Sports Analytics Research Articles

Effects of the Use of Sports Analytics and Team Attributes on Success in Regular Season of National Hockey League

Effects of the Use of Sports Analytics and Team Attributes on Success in Regular Season of National Hockey League

Data mining approach for business value analysis in basketball

With the rapidly increasing volume of data, novel methods and technologies for their analysis, and opportunities to support decision-making processes emerge in the domain of sports analytics. This, in particular, applies to analysing athletes’ performance and calculating related business added value in major sports leagues such as the National Basketball Association (NBA). Specifically, the financial success of a team/franchise depends not only on the results of the games but also on the success of attracting marketable individuals who bring higher business value. In that regard, this paper aims to demonstrate the potential and importance of data mining methods to uncover the factors influencing the decisions related to the player selection based on individual results, physical characteristics, and professional contract salaries in the NBA. For the study, 22 datasets were integrated into three large datasets. The data covers the period from 1946 (when the league was founded) to 2017. Data mining models were developed in RapidMiner, enabling correlation, cluster and regression analysis. Change in the factors affecting the selection of new players in recent years was uncovered, while the classification revealed, for example, that more than 50% of players have below-average coefficients of efficiency and individual result contribution. An artificial neural network algorithm was used to identify discrepancies for players with high-salary contracts as many do not meet high-performance standards. The study demonstrates how classification and prediction models can serve sports analysts and managers in making decisions related to future professional contracts and predict future salaries for active players, among other contributions.

The Athletic Intelligence Quotient and Performance in the National Football League

Introduction In the realm of competitive athletics, numerous variables have been examined for predictive utility with respect to player selection/development and outcomes on the field. Notwithstanding important advances, the current predictors only account for a modest amount of variance in outcomes of relevance in the National Football League (NFL). Objective The primary objective of this study was to investigate the predictive validity of a new measure of athletic intelligence, the Athletic Intelligence Quotient (AIQ), which is based on the empirically supported Cattell-Horn-Carroll (CHC) Theory of Intelligence. The predictive validity of the AIQ was determined in relation to performance metrics from 146 NFL players across several seasons. Results Hierarchical regression analyses indicate that specific AIQ factors accounted for a statistically significant increase in the explanation of variance beyond the current level of evaluation for several performance metrics (e.g., career approximate value; sacks, tackles, rushing yards). Further, specific factors of the AIQ are related to position specific statistics, offering the possibility that performance prediction can be focused in for the specific skills required by a given position. Discussion Given the recent impact of analytics in professional sports, and the significant findings noted in the current investigation, the authors discuss the potential importance of the AIQ in the selection and coaching processes.

RETRACTED: Research on the application of embedded wearable devices in campus football

Basic information is that the expression based on ``sports analytics'' is widely used these days. Sports Analytics based pitches have different applications in football that affect the assortment of fields. For example, these may indicate a competitor or group's performance, competitor's ability and market reputation, and a potential physical problem. The ever-increasing number of conferences and pioneers are happy to incorporate such "tools" into their teaching courses to improve their strategies. The show is divided into two important parts. The first is a written audit of existing progress in this regard. The latter part is primarily focused on research directed with football information. In these tests, using appropriate calculations, the player's condition on the field could be traced. By gathering information from past years, can predict a player's goal-scoring performance over the next season. Besides, to be more precise, the player can adjust the shots' size in each match, which corresponds to the objective scoring probability.

Recent Progress in Wearable Biosensors: From Healthcare Monitoring to Sports Analytics.

Recent advances in lab-on-a-chip technology establish solid foundations for wearable biosensors. These newly emerging wearable biosensors are capable of non-invasive, continuous monitoring by miniaturization of electronics and integration with microfluidics. The advent of flexible electronics, biochemical sensors, soft microfluidics, and pain-free microneedles have created new generations of wearable biosensors that explore brand-new avenues to interface with the human epidermis for monitoring physiological status. However, these devices are relatively underexplored for sports monitoring and analytics, which may be largely facilitated by the recent emergence of wearable biosensors characterized by real-time, non-invasive, and non-irritating sensing capacities. Here, we present a systematic review of wearable biosensing technologies with a focus on materials and fabrication strategies, sampling modalities, sensing modalities, as well as key analytes and wearable biosensing platforms for healthcare and sports monitoring with an emphasis on sweat and interstitial fluid biosensing. This review concludes with a summary of unresolved challenges and opportunities for future researchers interested in these technologies. With an in-depth understanding of the state-of-the-art wearable biosensing technologies, wearable biosensors for sports analytics would have a significant impact on the rapidly growing field—microfluidics for biosensing.

