This repository analyzes FIFA team performances based on metrics such as goals scored, goals conceded, possession, and shot conversion rates. The analysis aims to provide insights into team efficiency, defensive performance, and the relationship between possession and match outcomes. Moreover, I wanted to test my SAS skills to the test, so instead of doing with python, I chose SAS.
The data used in this analysis is obtained through web scraping from FIFA. The dataset has 78 attributes and 64 observations. For this study, I used the following key variables:
- Goals scored: The number of goals scored by each team.
- Goals Conceded: The number of goals conceded by each team.
- Possession Difference: The difference in possession percentage between the home and away teams.
- Shot Conversion Rate: The ratio of goals scored to total attempts by each team.
- Defensive Metrics: Clean sheets, goal preventions, defensive pressures, and forced turnovers.
The dataset consists of observations for various teams, detailing their performance across multiple matches.
The analysis employs several statistical techniques, including:
- Descriptive Statistics: Summarize the data, including means and standard deviations for various performance metrics.
- Logistic Regression: To model the probability of a team winning based on possession difference. This technique helps in understanding the relationship between possession and match outcomes.
- Data Visualization: Although not included in this document, visualizations can be created to represent the data graphically for better insights.
- Goals Conceded: The analysis shows that teams like England, Japan, and the Netherlands conceded a total of 4 goals, while teams like Tunisia conceded only 1 goal throughout the tournament
- Possession and Match Outcomes: The logistic regression results indicate that possession difference has a weak correlation with match outcomes. The mean possession difference for winning teams was slightly negative, suggesting that teams can win even with less possession. This was evident with the data from France. France's lack of high possession but more wins highlights France's tactical flexibility, efficiency in attack, defensive solidity, and pragmatic approach to achieving results.
- Shot Conversion Rates: The analysis reveals that teams like Saudi Arabia and Ghana had the highest shot conversion rates, indicating their efficiency in converting chances into goals
- Defensive Performance: Teams like Morocco and Croatia demonstrated strong defensive metrics, with Morocco achieving four clean sheets and a high defensive performance score
The results of the analysis are summarized as follows:
- Total Goals Conceded by Team: A detailed table of goals conceded by each team is provided, highlighting the defensive strengths and weaknesses of each team
- Logistic Regression Model: The model fit statistics indicate that the model did not show significant predictive power regarding match outcomes based on possession difference, with a likelihood ratio p-value of 0.4059
- Team Efficiency: The shot conversion rates highlight the most efficient teams in terms of scoring relative to their attempts
This analysis provides valuable insights into team performances in the FIFA tournament, highlighting the importance of offensive and defensive strategies. The findings can help coaches, analysts, and fans better understand the dynamics of the game.
Future analyses could include:
- Incorporating more advanced machine learning techniques to predict match outcomes.
- Analyzing player-level data for a more granular understanding of performance.
- Visualizing the data to enhance interpretability and insights.
Thanks to the FIFA organization for providing the data and the community for their contributions to football analytics.
Feel free to clone this repository and explore the analysis further! If you have any questions or suggestions, please open an issue or submit a pull request.