#importing libraries
import pandas as pdPremier League Performance Metrics and Results: A Dynamic Analysis
Deciphering the Statistics That Support Football Performance in the 2021–2022
Dataset
#importing csv into pandas dataframe
import_data = pd.read_csv("data/soccer21-22.csv")Full Preview of Data
| variable | class | description |
|---|---|---|
| Date | character | The date when the match was played |
| HomeTeam | character | The home team |
| AwayTeam | character | The away team |
| FTHG | double | Full time home goals |
| FTAG | double | Full time away goals |
| FTR | character | Full time result |
| HTHG | double | Halftime home goals |
| HTAG | double | Halftime away goals |
| HTR | character | Halftime results |
| Referee | character | Referee of the match |
| HS | double | Number of shots taken by the home team |
| AS | double | Number of shots taken by the away team |
| HST | double | Number of shots on target by the home team |
| AST | double | Number of shots on target by the away team |
| HF | double | Number of fouls by the home team |
| AF | double | Number of fouls by the away team |
| HC | double | Number of corners taken by the home team |
| AC | double | Number of corners taken by the away team |
| HY | double | Number of yellow cards received by the home team |
| AY | double | Number of yellow cards received by the away team |
| HR | double | Number of red cards received by the home team |
| AR | double | Number of red cards received by the away team |
Introduction
Evan Gower’s work on Kaggle has made it possible to obtain statistics from 380 matches, giving us a detailed glimpse at the English Premier League season of 2021–2022. Along with comprehensive statistics for both home and away sides, such as goals, shots, fouls, and cards, it also provides important game information such team names, match dates, and referees. This dataset, which contains data on halftime performance as well as full-time results, is an invaluable resource for anyone wishing to examine the factors that affect football match outcomes, from individual player affects to team plans.
Dataset Description:
The 2021–2022 English Premier League (EPL) season’s match-day statistics are contained in this dataset, which is sourced from Evan Gower’s work on Kaggle. It is taken from the official Premier League website and has been carefully cleaned. It includes 380 matches, which represents the whole season. Every entry contains an abundance of information, such as the teams involved, the date of the match, the referee, and a variety of in-game statistics for both the home and away sides, including goals, shots, fouls, and cards. It also provides full-time results and halftime analytics, providing a detailed analysis of each match’s dynamics. This dataset offers a basis for more in-depth research into the variables impacting team performance and match results in addition to providing a numerical narrative of the season.
Reasons for Choosing this Dataset:
This dataset captures the attention of millions because of the rich detail and storytelling potential. It is a goldmine for football enthusiasts and can provide deep details within one of the world’s most followed sports leagues. Football is a tasteful concoction of skill, strategy, and raw athleticism. This complex tapestry allows for a broad spectrum of unique opportunities to be statistically explored: fouls affecting game outcomes, team discipline, and shot efficacy. It also enables a multifaceted investigation into the relationship between in-game occurrences and success or failure, offering insights that may be of use to observers, coaches, and even players who are trying to improve their tactics. By include halftime data, the analysis gains additional depth and an interesting new perspective on momentum fluctuations within games.
Questions
Question 01: What is the connection between in-game metrics such as shots on goal, fouls committed, and cards received, and the outcomes of soccer matches? Can these metrics help in creating a predictive model to forecast whether the match results will favor the home or away team?
Question 02: How do the outcomes of soccer matches vary under the premise that all games concluded at halftime? How does this disparity differ from team to team, and what is the influence of the overall outcome of the league championship?
Analysis plan
For Question 01:
Introduction: The first query explores the complex connections between in-game metrics—like shots on goal, fouls, and cards—and how those relationships affect the results of football matches. Our goal is to use these factors to investigate if they can accurately forecast match outcomes that will benefit the home team or the away team. This question is especially fascinating since it addresses fundamental football dynamics and provides information on how different gameplay elements affect a team’s eventual success or failure.
