Basketball Day

In a post last week, I shared one of my favorite in-class exercises on coin-flipping and random walks. A related and equally fun day in my commodity price analysis class is what I call “basketball day.” On this day, I wear my special edition Uncle Drew Kyrie high tops to class and a University of Illinois basketball jersey. I also show an old picture of me playing basketball in high school (BW of course). I am not sure that they believe the kid in the picture is really me. I do not explain the reason for the unusual attire until later in the lecture. I also show a more recent picture of me shooting baskets in my driveway to show that I still love the game. Also note that I am wearing the Kyries in the picture on the right.

I begin the formal part of the lecture by reviewing the random walk model and its implications for the technical analysis of commodity prices. This usually prompts an interesting discussion about whether technical methods, like charting, work in real-world markets. As I said before, I want students to make up their own minds on this crucial question for commodity market analysis. I also think that students deserve to know what I think on the issue.

I explain to students that there are two options with regard to technical analysis: #1) commodity prices do not follow a random walk and technical analysis has predictive value; or #2) commodity prices actually do follow a random walk but people have mistaken perceptions about price movements. I lean towards option #2 because there is considerable evidence that we humans have a tendency to see patterns in randomness where there is none. This is where the basketball part of the lecture comes in. I introduce one of my all-time favorite research papers to the students, “The Hot Hand in Basketball: On the Misperception of Random Sequences,” by Gilovich, Vallone, and Tversky. This famous 1985 paper investigates the validity of the belief that “hot hands” and “streak shooting” are common in basketball. I share the fascinating evidence contrary to this belief in the paper. Since I was a huge Dr. J fan back in the day, I particularly like this table with shooting streaks from the Philadelphia 76ers from their 1980-81 season:

On average, the unconditional probabilities are shockingly close to the conditional probabilities. Every year, I see the sports fans in the class shaking their heads in disbelief at this evidence. Another teachable moment has arrived.

I am not sure I ever convince the basketball players and fans in the class that hot hands are not real, but we sure have an interesting conversation. I end the lecture by sharing with the students my personal struggle regarding the validity of technical analysis. I have a hard time squaring all the really smart people I know that use technical analysis with a pure random walk in the markets. Not very satisfying but at least I am honest!

This year’s class agreed to take a group photo to commemorate “basketball day.” Photo credit goes to my amazing teaching assistant Maria Gerveni.

Browse other posts by category: