Business analytics in sport talent acquisition

DOI: 10.4018/IJBAN.290406 International Journal of Business Analytics Volume 9 • Issue 1 This article published as an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/) which permits unrestricted use, distribution, and production in any medium, provided the author of the original work and original publication source are properly credited. *Corresponding Author 1 Business Analytics in Sport Talent Acquisition: Methods, Experiences, and Open Research Opportunities Rocio de la Torre, Public University of Navarre, Spain Laura O. Calvet, Universitat Oberta de Catalunya, Spain David Lopez-Lopez, ESADE, Spain Angel A. Juan, Universitat Oberta de Catalunya, Spain https://orcid.org/0000-0003-1392-1776 Sara Hatami, Universitat Oberta de Catalunya, Spain ABSTRACT Recruitment of young talented players is a critical activity for most professional teams in different sports such as football, soccer, basketball, baseball, cycling, etc. In the past, the selection of the most promising players was done just by relying on the experts’ opinions but without systematic data support. Nowadays, the existence of large amounts of data and powerful analytical tools have raised the interest in making informed decisions based on data analysis and data-driven methods. Hence, most professional clubs are integrating data scientists to support managers with data-intensive methods and techniques that can identify the best candidates and predict their future evolution. This paper reviews existing work on the use of data analytics, artificial intelligence, and machine learning methods in talent acquisition. A numerical case study, based on real-life data, is also included to illustrate some of the potential applications of business analytics in sport talent acquisition. In addition, research trends, challenges, and open lines are also identified and discussed. KEywORdS Business Analytics, Machine Learning, Sports, Talent Acquisition 1. INTROdUCTION In the present day, finding and hiring talented workers has become one of the top priorities for many businesses. Inefficient hiring practices have a negative repercussion on any organization, and might impose considerable loses, both in terms of money and time. As pointed out by Davenport et al. (2010), those companies that are capable of attracting and retaining the best talented people are among the most competitive ones. According to Harris et al. (2011), in a globalized and highly competitive environment most organizations should start using data to measure and improve the contribution of their human resources (HR) to their performance. In the sports sector, Baker et al. (2017), De Bosscher and De Rycke (2017) and Hanlon et al. (2014) state that there is an increasing interest in understanding the costs and benefits of initiatives International Journal of Business Analytics Volume 9 • Issue 1 2 for early identification of talented players, as well as in unleashing the factors that influence athletes’ development. These authors also affirm that, while traditional statistical analysis was focused on match variables, such us goals scored o players’ position on the field, recent advances in sports analytics are focused on more complex issues like talent acquisition. As pointed out by Gerrard (2017), the 2011 film ‘Moneyball’1 highlighted the possibilities of analytics as a competitive strategy, particularly for small-market teams with relatively limited resources. Following Fried and Mumcu (2016), many coaches employ data on habits and performance indicators to assess the potential of their players. In fact, it is possible to use data to: (i) evaluate players’ performance; (ii) rank players (Pappalardo et al., 2019); (iii) estimate the value of players in transfer markets (Kim et al., 2019); (iv) locate the best position that a player can occupy in the field; or (v) forecast players’ goal scoring performance in the next season (Apostolou and Tjortjis, 2019). As illustrated in Figure 1, adapted from 21st Club2, clubs can use data to analyze the impact of a new player on the team’s overall performance level. The sports industry is being transformed by data analytics in the following dimensions: (i) at clubs level, e.g., soccer clubs like Liverpool, Barcelona3, Arsenal, Manchester City, or Milan are among the ones that already use data analysis to improve performance, analyze rivals, prevent injuries, optimize the management of the transfer market, and also the acquisition of new talent; (ii) regarding new entrants in data management and analysis, new platforms for data analysis and management appear to provide services to clubs, such as Wiscout4 and Scisports5; (iii) as regards as new entrants in data capture and generation, large companies such as Intel have launched the creation of wearable Internet-of-Things devices, which are capable of capturing players’ information in real time6; (iv) with respect to transformation of sports management professionals, new official university degrees dedicated to sports management have emerged, most of them with a special emphasis on data science; and (v) a transformation of sports enthusiasts, who have begun to consume complex data, both for their own digitization process as well as for their leisure and fun when using online betting applications. Figure 1. Player recommendation based on data analytics (adapted from 21st Club) International Journal of Business Analytics Volume 9 • Issue 1 3 This paper presents a comprehensive review of the state of the art regarding