Football League Matches Research Articles

AI-based betting anomaly detection system to ensure fairness in sports and prevent illegal gambling

This study develops a solution to sports match-fixing using various machine-learning models to detect match-fixing anomalies, based on betting odds. We use five models to distinguish between normal and abnormal matches: logistic regression (LR), random forest (RF), support vector machine (SVM), the k-nearest neighbor (KNN) classification, and the ensemble model—a model optimized from the previous four. The models classify normal and abnormal matches by learning their patterns using sports betting odds data. The database was developed based on the world football league match betting data of 12 betting companies, which offered a vast collection of data on players, teams, game schedules, and league rankings for football matches. We develop an abnormal match detection model based on the data analysis results of each model, using the match result dividend data. We then use data from real-time matches and apply the five models to construct a system capable of detecting match-fixing in real time. The RF, KNN, and ensemble models recorded a high accuracy, over 92%, whereas the LR and SVM models were approximately 80% accurate. In comparison, previous studies have used a single model to examine football match betting odds data, with an accuracy of 70–80%.

Factors that affect handball execution in Australian Football

ABSTRACT A handball is one of two technical skills used to dispose of the ball in Australian Football. Previous research has only considered handball effectiveness in the analysis of team performance and there is a need to understand whether there are other more important characteristics of handball execution that explain effectiveness (i.e., performance). 1342 handballs from Australian Football League matches were analysed. Ten characteristic variables were created that represent the context and execution of each handball included in the analysis. A mixed effects generalised linear model was used to evaluate the effect of the (ten) characteristics on handball outcome. Three out of ten characteristics were associated with handball outcome, and these related to the posture of the handballer and the type of defensive pressure applied to the handballer and the receiver of the handball. These findings explain both how to increase handball effectiveness and how to reduce the handball effectiveness of an opposition team. Given the important role of handballing in passing sequences and maintaining ball possession, the practical application of these findings could enhance overall team performance.

PREDIKSI HASIL PERTANDINGAN SEPAK BOLA LIGA PREMIER INGGRIS DENGAN ARTIFICIAL NEURAL NETWORK BACKPROPAGATION

This study aims to enhance the accuracy of predicting English Premier League football match outcomes by utilizing a partially updated Artificial Neural Network (ANN) model based on match outcome data from the period 2017 to 2021. In this research, various statistical features such as the number of goals scored in the first half and the number of shots on target were incorporated as inputs to the ANN model. The match outcome data was normalized to improve the model's performance. The ANN model employed multiple hidden layers with ReLU (Rectified Linear Unit) activation functions and was trained using the Backpropagation algorithm. Throughout the training process, the model was periodically updated to reflect changes in match patterns over time. The research findings reveal that the ANN model with partial updates can predict football match outcomes with an accuracy of 77.89% in the final iteration, with a Mean Squared Error (MSE) of 0.769 and a Mean Absolute Error (MAE) of 0.689. Additionally, the prediction results are visualized in the form of a distribution graph comparing actual match outcomes with the predictions from the final iteration, providing a visual representation of the model's performance. This study makes a significant contribution to the development of modeling techniques for forecasting football match outcomes and underscores the importance of partial updates in adapting to changes in match patterns over time, offering potential for improvements in football match analysis and prediction in the future

An observational study of player behaviours under varying task constraints in professional Australian football players

An evaluation of player behaviour and performance across different playing conditions in team sport is difficult. This study examined integrative (indegree and outdegree importance) and individual (goals, kicks, handballs, marks, tackles, kick and handball proficiency) level behaviours at one professional Australian Football club across three different conditions: small-sided games (SSGs), match simulation and competitive Australian Football League (AFL) matches. Estimated marginal means identified negligible changes for integrative measures across conditions. This may suggest larger changes to control parameters are required to modify these behaviours, as it relies on interdependent cooperative behaviour. Individual level behaviours demonstrated varied responses across changing conditions. AFL matches revealed the lowest kicks and marks per minute compared to the other conditions, whilst having the highest rate of tackles. Players had lower kick proficiency in conditions with a higher number of players. Moreover, one SSG with a floater (player who participates in attack for both teams) elicited a higher number of handballs per minute relative to every other condition. These findings provide implications for measuring the (in)stability of individual and integrative level behaviours and how changes to constraints may alter these metrics. This method may provide a foundation for future hypotheses testing and for player monitoring that can assist practitioners to examine the changes in players behaviours in response to changing constraints.

