Dark Mode
College Football Bowl Games Historical Data
Sports & Recreation
Tags and Keywords
Trusted By
"No reviews yet"
Free
About
This dataset offers a detailed collection of every college football bowl game played from 1902 to 2022 [1]. It provides a historical record for analysis, research, and general interest in college football. The data was gathered from Sports Reference and then prepared for data analysis [1].
Columns
- id: A Primary Key for each record [2, 3].
- year: The year in which the game took place. This column covers games from 1901 to 2021 [1, 4].
- date: The specific date of the game, ranging from 1st January 1902 to 10th January 2022 [1, 4, 5].
- day: The day of the week on which the game was played. There are 7 unique days recorded, with 'Saturday' being the most frequent [1, 5].
- winner_tie: The name of the winning team or if the game resulted in a tie. 'Alabama' is the most frequent winner [1, 5]. There are 166 unique winners or ties [5].
- winner_rank: The ranking of the winning team at the time of the game. Approximately 51% of entries for this column are missing, indicating that not all winning teams were ranked [1, 5]. There are 26 unique ranks, with '3' being the most common ranked winner [5].
- winner_points: The total points scored by the winning team. Scores range from 0 to 70, with a mean of 30.3 points [1, 6].
- loser_tie: The name of the losing team or if the game resulted in a tie. 'Alabama' and 'Michigan' are among the most frequent losers [2, 6, 7]. There are 171 unique losers or ties [7].
- loser_rank: The ranking of the losing team at the time of the game. Around 56% of entries for this column are missing [2, 7]. There are 26 unique ranks, with '4' being the most common ranked loser [7].
- loser_points: The total points scored by the losing team. Scores range from 0 to 61, with a mean of 17.1 points [2, 7].
- attendance: The total number of people who attended the bowl game. This column has a small percentage of missing data (1%), with a mean attendance of approximately 49,500 people and a maximum recorded attendance of 107,000 [2, 8].
- mvp: Details of the Most Valuable Player, including their position, team, and name [2]. This column has 11% missing data, with 1331 unique MVP entries [9].
- sponsor: The sponsor of the game [2]. This column has 46% missing data, with 'Outback Steakhouse' being the most common sponsor among the 166 unique entries [9].
- bowl_name: The specific name of the bowl game. 'Rose Bowl' is the most common, accounting for 7% of games, followed by 'Orange Bowl' at 6% [2, 9]. There are 77 unique bowl names [9].
Distribution
The dataset is provided in a tabular format [1], specifically as a CSV file (
collegefootballbowl.csv) [2]. It contains 1527 records or rows and 14 distinct columns [2, 3]. Data is well-structured, with a very low percentage of mismatched entries across all fields [3-9]. Key fields like year, date, winner_points, loser_points, and bowl_name are 100% valid with no missing entries [4-7, 9]. Some fields such as winner_rank, loser_rank, mvp, sponsor, and attendance have missing values [5, 7-9]. The id column serves as a Primary Key [2, 3].Usage
This dataset is ideal for:
- Sports analytics: Analysing team performance, scoring trends, and win/loss records over time.
- Historical research: Studying the evolution of college football bowl games and their significance.
- Statistical modelling: Developing models to predict game outcomes or attendance figures.
- Educational purposes: Providing a real-world dataset for data analysis and visualisation exercises.
- Fan engagement: Providing detailed information for college football enthusiasts.
Coverage
The dataset covers college football bowl games spanning an extensive time range from 1902 to 2022 [1]. Geographically, the data pertains to United States college football [2]. There are no specific demographic details within the dataset itself, as it focuses on game statistics. Data availability for rankings and sponsors varies, with a notable percentage of missing values for these attributes [5, 7, 9].
License
CC0: Public Domain
Who Can Use It
- Data Scientists and Analysts: For quantitative analysis and pattern recognition in sports data.
- Sports Researchers: To investigate historical trends and the development of college football.
- Educators and Students: As a practical dataset for learning data manipulation, analysis, and visualisation.
- Sports Enthusiasts: For personal exploration of game statistics and historical records.
Dataset Name Suggestions
- College Football Bowl Game History
- Historic College Bowl Results
- American College Football Bowls Data
Attributes
Original Data Source: College Football Bowl Games Historical Data
QUALITY