data-driven approaches for talent acquisition in sports. In addition, the paper discusses the most common used methods for talent acquisition. Some of these methods belong to the fields of artificial intelligence (AI) and machine learning (ML), which are also introduced in the context of sports analytics. Finally, the paper identifies and discusses trends, challenges, and open research lines related to this research area. The rest of the paper is organized as follows. Section 2 bring out the concept of talent analytics, specially in the sports sector. Section 3 provides a short introduction to the fields of AI and ML, thus allowing the unfamiliar reader to follow the following sections. Section 4 discusses different datadriven analytic approaches. Afterwards, Sections 5 and 6 review research papers on data analytics in recruitment and sports talent acquisition, respectively. Section 7 provides an original case study based on a real-life dataset of European soccer players, where some of the potential of ML methods is illustrated. Trends, challenges, and open research lines are discussed in Section 8. Finally, Section 9 summarizes the main contributions of this paper. 2. TALENT ANALyTICS IN SPORTS Talent analytics (TA) represents a groundbreaking opportunity for many organizations in the sports sector. TA is defined by Bassi (2011) as “an evidence-based approach for making better decisions on the people side of the business; it consists of an array of tools and technologies, ranging from simple reporting of HR metrics all the way up to predictive modeling”. During the last years, professional and elite sport teams are giving a considerable attention to implement business-oriented TA in their strategy. Hence, for Davids and Arau´jo (2019) the main challenge is not the management of talent among the players who already belong to a team, but the acquisition of young and talented players for the future. When considering young people, evolution and forecasting models that go beyond data on the current performance level should be built as well (Webb et al., 2020; Pifer, 2019; Ford and Williams, 2017). As Williams and Reilly (2000) state, multi-dimensional data (including physical, psychical, and sociological characteristics) has to be considered. Likewise, Davcheva (2014) maintains that data obtained from social networks could be used as a predictor of potential children’s talent, while Martin (2015) explores the influence of the child socio-economic status on its performance and future evolution. Other authors (Kandra´ˇc et al., 2019; Pickering et al., 2019; Loland, 2015; Webborn et al., 2015; Cˆot´e, 1999) discuss about whether talent is an innate skill or not, and if data analysis should be oriented towards innate capacities and genes rather than to performance. Given the number of factors that may affect decisions related to talent acquisition, authors such as Vaeyens et al. (2008), Gerrard (2017), and Bergkamp et al. (2019) conclude that talent acquisition is a multi-dimensional challenge, and one that is not only based on the skills of individual players but also on the whole team, i.e., the current team configuration has to be considered as well. In this context, Gerrard (2017) propose the use of simulation models as more effective tools than expert judgment, especially in a multi-dimensional environment like the one being considered. Using data analytics, Gandelman (2009) was able to show that outside opportunities were higher for soccer players with a superior socioeconomic background and a better education. He also found evidence of racial discrimination in the Uruguayan soccer market, where obtaining a professional contract was easier to white players. Similarly, Berri and Brook (2010) used data analytics to identify some efficiencies regarding the evaluation of players’ performance in the National Hockey League. Employing Bayesian analysis combined with Markov Chain Monte Carlo estimation, Rimler et al. (2010) studied the technical efficiency of a basketball team. They were not able to find significant differences, in technical efficiency levels, across teams playing in the same category. Bryson et al. (2013) investigated how the salary of a soccer player might depend upon his ability to play with both feet. After analyzing data from the five main European leagues, they concluded that the aforementioned players tend to receive a noticeable salary premium. International Journal of Business Analytics Volume 9 • Issue 1 4 3. ARTIFICIAL INTELLIGENCE ANd MACHINE LEARNING The popularity of AI and ML methods has been constantly increasing during the last decade (Joshi, 2019). The disciplines in which they have been originated are numerous, including: Computer Science, Mathematics, Electrical Engineering, Statistics, Signal Processing, etc. These fields find applications in a wide range of industries, such as image processing, natural language processing, and online shopping. Figures 2 and 3 shows the variety of AI-related and ML-related research disciplines, and the impact generated measured in terms of the number of indexed publications in the Web of Science (WoS). Figure 2. WoS-indexed contributions in Artificial Intelligence according to its research area Figure 3. WoS-indexed contributions in Machine Learning according to its research area International Journal of Business Analytics Volume 9 • Issue 1 5 According to Broussard et al. (2019), AI refers to machines capable of performing one or more tasks associated with the human nature, for instance: learn and