Before the ball tipped off on the 2026-27 NBA regular season, we published our full 30-team win projections and conference standings generated by our proprietary quantitative models. Unlike traditional power rankings driven by offseason headlines and superstar name recognition, our system evaluates how a team’s entire 10-man rotation (from starting backcourt efficiency and wing spacing […]
Usually, we project total team wins based on the first few weeks of the given NBA season. This year, we wanted to try something different by projecting the number of wins each team will have just based on what we know about them before the beginning of the season. Every NBA offseason brings a wave […]
Updated: March 17, 2025 Based on our ML algorithm as previously described here: https://relativitydigest.com/2024/12/28/nba-2024-2025-win-predictions/ Our updated win projections for the season are as follows (all predictions +/- 5 Wins) Team Predicted Total Wins (+/- 5) CLE 58 OKC 57 BOS 53 LAL 52 MIL 51 MIN 50 NYK 48 IND 48 MEM 48 LAC 47 […]
Below, we demonstrate why trying to identify a single player as the “Greatest of all time” is a fallacy.
Using a lot of data wrangling and NLP, I created a movie and tv show recommender: Basically, the user selects from the drop-down list their movie/tv show of interest, and the algorithm will recommend a combination of movies and tv shows that the user is most likely to be interested in. I tried to include […]
I wanted to test out R on Apple’s new M1/ARM architecture. Here’s how it went:
Many textbooks when introducing ordinary least squares as a method to do regression (usually via introducing linear regression) don’t explain / derive the coefficients. Here, I explain how to do it using tensor/index notation.
Using mathematics to analyze who is a “better” player By: Dr. Ikjyot Singh Kohli It seems that nowadays one cannot escape the never-ending debate of “Who is better? Michael Jordan or Lebron James?” The situation is made worse by organizations like ESPN that have endless debates on their various shows combined with releasing lists of […]
By: Dr. Ikjyot Singh Kohli I wrote an extensive script in R that takes the most recent data available for the number of new/confirmed COVID-19 cases per day by location and computes the probability using statistical learning that a selected location will observe a new COVID-19 case (probability of observing a non-zero daily percentage return). […]