Tactically Maximize Game Advantage by Predicting Football Substitutions Using Machine Learning

Football (also known as Soccer), boasts a staggering fan base of 3.5 billion individuals spread across 200 countries, making it the world’s most beloved sport. The widespread adoption of advanced technology in sports has become increasingly prominent, empowering players, coaches, and team management to enhance their performance and refine team strategies. Among these advancements, player substitution plays a crucial role in altering the dynamics of a match. However, due to the absence of proven methods or software capable of accurately predicting substitutions, these decisions are often based on instinct rather than concrete data. The purpose of this research is to explore the potential of employing machine learning algorithms to predict substitutions in Football, and how it could influence the outcome of a match. This study investigates the effect of timely and tactical substitutions in football matches and their influence on the match results. Machine learning techniques such as Logistic Regression (LR), Decision tree (DT), K-nearest Neighbor (KNN), Support Vector Machine (SVM), Multinomial Naïve Bayes (MNB), Random Forest (RF) classifiers were implemented and tested to develop models and to predict player substitutions. Relevant data was collected from the Kaggle dataset, which contains data of 51,738 substitutions from 9074 European league football matches in 5 leagues spanning 6 seasons. Machine learning models were trained and tested using an 80-20 data split and it was observed that RF model provided the best accuracy of over 70% and the best F1-score of 0.65 on the test set across all football leagues. SVM model achieved the best Precision of almost 0.8. However, the worst computation time of up to 2 min was consumed. LR showed some overfitting issues with 100% accuracy in the training set, but only 60% accuracy was obtained for the test set. To conclude, based on the time of substitution and match score-line, it was possible to predict the players who can be substituted, which can provide a match advantage. The achieved results provided an effective way to decide on player substitutions for both the team manager and coaches.

Real time prediction of match outcomes in Australian football

ABSTRACT This study aimed to determine whether machine learning models based on technical performance and not score margin could be used to predict end-of-match outcome of Australian football matches in real-time. If efficacious, these models could be used to generate insights about team performance and support the decision-making of coaches during matches. A database of 168 team technical performance indicators from 829 Australian Football League matches played between 2017 and 2021 was used. Two feature sets (data-driven and data-informed) were used to train and evaluate six models (generalised linear model, random forest and adaboost) on match outcome prediction (Win/Loss) over 120 epochs (a representation of normalised time during each match). All models performed well (mean classification accuracy = 73.5–75.8%) in comparison with a benchmark score-based model (mean classification accuracy = 77.4%). Data-informed feature sets performed better than data-driven in most cases. Classification accuracy was low at the start of a match (45.7–48.8%) but increased to a peak near the end of a match (87.2–92.7%). These findings suggest that any of the employed models can be used to formulate in-match decision support. The model which is best in practice will depend on factors such as time-cost trade-off, feasibility and the perceived value of its suggestions.

Anticipatory effects on side-step cutting biomechanics in Women’s Australian Football League players

ObjectivesReactive side-step cutting manoeuvres are linked to anterior cruciate ligament (ACL) injuries in Women’s Australian Football League (AFLW) matches. We explored knee joint moments and ground reaction forces (GRFs) in...

Optimization of the C4.5 Algorithm Using Particle Swarm Optimization and Discretization in Predicting the Results of English Premier League Football Matches

Football is one of the most popular sports. One of the most competitive football competitions is the English Premier League. This study aims to determine the prediction of the results of the football match in English Premier League. The prediction results in the form of home win, away win, and draw. This prediction uses data mining techniques, namely using the C4.5 algorithm as a classification algorithm with Particle Swarm Optimization as a feature selection method and Discretization as a preprocessing method. The dataset used was obtained from the football-data.co.uk website for four league seasons from the 2017/2018 season to the 2020/2021 season with a total of 1,520 instances. In this study, a comparison was made to the methods used to determine the increase in accuracy obtained. Based on ten times the data mining process, the final result of the best accuracy from using the C4.5 algorithm is 57.24%, then the C4.5 algorithm with Discretization gets an accuracy of 65.13%, and the C4.5 algorithm with Discretization and Particle Swarm Optimization gets accuracy of 71.05%. The conclusion is that the use of Discretization and Particle Swarm Optimization can improve the performance of the C4.5 algorithm in predicting the results of English Premier League matches with an increase in accuracy of 13.81%.

