A Website with Personalized NBA Statistics

SliderStats is the all-in-one website that allows users to compare an unlimited number of stats between players. Users can weigh each stat, providing a completely customizable and more importantly, an objective approach to comparing performance metrics. Any basketball fan, ranging from the average casual to full-time ESPN analyst can benefit from SliderStats.

Unlike other websites for basketball statistics that only show raw stats, SliderStats uses a weighted sum ranking to provide further insight into multiple stats between players within a season.

Visit our website: sliderstats.streamlit.app

Features

Multitude of Statistics

• Per Game Stats: Analyze player and team performance on a per-game basis.
• Total Stats: View cumulative statistics for players and teams.
• Head-to-Head Comparison: Compare stats between two players or teams.
• Clutch Stats: Examine performance during crucial moments of the game for specific players or teams.
• Playoff Stats: Focus on player performance during the playoffs.

Trademark Sliders

• Interactive Sliders: Use sliders to filter and explore data interactively.

Filtering

• Year: Filter ranking by year.
• Position: Filter ranking by position.
• Statistics: Choose which statistics to use.
• Range of Stats: Specify the range for each of your stats.

Player Rankings

After filtering and adjusting sliders, see the final rankings of players/teams and their stats.

Code

For our data, we use an API from github called nba_api which provides continuously updated data from the NBA's most current games and even games from decades before.

We use a variety of Python libraries, such as Pandas and Streamlit, to format and display the statistics. Below is a snippet of the code used to filter statistics.

for stat in selected_stats_per_game: sliders[stat] = st.sidebar.slider(f'Importance of {stat}', -1.0, 1.0, 0.5, 0.01) st.sidebar.title("Range of Stats") for stat in selected_stats_per_game: if stat in filtered_data_perGame: sliders_filter[stat] = st.sidebar.slider(f'Range of {stat}', float(filtered_data_perGame[stat][0]), float(filtered_data_perGame[stat][1]), (float(filtered_data_perGame[stat][0]), float(filtered_data_perGame[stat][1])))

# Exclude all of the rows that are not in the player's range
filtered_df = season_perGame_active_players.copy()
for col in season_perGame_active_players.columns[1:]:
    if col in sliders_filter:
        filtered_df = filtered_df[filtered_df[col].between(*sliders_filter[col])]

season_perGame_active_players = filtered_df

# Filter the DataFrame based on selected stats
filtered_data = season_perGame_active_players[selected_stats_per_game]


# Calculate weighted sum based on user-defined importance
normalized_dataPG = (filtered_data - filtered_data.min()) / (filtered_data.max() - filtered_data.min())
# Calculate weighted sum based on user-defined importance
weighted_totalsPG = normalized_dataPG * pd.Series(sliders)

# Rounding for all per game stats
season_perGame_active_players[columns_to_keepPG] = season_perGame_active_players[columns_to_keepPG].round(decimals=2)

# Calculate the total weighted sum for each player
season_perGame_active_players['Weighted_Sum'] = weighted_totalsPG.sum(axis=1)
# Display sorted players based on weighted sum
sorted_playersPG = season_perGame_active_players.sort_values(by=['Weighted_Sum', 'PLAYER_NAME'], ascending=[False, True])

if drop_unused_stats:
    sorted_playersPG = pd.concat([sorted_playersPG['PLAYER_NAME'], sorted_playersPG[selected_stats_per_game], sorted_playersPG['Weighted_Sum']], axis = 1)

st.write("Season Per Game Stats (Regular Season) for Active NBA Players:")
st.write(sorted_playersPG)

Technologies

• Python
• Pandas
• Streamlit
• nba_api

Contributors

Below are the key roles and team members involved in this project:

• Project Managers: Juan Diaz and Jason Yang
• Project Members: Alex Qi and Ryan Uyeki
• Media Members: Ravinit Chand and Yaelle Kretchmer
• Business Member(s): Aashritha Javvaji