Analysing Long Short Term Memory Models for Cricket Match Outcome Prediction

As the technology advances, an ample amount of data is collected in sports with the help of advanced sensors. Sports Analytics is the study of this data to provide a constructive advantage to the team and its players. The game of international cricket is popular all across the globe. Recently, various machine learning techniques have been used to analyse the cricket match data and predict the match outcome as win or lose. Generally these models make use of the overall match level statistics such as teams, venue, average run rate, win margin, etc to predict the match results before the beginning of the match. However, very few works provide insights based on the ball-by-ball level statistics. Here we propose a novel Recurrent Neural Network model which can predict the win probability of a match at regular intervals given the ball-by-ball statistics. The Long Short Term Memory (LSTM) Model takes as input the ball wise features as well as the match level details available from the training dataset. It gives a prediction of winning the match at any time stamp during the match. This level of insight will help the team to predict the probability of them winning the match after every ball and help them determine the critical in-game changes they should make in their game strategies.

Recognizing Events in Spatiotemporal Soccer Data

Spatiotemporal datasets based on player tracking are widely used in sports analytics research. Common research tasks often require the analysis of game events, such as passes, fouls, tackles, and shots on goal. However, spatiotemporal datasets usually do not include event information, which means it has to be reconstructed automatically. We propose a rule-based algorithm for identifying several basic types of events in soccer, including ball possession, successful and unsuccessful passes, and shots on goal. Our aim is to provide a simple procedure that can be used for practical soccer data analysis tasks, and also serve as a baseline model for algorithms based on more advanced approaches. The resulting algorithm is fast, easy to implement, achieves high accuracy on the datasets available to us, and can be used in similar scenarios without modification.

Comparing different feature selection algorithms for cardiovascular disease prediction

Determining the key features for the best model fitting in machine learning is not an easy task. The main objective of this study is to accurately predict cardiovascular disease by comparison among different feature selection algorithms. This study has employed a two-stage feature sub-set retrieving technique to achieve this goal: we first considered three well-established feature selection (filter, wrapper, embedded), and then, a feature sub-set was extracted using a Boolean process-based common “True” condition from these three algorithms. To justify the comparative accuracy and define the best predictive analytics, the well-known random forest, support vector classifier, k-nearest neighbors, Naive Bayes, and XGBoost model have been considered. The artificial neural network (ANN) has been considered as the benchmark for further comparison with all features. The experimental outcomes exhibit that the XGBoost Classifier integrated with the wrapper methods offers precise prediction results for cardiovascular disease. The proposed approach can also be applied in other domains such as sports analytics, bio-informatics, and financial analysis in contrast with healthcare informatics. This empirical study’s novelty is that the common “True” condition–based feature selection and comparison technique is entirely a new phenomenon in medical informatics.

Finger Gesture Tracking for Interactive Applications

This paper presents FinGTrAC, a system that shows the feasibility of fine grained finger gesture tracking using low intrusive wearable sensor platform (smart-ring worn on the index finger and a smart-watch worn on the wrist). The key contribution is in scaling up gesture recognition to hundreds of gestures while using only a sparse wearable sensor set where prior works have been able to only detect tens of hand gestures. Such sparse sensors are convenient to wear but cannot track all fingers and hence provide under-constrained information. However application specific context can fill the gap in sparse sensing and improve the accuracy of gesture classification. Rich context exists in a number of applications such as user-interfaces, sports analytics, medical rehabilitation, sign language translation etc. This paper shows the feasibility of exploiting such context in an application of American Sign Language (ASL) translation. Noisy sensor data, variations in gesture performance across users and the inability to capture data from all fingers introduce non-trivial challenges. FinGTrAC exploits a number of opportunities in data preprocessing, filtering, pattern matching, context of an ASL sentence to systematically fuse the available sensory information into a Bayesian filtering framework. Culminating into the design of a Hidden Markov Model, a Viterbi decoding scheme is designed to detect finger gestures and the corresponding ASL sentences in real time. Extensive evaluation on 10 users shows a recognition accuracy of 94.2% for 100 most frequently used ASL finger gestures over different sentences. When the size of the dictionary is extended to 200 words, the accuracy is degrades gracefully to 90% thus indicating the robustness and scalability of the multi-stage optimization framework.