Anticipatory Effects on Side-Step Cutting Biomechanics in Women’s Australian Football League Players

Introduction: Reactive side-step cutting manoeuvres are a common task linked to anterior cruciate ligament (ACL) injuries in Women’s Australian Football League (AFLW) matches. Knee loads associated with increased ACL loading and strain are typically elevated when performing unanticipated versus anticipated changes of direction. We explored knee joint moments and ground reaction forces (GRFs) in AFLW players when performing anticipated and unanticipated side-step cuts.

Risk factors for injuries related to player substitutions during Korean professional football league matches

Risk factors for injuries related to player substitutions during Korean professional football league matches

Statistical modelling of goalkicking performance in the Australian Football League

ObjectivesAustralian football goal kicking is vital to team success, but its study is limited. Develop and apply Bayesian models incorporating temporal, spatial and situational variables to predict shot outcomes. The models aim to (i) rank players on their goal kicking and (ii) create clusters of statistically similar players and rank these clusters to provide generalised recommendations about player types. DesignRetrospective longitudinal study with goal kicking data from three seasons, 2018–2020, 576 official Australian Football League matches, containing 26,818 attempts at goal from 778 players. MethodsThe Bayesian ordinal regression model enables descriptive analysis of goal kicking performance. The models include spatial variables of distance and kick angle, situational variables of shot type and player or cluster with interaction terms. Alternative models included situational variables of weather and player characteristics, spatial variables of stadium location and temporal variables of time and quarter. Approximate leave-one-out cross validation was used to test the model. ResultsOverall goal rate of 47% (12,600), behind rate of 35% (9373) with misses the remaining 18% (4845). Accuracy of both player and cluster model achieved 0.51 against an uninformed (predict goal) model result of 0.47. The models allow for analysis of goal kicking accuracy by distance and angle and analysis of player and player-type performance. ConclusionsWhile credible intervals for all players for set shots and general play were relatively large, some 95% credible intervals excluded zero. Therefore, it may be concluded that some players' goal kicking skill can be quantified and differentiated from other players.

PREDICTIONS OF TOP 5 EUROPEAN FOOTBALL LEAGUE MATCHES

Beating the bookmakers is something that is nearly impossible task to achieve. However, in this paper we propose a model to predict winning odds of major European league matches. Particularly, the predictions were done on top 5 (England, Spain, Italy, France and Germany) football league matches. To create an accurate predictor, we used Machine Learning to build an Artificial Neural Networks. Using the predicted probabilities, we introduce betting strategies and analyze their profitability using Monte Carlo simulations. To create more precise model, we set predictor variables such as ratings of the competing teams, head to head statistics and home advantage. We also estimate the predictability of top 5 European football competitions. The analysis is done based on historical data of matches and the closing odds on outcomes of matches. Finally, we deploy the results of the predictions by creating a web application which gives the predictions on the upcoming football matches of the top 5 European leagues.

Effects of Match Location, Quality of Opposition and Match Outcome on Match Running Performance in a Portuguese Professional Football Team.

The aim of this study was to analyze the effects of match location, quality of opposition and match outcome on match running performance according to playing position in a Portuguese professional football team. Twenty-three male professional football players were monitored from eighteen Portuguese Football League matches during the 2019–2020 season. Global positioning system technology (GPS) was used to collect time-motion data. The match running performance was obtained from five playing positions: central defenders (CD), fullbacks (FB), central midfielders (CM), wide midfielders (WM) and forwards (FW). Match running performance was analyzed within specific position and contextual factors using one-way analysis of variance (ANOVA) for repeated measures, standardized (Cohen) differences and smallest worthwhile change. CM and WM players covered significantly greater total distance (F = 15.45, p = 0.000, η2 = 0.334) and average speed (F = 12.79, p < 0.001, η2 = 0.294). WM and FB players covered higher distances at high-speed running (F = 16.93, p = 0.000, η2 = 0.355) and sprinting (F = 13.49; p < 0.001, η2 = 0.305). WM players covered the highest number of accelerations (F = 4.69, p < 0.001, η2 = 0.132) and decelerations (F = 12.21, p < 0.001, η2 = 0.284). The match running performance was influenced by match location (d = 0.06–2.04; CI: −0.42–2.31; SWC = 0.01–1.10), quality of opposition (d = 0.13–2.14; CI: –0.02–2.60; SWC = 0.01–1.55) and match outcome (d = 0.01–2.49; CI: −0.01–2.31; SWC = 0.01–0.35). Contextual factors influenced the match running performance with differential effects between playing positions. This study provides the first report about the contextual influence on match running performance in a Portuguese professional football team. Future research should also integrate tactical and technical key indicators when analyzing the match-related contextual influence on match running performance.

