What Matters in Winning College Basketball Games?

Data Source

• All data is scraped with python from sports-reference.com code is below.

Filters

• All college basketball teams from 2000 to 2019.

Key Points

• No one statistical item matters.

• FG percentage matters the most but not all the much.

• Was surprised the R^2 of steals per game isn't negative.

• I think going deeper would make more sense in how college basketball teams are winning.

---

---

Python Script

#pip install beautifulsoup4

#pip install lxml

#pip install requests

import pandas as pd

import requests

import bs4

from bs4 import BeautifulSoup

n=range(2000,2020)

url2 ='https://www.sports-reference.com/cbb/seasons/'

url3 ='-school-stats.html'

## initialize all of the lists but make sure they don't refer to the same empty list

pl_team_list, pl_wins_list, pl_loses_list, pl_total_points_list, pl_total_points_againest_list, \

pl_FG_percentage_list, pl_3_percentage_list, pl_FT_percentage_list, pl_total_rebounds_list, \

pl_assist_list, pl_steals_list, pl_turnovers_list = ([] for i in range(12))

for n in n:

all = url2+str(n)+url3

r = requests.get(all)

soup = BeautifulSoup(r.text, 'html.parser')

league_table = soup.find('table', class_ = 'per_match_toggle sortable stats_table')

for team in league_table.find_all('tbody'):

rows = team.find_all('tr')

for row in rows:

pl_team = row.find('td', class_ = 'left')

if pl_team == (None):

continue

pl_wins = row.find_all('td', class_ = 'right')[1]

if pl_wins == (None):

continue

pl_loses = row.find_all('td', class_ = 'right')[2]

if pl_wins == (None):

continue

pl_total_points = row.find_all('td', class_ = 'right')[16]

if pl_total_points == (None):

continue

pl_total_points_againest = row.find_all('td', class_ = 'right')[17]

if pl_total_points_againest == (None):

continue

pl_FG_percentage = row.find_all('td', class_ = 'right')[22]

if pl_FG_percentage == (None):

continue

pl_3_percentage = row.find_all('td', class_ = 'right')[25]

if pl_3_percentage == (None):

continue

pl_FT_percentage = row.find_all('td', class_ = 'right')[28]

if pl_FT_percentage == (None):

continue

pl_total_rebounds = row.find_all('td', class_ = 'right')[30]

if pl_total_rebounds == (None):

continue

pl_assist = row.find_all('td', class_ = 'right')[31]

if pl_assist == (None):

continue

pl_steals = row.find_all('td', class_ = 'right')[32]

if pl_steals == (None):

continue

pl_turnovers = row.find_all('td', class_ = 'right')[33]

if pl_total_rebounds == (None):

continue

#print(n,

# pl_team.text,

# pl_wins.text,

# pl_loses.text,

# pl_total_points.text,

# pl_total_points_againest.text,

# pl_FG_percentage.text,

# pl_3_percentage.text,

# pl_FT_percentage.text,

# pl_total_rebounds.text,

# pl_assist.text,

# pl_steals.text,

# pl_turnovers.text,

# )

#data = {'Year': n,

# 'Team': pl_team.text,

# 'Wins': pl_wins.text}

pl_team_list.append(pl_team.text)

pl_wins_list.append(pl_wins.text)

pl_loses_list.append(pl_loses.text)

pl_total_points_list.append(pl_total_points.text)

pl_total_points_againest_list.append(pl_total_points_againest.text)

pl_FG_percentage_list.append(pl_FG_percentage.int)

pl_3_percentage_list.append(pl_3_percentage.int)

pl_FT_percentage_list.append(pl_FT_percentage.int)

pl_total_rebounds_list.append(pl_total_rebounds.text)

pl_assist_list.append(pl_assist.text)

pl_steals_list.append(pl_steals.text)

pl_turnovers_list.append(pl_turnovers.text)

df = pd.DataFrame({

'Team': pl_team_list,

'Wins': pl_wins_list,

'Loses': pl_loses_list,

'Total Points': pl_total_points_list,

'Total Points Against': pl_total_points_againest_list,

'FG Percentage': pl_FG_percentage_list,

'3 Percentage': pl_3_percentage_list,

'FT Percentage': pl_FT_percentage_list,

'Total Rebounds': pl_total_rebounds_list,

'Assist': pl_assist_list,

'Steals': pl_steals_list,

'Turnovers': pl_turnovers_list

})

df.to_excel('C://Users/bphil/OneDrive/Desktop/ncaa_basketball_teams.xlsx')

df