game logs for the 2016-2017 season. Extracting all games played against ASU, we created a separate category called “at ASU.” During these games, opponents shot free throws with the curtain of distraction. We conducted all tests with a significance threshold of 0.05. A Shapiro-Wilk test (p-value = 0.232) allowed us to operate under the assumption that the data were normally distributed.
We were, therefore, able to conduct a one-way ANOVA test, hypothesizing that at least one distribution of free throw percentages significantly differed from the other distributions. This test returned a significant result (p-value = 0.022). When comparing several groups in this manner, a difference likely occurs by chance. With a one-way ANOVA, we cannot determine which of the groups were different, only that one might be different. Further testing was necessary.
To determine which, if any, groups were different, we carried out multiple comparisons via pairwise T-tests using Bonferroni adjusted p-values. The Bonferroni adjustment accounts for the number of tests being conducted. This controls for the likelihood of chance outcomes. Pairwise testing compares all possible pairings of categories to determine if any single pairing is different.
None of the pairwise tests returned significant results. We concluded that teams’ free throw percentages do not significantly differ from the “at ASU” group. Further, none of the teams’ free throw percentages significantly differ from each other’s.
When an opponent misses a free throw at ASU, it certainly seems like the result of the curtain of distraction. This assumption, however, perfectly exemplifies a confirmation bias — an attribution of new information to what we believe or hope to be true. In this case, we believe that Michael Phelps standing under the basket and wearing a collection of gold medals will shake a player’s concentration. But, in fact, players miss free throws all of the time.
From these tests, we cannot conclude that ASU’s curtain of distraction causes teams to shoot free throws at a significantly worse percentage than would be expected during any other away game.
So, if teams are not distracted in a way that significantly alters free throw shooting, should the curtain of distraction cease to exist? Absolutely not. At the university I attended, per tradition, we slap the backs of our hands together and bark like seals during the opponent’s free throws.
The curtain of distraction might not have statistical significance, but it is objectively better.