Trends in sports analytics for event management: A content analysis

This is a conceptual paper which explores recent trends in sports analytics for event management and the influence thereof on viewership, sponsorship and athletic performance. The content analysis here entails multiple case studies on sporting events; recommended practices for organising sporting events are highlighted using a site-ordered effects matrix. Analysis yields unprecedented outcomes related to sporting event ticket prices (relevant especially to fans and organisers), willingness of athletes to participate in mega events, technologies to be used to effectively reach out to passionate fans, and innovative practices to build a sound infrastructure involving sponsors, athletes and authorities.

Basketball lineup performance prediction using edge-centric multi-view network analysis

Sports analytics is one of the fast-growing applications of machine learning and data mining. The major goal in sport analytics applications is to make informed decisions to gain a competitive advantage. In this paper, we propose a method to analyse and predict lineup performance in basketball which may assist the team managers and coaches to make informed decisions. The constant lineup change in basketball games makes it vital to always have the best move ready at every moment of the game. A complex decision-making system is very important for the coaches to navigate their team through a season. This system must make the decisions based on the pros and cons of the teams, past results and lineups, and both teams’ performance under similar situations. To create this system, authors have constructed a signed, directed, and weighted network from the results of all matchups and lineups of the teams in NBA from 2007 to 2019. The proposed approach uses machine learning and graph theory to develop a new metric called Inverse Square Metric and edge-centric multi-view networks to predict the performance of a lineup in a given situation. The edge-centric approach provides a deep analysis of any condition between two teams from 16 different perspectives. The average accuracy achieved by ISM and edge-centric multi-view is 68% and 80%, respectively. In the end, the training and test datasets are chosen from different seasons and results are reasonable compared to the baseline methods.

A Bayesian adjusted plus-minus analysis for the esport Dota 2

Abstract Analytics and professional sports have become linked over the past several years, but little attention has been paid to the growing field of esports within the sports analytics community. We seek to apply an Adjusted Plus Minus (APM) model, an accepted analytic approach used in traditional sports like hockey and basketball, to one particular esports game: Defense of the Ancients 2 (Dota 2). As with traditional sports, we show how APM metrics developed with Bayesian hierarchical regression can be used to quantify individual player contributions to their teams and, ultimately, use this player-level information to predict game outcomes. In particular, we first provide evidence that gold can be used as a continuous proxy for wins to evaluate a team’s performance, and then use a Bayesian APM model to estimate how players contribute to their team’s gold differential. We demonstrate that this APM model outperforms models based on common team-level statistics (often referred to as “box score statistics”). Beyond the specifics of our modeling approach, this paper serves as an example of the potential utility of applying analytical methodologies from traditional sports analytics to esports.

Sports scheduling and other topics in sports analytics: a survey with special reference to Latin America

Recent years have witnessed vigorous development in the application of mathematical and computational techniques to many aspects of the organization and planning of sports competitions. Known as sports scheduling, this field is one of the subjects of the present survey, which reviews different problems tackled in the associated literature, the techniques employed for solving them and the results of their implementation in real-world cases. Other topics in sports analytics such as player performance, result prediction, fantasy games and analysis of rankings are also examined. Special attention is given to applications in Latin America. The main challenges facing sports scheduling and other areas of sports analytics are also discussed.

Strategic target setting in the heptathlon

Article published in Journal of Sports Analytics on 28 February 2020, available open access at: https://doi.org/10.3233/JSA-200351.

Sports analytics — Evaluation of basketball players and team performance

Given the recent trend in Data Science (DS) and Sports Analytics, an opportunity has arisen for utilizing Machine Learning (ML) and Data Mining (DM) techniques in sports. This paper reviews background and advanced basketball metrics used in National Basketball Association (NBA) and Euroleague games. The purpose of this paper is to benchmark existing performance analytics used in the literature for evaluating teams and players. Basketball is a sport that requires full set enumeration of parameters in order to understand the game in depth and analyze the strategy and decisions by minimizing unpredictability. This research provides valuable information for team and player performance basketball analytics to be used for better understanding of the game. Furthermore, these analytics can be used for team composition, athlete career improvement and assessing how this could be materialized for future predictions. Hence, critical analysis of these metrics are valuable tools for domain experts and decision makers to understand the strengths and weaknesses in the game, to better evaluate opponent teams, to see how to optimize performance indicators, to use them for team and player forecasting and finally to make better choices for team composition.