process human language, execute mechanical tasks that require complex maneuvering, solving computer-based complex problems that may involve large amount data in a very short time lapse. Others, like Li and Du (2017), define the term as “variety of human intelligent behaviors, such as perception, memory, emotion, judgment, or reasoning, that can be realized artificially by a machine, a system, or a network”. Most current applications are focused on the field of neural networks. For instance, deep neural networks are used for speech recognition (Tveter, 1997), image classification (Deepa and Devi, 2011), or prediction of words in a text (Battaglia et al., 2016). ML as subset of AI refers to a computer program that can learn how to produce a certain behavior, which was not explicitly programmed in it (Kotsiantis et al., 2007). Indeed, it can be capable of showing behaviors from which the programmer may be completely unaware of. As in human behavior, many aspects of learning and intelligence in AI are closely related to the representation of uncertainty. Therefore, probabilistic approaches are fundamental (Ghahramani, 2015). Probabilistic methods try to assign an uncertainty measure to the unknown variables, as well as a certain probability to known variables. Hence, the goal is to find the unknown values using probabilistic models. These models are classified into two main types, so called generative and discriminative: discriminative models try to forecast the changes in the output just considering the changes occurred in the input, while generative models are the ones in which the changes in the output can be explained as a consequence of changes in the input as well as changes in the state (Joshi, 2019). Current research regarding the frontier of probabilistic ML approaches (both discriminative as well as generative) is mainly focused on: (i) probabilistic programming (as a general framework for expressing probabilistic models as computer programs); (ii) Bayesian optimization (for globally optimizing unknown functions); (iii) hierarchical modeling for learning many related models; and (iv) probabilistic data compression (Ghahramani, 2015). ML approaches can be divided into supervised and unsupervised learning. The former are involved in many applications and deal with problems related to learning with guidance. In other words, the training data in supervised learning methods needs labeled samples. Thus, for instance, samples with class labels are required in a classification problem. Hence, the mathematical model learns its parameters from labeled samples with the main goal of making predictions on samples that the model has not seen before. Then, the classifier is used for assigning class labels to the testing instances in which the values of the predictor features are known, but the value of the class label remains unknown. Since the supervised classification is one of the tasks frequently developed by ‘intelligent systems’, it seems logical that a great number of techniques are based on AI and statistics. Meanwhile, unsupervised learning deals with problems that involve data without labels. In this case, the machine receives inputs but obtains neither outputs nor rewards from its environment (Ghahramani, 2003). Unsupervised approaches try to find trends and some kind of structure in the training data. That is, these approaches try to understand the origin of the data itself and to build representations of the inputs that can be used for decision-making, efficiently communicating the inputs to another machine or predicting future inputs. Clustering is a typical example of unsupervised learning. In unsupervised learning, the majority of the work can be considered as a learning process of a probabilistic model. When the scenario is not able to give the machine any supervision or reward, the machine can design a model that represents the probability distribution for a new input. This is achieved by just considering a previous useful input (e.g., stock prices or weather conditions). Probabilistic models that can be used in unsupervised learning are, among others: factor analysis, independent components analysis, principal components analysis, or Gaussians models. There are situations where the supervised methods are not the best option. The first and most important is the high cost of labeling. Moreover, having all the training data fully labeled can be practically impossible. In these cases, it is common to start with supervised methods –using a small International Journal of Business Analytics Volume 9 • Issue 1 6 set of labeled data–, and then improve the model in an unsupervised way–i.e., using a larger set of unlabeled data. AI and ML techniques are present in almost all sectors, including sports. In fact, most variables that can be quantified can also be predicted using AI and ML. The sport sector is full of quantifiable elements, which makes it ideal for the use of these techniques. For example, recruitment of players is one of the areas in sports where AI and ML are increasingly employed (Chavan, 2019; Herold et al., 2019; Musa et al., 2019; C´wiklinski et al., 2021). 