The influence of ball possession, passes and shots on target in winning premier league football matches

Football is the most popular sport in the world and involves the availability of information of footballers and teams to investigate success in this sport. In account of these aspects, the current study aims to analyze the influence of ball possession time, passes and shots on target in winning Premier League football matches. The methodology adopted consists of a quantitative, descriptive and documentary research, in which the sample is represented by all the 380 Premier League (PL) matches of the 2015/2016 season. The analysis process was made by the use of the association technique, present in the data mining process, with the support of the Weka software. The results suggest that shots on target, as well as its interaction with ball possession and passes - total passes, accuracy of passes, accuracy of passes inside the opponent’s field and long passes - influenced the final result of the match. Based on the results, the conclusion is that direct play alternatives, such as presenting less ball possession time and more long passes, associated with more shots on target, showed to be more beneficial for football clubs to win football matches.

Acceleration and High-Speed Running Profiles of Women's International and Domestic Football Matches.

Acceleration and deceleration are important given football is an intermittent sport with constant changes in velocity and direction. It is unclear, however, if the accelerations and decelerations performed by players differ between competition levels. The aim of the present study was to compare the acceleration, deceleration, and high-speed running profiles of players during international and domestic matches and to determine if differences were apparent across playing positions (defenders, midfielders, and attackers). GPS data from 21 Australian women's domestic football league matches over 2 seasons (2016–2018) and 15 Australian women's international matches (2017–2018) were collected and analyzed. Movement pattern data was collected using VX Sport and GPSports 10 Hz GPS receivers. Variables analyzed included: total distance, distance covered high-speed running (16–20 km·h−1) distance covered sprinting (> 20 km·h−1) and time spent accelerating and decelerating in four predetermined bands (1–2 m·s−2, 2–3 m·s−2, 3–4 m·s−2, and > 4 m·s−2). Results revealed that players competing in international matches covered significantly greater total distances, greater high-speed running distances and greater sprinting distances as well as spending a greater duration accelerating in band 4 compared to players in domestic competitions (p < 0.05). Players competing in international matches spent significantly less duration decelerating in bands 2 and 3, compared to players in domestic competitions. International defenders and midfielders recorded significantly higher total distances and high-speed running distance compared to players in domestic matches. Our findings suggest that preparing players for international-level competition should include progressive exposure to high-speed running and sprinting distances, as well as high magnitude accelerations. Furthermore, the higher running speeds experienced by players during international matches appears to be a result of less time spent decelerating. The optimal deceleration necessary for specific situations appears important and emphasizes the need for specific deceleration training. The increased effort of high-intensity activity that is required for players competing in international matches affects defenders and midfielders to the greatest degree. Gradual exposure to the increased running demands for midfielders and defenders competing in international matches is needed to improve performance and reduce the potential risk of injury.

Effects of a multi-component alcohol prevention intervention at sporting events: a quasi-experimental control group study.

Alcohol intoxication among spectators and related problems are common at sporting events. This study estimated the impact a multi-component community-based alcohol prevention intervention, implemented at Swedish Premier Football (soccer) League matches, had on intoxication levels among spectators and refusal rates of alcohol service to, and arena entry of, obviously intoxicated spectators. A quasi-experimental controlled study using a repeated cross-sectional design. Spectators and sport arenas in Stockholm and Gothenburg, Sweden. The intervention was implemented in Stockholm and consisted of community mobilisation and collaboration, training of staff and improved enforcement and policy work. Gothenburg was the comparison area and received usual care. Primary outcomes were spectators' mean breath alcohol concentration (BrAC), proportion of spectators with high intoxication levels, (defined as BrAC ≥0.1%), overserving at licensed premises inside arenas and refused arena entry of obviously intoxicated spectators. Baseline data were collected during 2015 and follow-up during 2016 and 2017. BrAC assessments (n=10 188), arena entry attempts (n=201) and alcohol purchase attempts at premises inside arenas (n=495) were collected. There was evidence that the refusal rates of alcohol purchase at premises inside arenas improved differently between study areas and over time (adjusted odds ratio [aOR]=0.28, 95% CI=0.07, 1.06, P=0.060, Bayes factor [BF]=8.60). In both study areas, the arena entry refusal rates improved over time (aOR=5.87, 95% CI=1.16, 29.83, P=0.033, BF=17.7), but evidence that the rates improved differently between study areas and over time was equivocal (aOR=0.57, 95% CI=0.09, 3.56, P=0.543, BF=1.2). Mean BrAC levels decreased over time in both study areas (β=-0.032, 95% CI=-0.050, -0.015, P<0.001), and differently (β=0.104, 95% CI=0.061, 0.146, P<0.001), but evidence for an interaction effect was equivocal (β=-0.002, 95% CI=-0.022, 0.018, P=0.868, BF=1.0). Evidence was equivocal regarding whether the proportion of spectators with high intoxication levels decreased differently between study areas and over time (aOR=1.17, 95% CI=0.91, 1.50, P=0.220, BF=2.2). A multi-component community-based alcohol prevention intervention at sporting events may have increased staff intervention toward obviously intoxicated spectators. It was not clear whether this translated into a reduction in intoxication, which can be explained by improvements in the comparison area.

