Sports Talent Identification and Artificial Intelligence: Future Champion or Suppressed Star?
Sports Talent Identification and Artificial Intelligence: Future Champion or Suppressed Star?
In recent years, sports scouting has transformed from instinctual practice to data-supported science. Football is one of the many sports undergoing a revolution in youth scouting. “Today, on the platform, there are 250,000 players. Last year, there was 130,000. Next year, there will be half a million.” says Benjamin Balkin, co-founder of Eyeball, a digital scouting firm that tracks the performances of amateur footballers all around the world.
Will the spread of these artificial intelligence technologies lead to more equitable access, or will it exacerbate pre-existing inequalities?
Talent Identification and Artificial Intelligence in Sport: the Great Leveller or the Great Divider ?
Artificial intelligence (AI) will reshape how we play, watch and experience sport. From elevating fan experiences to predicting injuries before they even happen, AI and sports’ marriage is undeniable and possibly unbreakable. Recent research shows the size of AI in the sports market. In 2024, it was USD 8.93 billion and expected to grow at a compound annual growth rate of 21.14% from 2025 to 2034, from USD 10.82 billion in 2025 to over USD 60.78 billion by 2034.
The NFL’s Digital Athlete uses AI and machine learning (ML) to build a complete view of the players’ experiences. The Digital Athlete records player positions and locations on the field during game week as well as player tracking data during team practices, which translates to over 500 million data points weekly. Soon, NFL teams will be able to create customised training and recuperation plans and perform real-time risk assessments for player injuries during games thanks to the Digital Athlete.
AI integration in sports scouting is a quickly developing topic. In order to locate and nurture athletic talent, scouting will increasingly depend on data-driven strategies that combine conventional methods with cutting-edge analytical tools.
This article covers three areas of AI in sports talent identification:
- Finding talent – the changes in sports scouting
- Predicting talent – forecasting future performance
- Measuring talent – comparing and ranking talent
After visiting these three areas, the article will evaluate what is good about AI in talent identification, the not-so-good, as well as the challenges that have/will arise.
Finding Talent
Why discover talent in the first place? It may appear obvious, nevertheless, a quick review. Firstly, professional sport is a competition. Countries, teams and individuals (should) want to win and if you are\have the best athletes, you are more likely to win. Thus, the best players are hunted. Secondly, the best ‘talent’ comes from all over the world, regardless of where and who the individual is, the best should be recruited and not just the most convenient talent. Thus, this second reason is about equality.
Thirdly, professional sport is a business. Businesses invest expecting to turn a profit. Clubs and countries invest hugely in player development pathways. The International Centre for Sports Studies named Benfica’s academy the world’s most lucrative in 2022. The club made €379 million from the sale of academy players since 2015. Thus, the best talent tends to be the most lucrative.
AI Scouting in Talent Identification
The process of identifying talent is intricate and requires a thorough comprehension of players’ tactical, psychological, technical, physical, and sociological abilities. Typically, scouts go watch ‘identified’ athletes during a game or practice session, administering conventional physical aptitude tests in controlled settings if they ‘see’ talent and/or potential talent. As a result, players depend heavily on scouts or agents to show off their abilities and get noticed. Identifying talent has now changed. AI will play a large role in sports talent identification.
Significant advantages of AI in talent spotting include consistency and objectivity. AI makes it possible to evaluate talent more accurately and impartially by minimising human prejudice. Furthermore, AI has the ability to democratise talent identification by reaching rural and neglected areas of the world where traditional scouting techniques might not be as effective.
AI has the potential to be a tremendous leveller, assisting developing countries and others with less developed sport ecosystems in identifying latent talent and overcoming financial and geographic obstacles. Conventional scouting frequently encounters geographic constraints. However, AI can search through enormous databases of youth competitions, online match footage, and even social media platforms to find undiscovered talent that human scouts might miss.