The variables involved: Full-time home and away goals ("FTHG" and "FTAG") are crucial indicators of match outcomes. Full-time results (“FTR”) will serve as the target variable for predictive modeling. Number of shots taken by the home and away team (“HS” and “AS”) and shots on target (“HST” and “AST”) reflect attacking performance. Number of fouls by the home and the away team (“HF” and “AF”) and cards received (“HY”, “AY”, “HR”, and “AR”) indicate team discipline and aggression.
Data Preparation: Combining match information with team-level season summaries will indeed provide valuable context for analysis and assist in identifying trends and outliers. The dataset’s consistency across all rows and columns, without any missing or erroneous values, ensures the reliability of our analysis. Additionally, no additional calculations or adjustments to rows are required, streamlining the process for further analysis.
Analysis: The dataset will be partitioned into training and testing sets to aid in model development and assessment. Regression techniques, including logistic regression or decision trees, will be applied to unveil associations between in-game performance metrics and match outcomes. This examination will enable the identification of patterns and relationships pivotal for constructing predictive models aimed at determining match results.
Discussion: The initial examination of the data, which concentrated on factors like as cards, fouls, and shots on goal, points to a complex relationship between these measurements and game results. Higher foul and card counts may not always indicate aggressive play, but shots on target and other planned offensive efforts may imply a more direct correlation between winning and succeeding. Because of the data’s consistency and lack of missing values, the analysis is streamlined and these associations may be examined in greater detail. Predictive modelling and trend analysis may provide trends that support received wisdom in football or unearth winning unexpected tactics.
For Question 02:
Introduction: In question two, we analyze a hypothetical scenarios to offer practical insights derived from the European Premier League’s 22nd season. By exploring how matches might end at halftime, we aim to analyze the real-world dynamics of team performance. Specifically, we will investigate whether successful teams exhibit early aggression in the first half or stage comebacks in the second half, and whether team rankings reflect consistent performance or fluctuate based on halftime strategies. This analysis offers practical implications for coaching strategies, player development, and tactical approaches, providing actionable insights to enhance team performance and strategic decision-making in professional football. By understanding the significance of second-half performances and their impact on league standings, teams can refine their approaches to enhance their competitive edge and adapt to the evolving landscape of professional football.
The variables involved: Full-time home and away goals (“FTHG” and “FTAG”) are necessary to determine team placement. Full-time results (“FTR”) denote the final actual outcome of the match. Half-time goals for the home and away teams, as well as the half-time result (“HTHG”, “HTAG”, and “HTR”), are important variables for the premise of our question. Additionally, the home team and away team (“HomeTeam” and “AwayTeam”) will identify who will win the league in our situation.
Data Preparation: In the data preparation phase, thorough checks ensure the integrity and consistency of the dataset. Any identified discrepancies are meticulously documented to ensure transparency. Visualizations are employed to illustrate changes in match outcomes and league standings, providing insights into the impact of simulated halftime results on soccer match dynamics and championship outcomes. Notably, the dataset exhibits no missing values.
Analysis: The dataset is organized chronologically, facilitating the computation of halftime results based on goals scored by both home and away teams (HTHG and HTAG). These halftime results are then compared with full-time results (FTR). A function is developed to iterate through the data, calculating halftime results (win, draw, loss) for each team and ranking them based on points. The recalibration of match outcomes assumes all games conclude at halftime, enabling a simulated scenario where leading teams at halftime are credited with wins, and tied matches result in draws for both teams. The reevaluated outcomes are subsequently contrasted with current league standings to identify any shifts in team rankings and overall league dynamics. Visualizations are utilized to enhance the interpretation of play styles and consistent team performance.
Discussion: The crucial significance that second-half play plays in football matches is highlighted by the simulation of halftime conclusions. According to preliminary research, some clubs might depend more on strong finishes to win games, which could cause the league rankings to change in this fictitious situation. This study not only emphasises how crucial it is to continue performing for the duration of the game, but it also encourages discussion about team tactics and how well they can be adjusted to suit various game developments. The dependability of these insights is increased by the absence of missing data, offering a strong basis for conjectural debates about team dynamics and league results.