Had some free time and trained a XGBoost model on historical transfer data to project next year’s BPM for players in the portal. The photo below is the model’s top 10 players. The numbers in grey are a 90% confidence interval of the 26-27 BPM value.
Latest posts 5
Gator-nomics.
New Substack post on the Economics based strategy from the Florida Gators and their attempt to foul vs Iowa.
open.substack.com/pub/jruhanen/p…
New Substack post on the Economics based strategy from the Florida Gators and their attempt to foul vs Iowa.
open.substack.com/pub/jruhanen/p…
jruhanen.substack.com Gator-nomics The reason behind why Florida tried to foul Iowa while leading by 2. This chart represents what the current NBA Rookie Class PPS was compared to their Expected PPS in the 24-25 college szn.
I included their NBA xFG difference, but this is unavailable for the majority of these players. I assume this is due to lack of an adequate sample of shots.
I included their NBA xFG difference, but this is unavailable for the majority of these players. I assume this is due to lack of an adequate sample of shots.
This is a new stat I developed called CAST. It stands for Contests, Assists, Screening, Tenacity. The metric is meant to measure who the best complimentary pieces are in the NBA.
Bottom 10 and more information in the following thread…
Bottom 10 and more information in the following thread…
Here is my personal portfolio! You can use a demo player data base where I use my scouting reports. There is also a Hustle Stat app using XML data. On top of this, I attached a demo postgame report. There is also a link to my substack and a contact me. jruhanen.github.io 1/3