A Bayesian approach for major European football league match prediction

This paper presents a Bayesian Approach for Major European Football League match prediction. In this study, four variants of Bayesian approaches are investigated to observe the impact of different structural learning algorithms within the family of Bayesian Network which are Naive Bayes (NB), Tree Augmented Naive Bayes (TAN) and two General Bayesian Networks (GBN); K2 algorithm with BDeu scoring function (GBN-K2) and Hill Climbing algorithm with MDL scoring function (GBNHC). The predictive performance of all Bayesian approaches is evaluated and compared based on football match results from five major European Football League consisting of three complete seasons of 1,140 matches. The results showed that GBN-HC gained 92.01% of accuracy while GBN-K2 and TAN produced comparable results with 91.86% and 91.94% accuracy, respectively. The lowest result was produced by NB, with only 72.78% accuracy. The results suggest that TAN requires further exploration in football prediction with its ability to cater the minimal dependency among attributes in a small-sized dataset.

Evaluation of football players' actions in individual risky situations of offensive play as a model determinant of a sports game

The essence of effective training in a sports game is the use of means and forms of teaching in the organisation of a game, which most faithfully reflect the environment of the player's actual activity in natural conditions. Thus, there is a need to accurately identify the conditions of competition in a sports game. One of the types of such activities is observation, which being an elementary method of cognition, based on deliberate, targeted and intended as well as systematic perception of the studied subject, process or phenomenon. Research objective and questions The aim of the work is to assess the level of absolutely dependent cooperation (synergy) during a team sports game in changing game situations, in achieving the highest offensive objective, i.e. scoring a goal. - Is the effectiveness of achieving the highest objective of the game in offensive actions (i.e. scoring a goal) dependent on the level of player synergy? - Is the size of the conflict in the sports objective of the game dependent on creative player involvement? Material and test method - In the work, detailed observation was carried out with regard to players' actions in achieving the highest objective of the game, i.e. scoring a goal. The observation concerned analysis of 134 goals scored in 61 UEFA Champions League and European League football matches in 2009-2020. - In this study, the method of observation was used, with the possibility of repeatedly observing events being the subject of research. - In the research analysis, the use of film material of recorded games on a DVD was taken into account, where the observed effects were applied to an observation sheet. - The situations in which goals were scored and the research material comprised teams that competed in a given match. - The observation of scoring goals took the type of situation and level of interaction between players into account. Results Analysis of the obtained results can be inferred: 1. In achieving the highest objective of offensive actions, a high level of synergy prevails, requiring anticipation of events. 2. The high level of synergistic activities increases with the difficulty of situational tasks. 3. Due to the multi-faceted nature of team games, the player's training process should also affect his/her mental sphere.

Incidence and Characteristics of Immediate Substitutions of Injured Players in Korean Professional Football League Matches

Incidence and Characteristics of Immediate Substitutions of Injured Players in Korean Professional Football League Matches

DEEP MULTILAYER NEURAL NETWORK FOR PREDICTING THE WINNER OF FOOTBALL MATCHES

In this work, we draw attention to prediction of football (soccer) match winner. We propose the deep multilayer neural network based on elastic net regularization that predicts the winner of the English Premier League football matches. Our main interest is to predict the match result (win, loss or draw). In our experimental study, we prove that using open access limited data such as team shots, shots on target, yellow and red cards, etc. the system has a good prediction accuracy and profitability. The proposed approach should be considered as a basis of Oracle engine for predicting the match outcomes.