4. TyPES OF ANALySIS IN TALENT ANALyTICS Data analytic approaches, which are typically based on ML and statistics methods, can bring insights that are critical for improving operational and business outcomes of many organizations. These approaches play a role as powerful tools in the seeking and hiring young talent. Talent analytics is a systematic process that applies statistics, technology, and expertise to large sets of people data to discover the meaningful patterns that allow for supporting decision-making in recruitment. Three common types of analytics –descriptive, predictive, and prescriptive– are used in TA, people and human resource analytics frameworks to measure efficiency, effectiveness, quality of recruitment, and impact (Necula and Strimbei, 2019). From descriptive to prescriptive analysis, not only the model increases in the complexity of the data being used, but also the analysis progressively gets more sophisticated. Each of the aforementioned types are described next: Descriptive Analytics: reporting / visualization is the first step for carrying out statistical analyses, and it is used to describe the basic features of the data via applying to the collected data through the structured questionnaire (Marrybeth et al., 2019); it plays a relevant role in providing a view into activity –such as requisition volume, talent pool size, source of hires, etc.–, as well as to reveal the levels of activity and efficiency in each candidate generation. Predictive Analytics: uses data to find patterns and employs them to predict the future; it permits to identify statistical relationships between a set of activities and the expected outcomes, that will help to: (i) forecast what will happen in the future or to explain the obtained outcome (like a candidate’s likely cultural fit, level of performance, and retention); or (ii) notice potential talent shortages or skills gaps, and market availability (workforce planning); also, predictive analytics is mainly related to selection or rejection of candidate, acceptance of provided offer by selected candidates and root cause analysis for offer decline (Srivastava et al., 2015); predictive analysis finds answers for questions such as ‘what will the future look like?’, ‘what tactics most influence business outcomes?’, etc.; the result of this predictive analysis help shorten the entire recruiting process while improving the hiring process. Prescriptive Analytics: goes a step further in the future and attempts to provide and suggest better decisions using data techniques such as decision modeling, ML, heuristic, simulation, neural networks; these decisions are based on the results provided by predictive analysis; it tries to evaluate the effect and impact of the provided decisions in order to modify them before implementation; it usually results in rules and recommendation for next steps (Attaran and Attaran, 2018); for example talent acquisition team receives the hire or not hire suggestions or strategy recommendation from predictive analysis. A common data analytics framework used in TA is shown in Figure 4. The most common data-driven methodologies used in TA may classified in International Journal of Business Analytics Volume 9 • Issue 1 7 regression, classification, clustering, association rule mining, and anomaly detection. The reader interested in a more complete introduction to TA is referred to Davenport et al. (2010), which illustrates uses and describes the fundamentals to build a capacity in this domain, i.e.: access to high-quality data, enterprise orientation, analytical leadership, and strategic targets. In this context, Nocker and Sena (2019) discusses the advantages and costs induced (in terms of data governance and ethics) by using TA within an organization. The authors present a number of case studies to analyze the positive effect of TA usage in organizational decision-making processes and determine the key channels through which the TA adoption improve HR management and, subsequently, the whole organization function. 5. dATA ANALyTICS IN RECRUITMENT A more data-driven culture is becoming increasingly popular among companies and governments. HR constitutes an example of a business’ department that has dramatically changed during the last decades due to the use of data analytics methodologies and technologies. Indeed, companies are increasingly adopting sophisticated methods for study employee’s data in order to improve the decision-making process, so they can strength their competitive advantage (Davenport et al., 2010). According to Rana et al. (2019), TA shows the potential within the decisions regarding hiring, training, improving productivity, and retaining talent, all of them with the main purpose to make a company more competitive. Gartner, Inc.7, points out that the volume of data and metrics available for HR has increased exponentially, while 70% of companies expect to increase the resources they dedicate to TA in the coming years. Even so, only 21% of the HR leaders believe that their organizations are effective at using talent data to inform business decisions. Dey and De (2015) points out five key areas where predictive analytics can create value in HR: (i) employee profiling and segmentation, employee attrition, and loyalty analysis; (ii) forecasting of HR capacity and recruitment needs; (iii) appropriate recruitment profile selection; (iv) employee sentiment analysis; and (v) employee fraud risk management. Figure Figure 4. Data analytics framework using in Talent Analytics International Journal of Business Analytics Volume 9 • Issue 1 8 5 lists the main TA applications, methodologies, and technologies. It is based on Kaur and Fink (2017), which offers a review of key approaches, competencies and tools, building on 22 interviews with academics, consultants and practitioners at 16 corporations, as well as on other TA experts. 