- shift from shooting to performance in general Can one's shooting in college basketball be used as a predictor for their NBA shooting?
#currently needs a lot of tweeking! (look at last year college / first year rookie season only) data collected using nba_py python library, developed by seemethere and found at https://github.com/seemethere/nba_py current data takes into regards all seasons of a player and all season of in college - maybe tweek this.
College Statistics
NBA Statistics (Regular Season)
NBA TS vs College FT
p-value = 0.02856, reject the null hypothesis statistical evidence of a relationship between NBA TS and College FT
Looking towards this year's draft - regardless of defensive ability, who is predicted to be a good shooter? Finally can work on this since exams are over!!
to do list:
- get stats of active nba players - Done
- get stats of their college statistics - Done <- somehow I missed player names by accident
- introduction - is college ts/fg/3pt/ft% a good predictor for nba ts? fg? 3-pt? regression analysis - simple linear regression - 3x3 plot regression analysis - simple linear regression of true shooting confidence intervals t-testing - reject/don't reject slope 0 (existence of a relationship) - Done
- looking draft to draft, what differentiated the good players from the rest? Was there any way to determine a bust? interactive plots from ... to today - distributions of various statistics (PER/ppg/apg/rpg/shooting splits) - see if anyone stood out. Try multiple regression? Response variable being first year NBA PER, relating to college PER, defensive stats, etc. <- t-tests/f-tests to find variables and obtain model.
- looking towards this year's draft