# Files:
**train_NBA.csv** — Historical NBA draft data from 1989 to 2015, including player stats and the target variable: value_over_replacement.

**test_NBA.csv** — Draft data from 2016 to 2021. The target variable is hidden; your model must predict it.

**submission_NBA.csv** — Shows the correct submission format. Contains id and placeholder values for value_over_replacement.

## What You’re Predicting:
Your task is to predict the value_over_replacement (VORP) for players in the test set — a numerical value that reflects a player’s total career contribution relative to a replacement-level player.

## Columns
Below is a description of the most important columns in the dataset:

Column Name,	Description
id:	Unique identifier for each player (used in submission)
year:	Year the player was drafted
rank:	Prospect scouting rank prior to the draft
overall_pick:	Overall draft position (e.g., 1st pick, 2nd pick, etc.)
team:	NBA team that drafted the player
player:	Player's full name
college:	College or international team the player came from
years_active:	Number of seasons played in the NBA
games:	Total games played in career
minutes_played:	Total career minutes played
points:	Total points scored
total_rebounds:	Total rebounds recorded
assists:	Total assists made
field_goal_percentage:	FG% (career shooting accuracy from the field)
3_point_percentage:	3P% (career shooting accuracy from 3-point range)
free_throw_percentage:	FT% (career shooting accuracy from the free-throw line)
average_minutes_played:	Minutes played per game over career
points_per_game:	Average points per game
average_total_rebounds:	Average rebounds per game
average_assists:	Average assists per game
win_shares:	Total career win shares (advanced metric)
win_shares_per_48_minutes:	Win shares normalized per 48 minutes played
box_plus_minus:	Advanced stat estimating performance vs. league average
value_over_replacement:	**Target variable** — estimated value above a replacement-level player

## Notes:
Some players may have short careers, impacting totals.

Shooting percentages are expressed as decimals (e.g., 0.455 for 45.5%).

Not all players played in the NBA (e.g., international draftees who didn’t sign).