5.1. Recruitment and Talent Acquisition According to Wikipedia, recruitment may be defined as “the process of attracting, shortlisting, selecting, and appointing suitable candidates for jobs within an organization, and is a key function of human resource management.” Another interesting definition is provided by Breaugh (2008), which defines external recruitment as “an employer’s actions that are intended to: (i) bring a job opening to the attention of potential job candidates who do not currently work for the organization; (ii) influence whether these individuals apply for the opening; (iii) affect whether they maintain interest in the position until a job offer is extended; and (iv) influence whether a job offer is accepted”. Recruitment plays an essential role in determining the effectiveness of organizations, and it is composed of several sub-processes: (i) job analysis, which consists in documenting the knowledge, skills, abilities, and other characteristics (KSAOs) required for a job; (ii) sourcing, which is the process of attracting or identifying candidates; and finally (iii) screening and selection. Organizations apply recruitment strategies to identify hiring vacancy, establish timelines, and define goals throughout the recruitment process. Each organization designs its own strategies for recruitment, but there are frequent approaches such as using social networks for external recruitment advertisement or employing standard psychological tests, group discussion and a number of interviews to assess a variety of KSAOs. Thus, the recruitment process is complex and requires big amounts of effort and investment. Bhattacharyya (2015) defines talent acquisition as a strategic approach aiming to identify, attract, and bring onboard top talent to meet dynamic business needs. According to this author, recruiting is more tactical and focuses mostly on immediate hiring needs, i.e., a process of filling the open positions. Figure 6 allows us to check the growing popularity of this research field. More specifically, it shows the evolution, from 2000 to 2019, in the number of works reported by the Web of Science when searching for: (i) “Talent acquisition” or “Talent recruitment” (which sometimes are used as synonyms in the literature) in “Social Science” (in orange); and (ii) the same but in “sports” (in blue). Clearly, there has been a positive and sustained trend during the last 10 to 15 years. 5.2. Studies on data-driven Recruitment Here, we describe a few recent and representative works on recruitment using data-intensive methodologies and techniques. For instance, Mohapatra and Sahu (2017) presents a case study to highlight the misconceptions in capacity planning methods (i.e. hiring process) and in the detection of bottlenecks in the hiring pipeline by using recruitment funnel technique and some metrics to check efficiency of the aforementioned hiring process. Moreover, the authors depict appropriate sources of hiring based on the performance of candidates hired from those sources. They identify successful profiles in the company through computing correlations between selection parameters and performance scores. Finally, they provide a recruitment strategy and future road map for the case study. Kaur and Fink (2017) investigates what is required to set up and run an effective TA function and the structures, system and skills that enable it. The authors discuss the answer to these questions through the collection and analysis of data from 22 interviews with academics, consultants, and different industries. The analysis shows that data infrastructure and reporting, advanced analytics, and organizational research are three components of a mature TA function. Azar et al. (2013) provides a decision-making tool to help managers during the recruitment process. Authors claim that the tool, through using data mining techniques, is able to discover patterns International Journal of Business Analytics Volume 9 • Issue 1 9 Figure 5. Scheme of talent analytics. Source: based on Kaur and Fink (2017) Figure 6. Evolution of the number of related works from 2000. Data source: Web of science. Dashed lines represent the tendencies calculated as moving averages. International Journal of Business Analytics Volume 9 • Issue 1 16 REFERENCES Alamar, B. C. (2013). Sports analytics: A guide for coaches, managers, and other decision makers. Columbia University Press. Apostolou, K., & Tjortjis, C. (2019). Sports analytics algorithms for performance prediction. In 10th International Conference on Information, Intelligence, Systems and Applications (IISA). IEEE. Attaran, M., & Attaran, S. (2018). Opportunities and challenges of implementing predictive analytics for competitive advantage. International Journal of Business Intelligence Research, 9. Azar, A., Rajaeian, A., Sebt, M. V., & Ahmadi, P. (2013). A model for personnel selection with a data mining approach: A case study in a commercial bank. SA Journal of Human Resource Management, 11(1), 1–10. doi:10.4102/sajhrm.v11i1.449 Babbitt, D. G. (2019). A development model to guide the recruiting of female shot putters at the NCAA Division I Championship level. Journal of Sports Analytics, 5(3), 181–190. doi:10.3233/JSA-180275 Baboota, R., & Kaur, H. (2019). Predictive analysis and modelling football results using machine learning approach for English premier league. International Journal of Forecasting, 35(2), 741–755. doi:10.1016/j. ijforecast.2018.01.003 Baker, J., Cobley, S., Schorer, J., & Wattie, N. (2017). Talent identification and development in sport: An introduction. In Routledge