Exploring the Importance of Players’ Characteristics and Performance on Serve and Return of Serve in Winning the Women’s Singles Grand Slam Tennis Tournaments

Sports analytics has become very popular in recent years. Many researchers have been using the tools and technique to study tennis player performance in various aspects during matches. However, much of the studies were performed on male players while much less attention were made on female players. This study explores the female players’ characteristics and serve performance that leads to winning matches in women’s singles 2018 Grand Slam tennis tournaments. Using data provided by the official website of the Women Tennis Association (WTA), 1016 single matches in all four 2018 Grand Slam Tournaments were analysed to evaluate winning traits in tennis specifically looking at the two most important strokes in tennis which are the serve and return of serve. Additionally, other personal player characteristics that would influence the match outcomes were also evaluated. To achieve the objectives of the study, odds ratio and chi-square test were used to compare between the match winner and match loser. The study revealed that percentage of first serve return won (PFSRW) has the highest odds ratio value (OR=1.263) followed by the percentage of first serve won (PFSW) with odds ratio equal to 1.244. These showed that having a good first serve, plays an important factor in increasing the winning odds as compared to second serve either in serve or in return of serve. This study also reveals that a taller player had the advantage over a shorter player for faster serve speed (p-value<0.05).

On-Line Visual Tracking with Occlusion Handling.

One of the core challenges in visual multi-target tracking is occlusion. This is especially important in applications such as video surveillance and sports analytics. While offline batch processing algorithms can utilise future measurements to handle occlusion effectively, online algorithms have to rely on current and past measurements only. As such, it is markedly more challenging to handle occlusion in online applications. To address this problem, we propagate information over time in a way that it generates a sense of déjà vu when similar visual and motion features are observed. To achieve this, we extend the Generalized Labeled Multi-Bernoulli (GLMB) filter, originally designed for tracking point-sized targets, to be used in visual multi-target tracking. The proposed algorithm includes a novel false alarm detection/removal and label recovery methods capable of reliably recovering tracks that are even lost for a substantial period of time. We compare the performance of the proposed method with the state-of-the-art methods in challenging datasets using standard visual tracking metrics. Our comparisons show that the proposed method performs favourably compared to the state-of-the-art methods, particularly in terms of ID switches and fragmentation metrics which signifies occlusion.

Sport Analytics: Science or Alchemy?

Sport analytics promises to use Big Data and sophisticated statistical methods to identify effective strategies in sports—“the Moneyball moment.” However, much like alchemy, sport analytics is characterized by opacity and secrecy, and outside of baseball, evidence of success that would meet the usual scientific criteria is limited. An example is used to demonstrate that quite simple models can match more complex ones in terms of prediction. Like alchemy, sport analytics can deliver important advances in our understanding, but some problems need to be addressed. These include the need to incorporate theory, reconciling the pursuit of profit with scientific principles, and focusing on prediction as a measure of progress.

University laboratory of biomechanics and sport analytics and engineering in Gdansk as an important science institution

University laboratory of biomechanics and sport analytics and engineering in Gdansk as an important science institution

A family of parsimonious mixtures of multivariate Poisson‐lognormal distributions for clustering multivariate count data

Multivariate count data are commonly encountered through high‐throughput sequencing technologies in bioinformatics, text mining, or sports analytics. Although the Poisson distribution seems a natural fit to these count data, its multivariate extension is computationally expensive. In most cases, mutual independence among the variables is assumed; however, this fails to take into account the correlation among the variables usually observed in the data. Recently, mixtures of multivariate Poisson‐lognormal (MPLN) models have been used to analyze such multivariate count measurements with a dependence structure. In the MPLN model, each count is modeled using an independent Poisson distribution conditional on a latent multivariate Gaussian variable. Owing to this hierarchical structure, the MPLN model can account for over‐dispersion as opposed to the traditional Poisson distribution and allows for correlation between the variables. Rather than relying on a Monte Carlo‐based estimation framework, which is computationally inefficient, a fast variational expectation–maximization (EM)‐based framework is used here for parameter estimation. Further, a family of parsimonious mixtures of Poisson‐lognormal distributions is proposed by decomposing the covariance matrix and imposing constraints on these decompositions. Utility of such models is shown using simulated and benchmark datasets.