Handbook of Talent Identification and Development in Sport (pp. 1–8). Routledge. doi:10.4324/9781315668017-1 Baker, R. E., & Kwartler, T. (2015). Sport analytics: Using open source logistic regression software to classify upcoming play type in the NFL. Journal of Applied Sport Management, 7. Barron, D., Ball, G., Robins, M., & Sunderland, C. (2020). Identifying playing talent in professional football using artificial neural networks. Journal of Sports Sciences, 38(11-12), 1–10. doi:10.1080/02640414.2019.17 08036 PMID:31941425 Bassi, L. (2011). Raging debates in HR analytics. People and Strategy, 34, 14. Battaglia, P., Pascanu, R., Lai, M., & Rezende, D. J. (2016). Interaction networks for learning about objects, relations and physics. Advances in Neural Information Processing Systems, 4502–4510. Beckwith, G., Callahan, T., Carlson, B., Fondren, T., Harris, R., Hoege, J., Martin, T., Menna, C., Summer, E., & Scherer, W. (2019). Systems analysis for university of virginia football recruiting and performance. In 2019 Systems and Information Engineering Design Symposium (SIEDS). IEEE. doi:10.1109/SIEDS.2019.8735611 Bergkamp, T. L., Niessen, A. S. M., den Hartigh, R. J., Frencken, W. G., & Meijer, R. R. (2019). Methodological issues in soccer talent identification research. Sports Medicine (Auckland, N.Z.), 49(9), 1–19. doi:10.1007/ s40279-019-01113-w PMID:31161402 Berri, D. J., & Brook, S. L. (2010). On the evaluation of the “most important” position in professional sports. Journal of Sports Economics, 11(2), 157–171. doi:10.1177/1527002510363097 Bhandari, I., Colet, E., Parker, J., Pines, Z., Pratap, R., & Ramanujam, K. (1997). Advanced scout: Data mining and knowledge discovery in NBA data. Data Mining and Knowledge Discovery, 1(1), 121–125. doi:10.1023/A:1009782106822 Bhattacharyya, D. K. (2015). The magnetic organization: Attracting and retaining the best talent. SAGE Publications India. Breaugh, J. A. (2008). Employee recruitment: Current knowledge and important areas for future research. Human Resource Management Review, 18(3), 103–118. doi:10.1016/j.hrmr.2008.07.003 Broussard, M., Diakopoulos, N., Guzman, A. L., Abebe, R., Dupagne, M., & Chuan, C. H. (2019). Artificial intelligence and journalism. Journalism & Mass Communication Quarterly, 96(3), 673–695. doi:10.1177/1077699019859901 International Journal of Business Analytics Volume 9 • Issue 1 17 Brown, M. R., Delau, S., & Desgorces, F. D. (2010). Effort regulation in rowing races depends on performance level and exercise mode. Journal of Science and Medicine in Sport, 13(6), 613–617. doi:10.1016/j. jsams.2010.01.002 PMID:20227342 Bryson, A., Frick, B., & Simmons, R. (2013). The returns to scarce talent: Footedness and player remuneration in european soccer. Journal of Sports Economics, 14(6), 606–628. doi:10.1177/1527002511435118 Chavan, A. (2019). Recruitment of Suitable Football Player by using Machine Learning Techniques (Ph.D. thesis). National College of Ireland. Cokins, G., & Schrader, D. (2017). The sports analytics explosion. OR/MS Today, 44, 28–33. C^ot’e, J. (1999). The influence of the family in the development of talent in sport. The Sport Psychologist, 13(4), 395–417. doi:10.1123/tsp.13.4.395 C´wiklinski, B., Gie-lczyk, A., & Chora’s, M. (2021). Who will score? a machine learning approach to supporting football team building and transfers. Entropy (Basel, Switzerland), 23(1), 90. doi:10.3390/e23010090 PMID:33435241 Davcheva, P. (2014). Identifying sports talents by social media mining as a marketing instrument. In Annual SRII global conference. IEEE. doi:10.1109/SRII.2014.38 Davenport, T. H., Harris, J., & Shapiro, J. (2010). Competing on talent analytics. Harvard Business Review, 88, 52–58. PMID:20929194 Davids, K., & Arau’jo, D. (2019). Innate talent in sport: Beware of an organismic asymmetry–comment on Baker & Wattie. In Current Issues in Sport Science. CISS. De Bosscher, V., & De Rycke, J. (2017). Talent development programmes: A retrospective analysis of the age and support services for talented athletes in 15 nations. European Sport Management Quarterly, 17(5), 590–609. doi:10.1080/16184742.2017.1324503 Deepa, S., & Devi, B. A. (2011). A survey on artificial intelligence approaches for medical image classification. Indian Journal of Science and Technology, 4(11), 1583–1595. doi:10.17485/ijst/2011/v4i11.35 Dey, T., & De, P. (2015). Predictive analytics in HR: A primer. A White Paper, Tata Consultancy Services. Available: http://www.tcs.com/SiteCollectionDocuments/White-Papers/PredictiveAnalytics-HR-0115-1.pdf Dronyk-Trosper, T., & Stitzel, B. (2017). Lock-in and team effects: Recruiting and success in college football athletics. Journal of Sports Economics, 18(4), 376–387. doi:10.1177/1527002515577885 Edelmann-Nusser, J., Hohmann, A., & Henneberg, B. (2002). Modeling and prediction of competitive performance in swimming upon neural networks. European Journal of Sport Science, 2(2), 1–10. doi:10.1080/17461390200072201 Ford, P. R., & Williams, A. M. (2017). Sport activity in childhood: Early specialization and diversification. In Routledge Handbook of Talent Identification and Development in Sport (pp. 116–132). Routledge. doi:10.4324/9781315668017-9 Fried, G., & Mumcu, C. (2016). Sport analytics: A data-driven approach to sport business and management. Taylor & Francis. doi:10.4324/9781315619088 Gandelman, N. (2009). Selection biases in sports markets. Journal of Sports Economics, 10(5), 502–521. doi:10.1177/1527002509332237 Gaviao, L. O., Sant’Anna, A. P., Alves Lima, G. B., & de Almada Garcia, P. A. (2020). Evaluation of soccer players under the moneyball concept. Journal of Sports Sciences, 38(11-12), 1221–1247. doi:10.1080/026404 14.2019.1702280 PMID:31876264 Gerrard, B. (2017). The role of analytics in assessing playing talent. In Routledge Handbook of Talent Identification and Development in Sport. Routledge. doi:10.4324/9781315668017-30 Ghahramani, Z. (2003). Unsupervised learning. In Summer School on Machine Learning (pp. 72–112). Springer. Ghahramani, Z. (2015). Probabilistic machine learning and artificial intelligence. Nature, 521(7553), 452–459. doi:10.1038/nature14541 PMID:26017444 International Journal of Business Analytics Volume 9 • Issue 1 18 Giuffrida, M. (2014). Unleashing the power of talent analytics in federal government. Public Management, 43, 7. Hanlon, C., Morris, T., & Nabbs, S. (2014). Program providers’ perspective: Recruitment and retention strategies for women in physical activity programs. Sport Management Review, 17(2), 133–144. doi:10.1016/j. smr.2013.04.001 Harris, J. G., Craig, E., & Light, D. A. (2011). Talent and analytics: New approaches, higher ROI. The Journal of Business Strategy, 32(6), 4–13. doi:10.1108/02756661111180087 Herold, M., Goes, F., Nopp, S., Bauer, P., Thompson, C., & Meyer, T. (2019). Machine learning in men’s professional football: Current applications and future directions for improving attacking play. International Journal of Sports Science & Coaching, 14(6), 798–817. doi:10.1177/1747954119879350 Joshi, A. V. (2019). Machine Learning and Artificial Intelligence. Springer. Kaur, J., & Fink, A. A. (2017). Trends and practices in talent analytics. Society for Human Resource Management (SHRM)-Society for Industrial-Organizational Psychology (SIOP) Science of HR White Paper Series 20. Kim, Y., Bui, K. H. N., & Jung, J. J. (2019). Data-driven exploratory approach on player valuation in football transfer market. Concurrency and Computation, 33(3), e5353. doi:10.1002/cpe.5353 Koseler, K., & Stephan, M. (2017). Machine learning applications in baseball: A systematic literature review. Applied Artificial Intelligence, 31(9-10), 745–763. doi:10.1080/08839514.2018.1442991 Kotsiantis, S.B., Zaharakis, I., & Pintelas, P. (2007). Supervised machine learning: A review of classification techniques. Emerging Artificial Intelligence Applications in Computer Engineering, 160, 3–24. Li, D., & Du, Y. (2017). Artificial intelligence with uncertainty. CRC Press. doi:10.1201/9781315366951 Loland, S. (2015). Against genetic tests for athletic talent: The primacy of the phenotype. Sports Medicine (Auckland, N.Z.), 45(9), 1229–1233. doi:10.1007/s40279-015-0352-5 PMID:26121951 Mankin, J.A., Rivas, J., & Jewell, J.J. (2019). The effectiveness of college football recruiting ratings in predicting team success: A longitudinal study. Research in Business and Economics Journal, 14, 1–19. Marrybeth, J., Brackz, J., Hadson, W., & Zlong, M. (2019). Talent acquisition and talent engagement practices significant impact over employee satisfaction. International Research Journal of Management, IT and Social Sciences, 6, 244–252. Martin, L. (2015). Is socioeconomic status a contributing factor to tennis players’ success. Journal of Medicine and Science in Tennis, 20, 116–121. Mohapatra, M., & Sahu, P. (2017). Optimizing the recruitment funnel in an ITES company: An analytics approach. Procedia Computer Science, 122, 706–714. doi:10.1016/j.procs.2017.11.427 Musa, R. M., Taha, Z., Majeed, A. P. A., & Abdullah, M. R. (2019). Machine learning in sports: identifying potential archers. Springer. doi:10.1007/978-981-13-2592-2 Nassib, S. H., Mkaouer, B., Riahi, S. H., Wali, S. M., & Nassib, S. (2020). Prediction of gymnastics physical profile through an international program evaluation in women artistic gymnastics. Journal of Strength and Conditioning Research, 34(2), 577–586. doi:10.1519/JSC.0000000000001902 PMID:31386634 Necula, S. C., & Strimbei, C. (2019). People analytics of semantic web human resource r’esum’es for sustainable talent acquisition. Sustainability, 11(13), 3520. doi:10.3390/su11133520 Nocker, M., & Sena, V. (2019). Big data and human resources management: The rise of talent analytics. Social Sciences, 8(10), 273. doi:10.3390/socsci8100273 Ofoghi, B., Zeleznikow, J., MacMahon, C., & Dwyer, D. (2013). Supporting athlete selection and strategic planning in track cycling omnium: A statistical and machine learning approach. Information Sciences, 233, 200–213. doi:10.1016/j.ins.2012.12.050 Pappalardo, L., Cintia, P., Ferragina, P., Massucco, E., Pedreschi, D., & Giannotti, F. (2019). Playerank: Datadriven performance evaluation and player ranking in soccer via a machine learning approach. ACM Transactions on Intelligent Systems and Technology, 10(5), 10. doi:10.1145/3343172 International Journal of Business Analytics Volume 9 • Issue 1 19 Peng, K., Cooke, J., Crockett, A., Shin, D., Foster, A., Rue, J., Williams, R., Valeiras, J., Scherer, W., & Tuttle, C. (2018). Predictive analytics for university of virginia football recruiting. In Systems and Information Engineering Design Symposium (SIEDS). IEEE. doi:10.1109/SIEDS.2018.8374745 Pickering, C., Kiely, J., Grgic, J., Lucia, A., & Del Coso, J. (2019). Can genetic testing identify talent for sport? Genes, 10(12), 972. doi:10.3390/genes10120972 PMID:31779250 Pifer, N. D. (2019). Data analytics in football: Positional data collection, modeling, and analysis. Journal of Sport Management, 33(6), 574–574. doi:10.1123/jsm.2019-0308 Ramirez, M. C., Viloria, A., Mun˜oz, A. P., & Posso, H. (2017). Application of multiple linear regression models in the identification of factors affecting the results of the Chelsea football team. International Journal of Control Theory and Applications, 10, 7–13. Rana, G., Sharma, R., & Goel, A. K. (2019). Unraveling the power of talent analytics: Implications for enhancing business performance. In Business Governance and Society (pp. 29–41). Springer. doi:10.1007/978-3-31994613-9_3 Rimler, M. S., Song, S., & Yi, D. T. (2010). Estimating production efficiency in men’s ncaa college basketball: A bayesian approach. Journal of Sports Economics, 11(3), 287–315. doi:10.1177/1527002509337803 Sivaram, N., & Ramar, K. (2010). Applicability of clustering and classification algorithms for recruitment data mining. International Journal of Computers and Applications, 4(5), 23–28. doi:10.5120/823-1165 Srivastava, R., Palshikar, G., & Pawar, S. (2015). Analytics for improving talent acquisition processes. Proceedings of 4th international conference on advanced data analysis, business analytics and intelligence (ICADABAI 2015). Tveter, D. (1997). The pattern recognition basis of artificial intelligence. IEEE Press. Vaeyens, R., Lenoir, M., Williams, A. M., & Philippaerts, R. M. (2008). Talent identification and development programmes in sport. Sports Medicine (Auckland, N.Z.), 38(9), 703–714. doi:10.2165/00007256-20083809000001 PMID:18712939 Van Eck, N. J., & Waltman, L. (2010). Software survey: VOSviewer, a computer program for bibliometric mapping. Scientometrics, 84(2), 523–538. doi:10.1007/s11192-009-0146-3 PMID:20585380 Walter, L., Citera, A., Knowles, K., Lowen, M., Oldenburg, C., Shahin, H., Scherer, W., & Tuttle, C. (2017). Implementation of a recruit visualization tool for UVA football. In Systems and Information Engineering Design Symposium (SIEDS). IEEE. doi:10.1109/SIEDS.2017.7937710 Webb, T., Dicks, M., Brown, D. J., & O’Gorman, J. (2020). An exploration of young professional football players’ perceptions of the talent development process in England. Sport Management Review, 23(3), 536–547. doi:10.1016/j.smr.2019.04.007 Webborn, N., Williams, A., McNamee, M., Bouchard, C., Pitsiladis, Y., Ahmetov, I., Ashley, E., Byrne, N., Camporesi, S., Collins, M., Dijkstra, P., Eynon, N., Fuku, N., Garton, F. C., Hoppe, N., Holm, S., Kaye, J., Klissouras, V., Lucia, A., & Wang, G. etal. (2015). Direct-to-consumer genetic testing for predicting sports performance and talent identification: Consensus statement. British Journal of Sports Medicine, 49(23), 1486–1491. doi:10.1136/bjsports-2015-095343 PMID:26582191 Williams, A. M., & Reilly, T. (2000). Talent identification and development in soccer. Journal of Sports Sciences, 18(9), 657–667. doi:10.1080/02640410050120041 PMID:11043892 International Journal of Business Analytics Volume 9 • Issue 1 20 Rocio de la Torre is an Assistant Professor in the Department of Business Management at the Public University of Navarre (Spain). She is also a researcher from INARBE-Institute for Advanced Research in Business and Economics. She holds a Ph.D. and a Bachelor’s Degree in Industrial Engineering from the Universitat Politecnica de Catalunya. Her major research areas are mathematical programming for strategic planning decisions in knowledge-intensive organizations (KIOs) and supply chain design. Laura O. Calvet is a Lecturer of Statistics in the Computer Science Dept. at the Universitat Oberta de Catalunya (UOC) and Lecturer of Mathematics & Project Management at the Escola Universitària Salesiana de Sarrià (EUSS). She holds a M.S. in Applied Statistics and Operations Research completed at Universitat Politècnica de Catalunya (UPC) & Universitat de Barcelona (UB) and a Ph.D. in Network and Information Technologies completed at the UOC. She is a member of the ICSO@IN3 research group. Her main lines of research are: - Design of optimization algorithms relying on the use of metaheuristics, machine learning and/or simulation applied to sustainable logistics & computing - Applied statistics & economics: applications in health, disaster management, & learning. David Lopez-Lopez is an academic collaborator at ESADE. He holds a PhD in digital transformation, and a joint MBA from ESADE (Spain) and the University of Duke (USA). He is also managing partner in Fhios, that employs more than 180 consultants. Angel A. Juan is a Full Professor of Operations Research & Industrial Engineering in the Computer Science Dept. at the Universitat Oberta de Catalunya (Barcelona, Spain). He is also the Director of the ICSO research group at the Internet Interdisciplinary Institute and Lecturer at the Euncet Business School. Dr. Juan holds a Ph.D. in Industrial Engineering and an M.Sc. in Mathematics. He completed a predoctoral internship at Harvard University and postdoctoral internships at Massachusetts Institute of Technology and Georgia Institute of Technology. His main research interests include applications of simulation, metaheuristics, and machine learning methods in different fields, including: logistics & transportation, finance, and smart cities. He has published over 110 articles in JCR-indexed journals and over 250 papers indexed in Scopus. ENdNOTES 1 https://en.wikipedia.org/wiki/Moneyball_(film) 2 www.21stclub.com 3 https://barcainnovationhub.com/es/event/barca-sports-analytics-summit-2019/ 4 https://wyscout.com 5 https://www.scisports.com 6 www.intel.co.uk/content/www/uk/en/it-management/cloud-analytic-hub/data-powered-football.html 7 https://www.gartner.com/en/human-resources/insights/talent-analytics