Untapped Talent
Aligning with the spirit of the 2024 Paris Olympic Games slogan, “Games Wide Open”, a new AI-powered talent-spotting system that evaluated athletic potential was made available for fans to test. In order to assess participants’ power, explosiveness, endurance, response time, strength, and agility, Intel developed a system that collected data from a variety of physical tests, such as running, jumping, and grip strength measurements.
In Senegal, this technology has already been used to evaluate more than 1,000 youngsters across five distinct villages. One child in particular showed outstanding promise out of the 48 children with significant potential identified by the system. These kids have been given the opportunity to participate in sports program run by the Senegalese National Olympic Committee.
Another example is how Ecuadorian centre-backs have gained international recognition with the help of Hudl Wyscout – the largest collection of football videos in the world. The often go-to places for talent in South America such as Brazil and Argentina is shifting. Whilst Hudl Wyscout itself suggests three reasons for the change in attitude to talent identification, it singles out global visibility of players as being the most crucial. Scouts and recruiters can now view the potential on display in Ecuador, which previously went unnoticed.
Clubs are using AI scouting to great benefit. The Portuguese football club Benfica is unable to generate the same amount of sponsorship money or commercial revenue as clubs in the Premier League or La Liga. As mentioned earlier, since 2015, the club has made £426m million from the sale of former academy players since 2015. The team is currently regarded as a true talent factory, producing everything from center backs to elite forwards. They have achieved this using the system Eyeball, a digital football scouting firm.
AI in sports talent identification is not only the future, but also the present. There is no doubt that all sports are already using AI in their talent recruitment programs. Major League Baseball’s Statcast system tracks player movements, pitch characteristics, and other data points to providing coaches and scouts with comprehensive data-driven performance insights to identify promising young talent.
Images taken from aiScout website.
Using a mobile device and artificial intelligence, the fully automated Talent Analysis & Development Platform aiScout can analyse athletes all around the world. helping scouts from national teams and professional teams like Burnley FC and Chelsea FC visit colleges and private academies. Rebekah who lives in Jersey never had scouts attend her games. After downloading the aiScout app and participating in the Chelsea FCW drills, the app evaluated and scored her, leading to her being scouted. She is now a member of Chelsea FC Women’s u18 team.
By using AI video analysis throughout their youth system, Ajax was able to reduce scouting time by 70% and improve talent discovery accuracy by 45%. Every touch, pass, and movement is automatically tracked and analysed by the technology, freeing up coaches to concentrate on player development rather than tedious film analysis.
Automated Video Analysis
Because AI can identify patterns that are invisible to the human eye, it also makes it easier to study sports footage and identify areas for improvement. AI-assisted scouting is quicker than human scouting, and the computer will undoubtedly choose the player’s particular actions and frames that are most relevant for assessment. AI can go through vast amounts of data to identify a player who would be a good fit that a scout might not be aware of.
The way teams evaluate player performance has been completely transformed by computer vision technology. Every move, pass, and choice a player makes throughout a game may be tracked by AI-powered video analysis, which can reveal details that the human eye might miss. Teams can assess players more fully and reliably by automating the video analysis process.
AI can identify prospective athletes from submitted mobile phone videos by utilizing machine-learning algorithms and video processing, greatly expanding the pool of talent. These inputs allow the algorithms to identify which ones merit more research, directing human scouts’ attention to the most promising ones.
Automated Video Analysis will also greatly help identify talent in positions, particularly defenders in team sports, that are deficient in high-quality metrics. A defender’s actions are greatly influenced by the team’s style; for example, a system that encourages crosses may be reflected in high aerial dual numbers, while tactical setup may have a greater influence on pass volume than individual skill. Important characteristics like scanning for danger, leadership, awareness, and posture, on the other hand, lack clear data counterparts. In the end, having access to high-quality video is essential for assessing defensive talent.
Importantly, sports video annotation is a crucial method for training AI models to understand and assess sporting events. Visual annotation involves the identification and masking of visual data to build complex computer vision models. In sports, it annotates game stages, player positions, and activities. This enables AI to offer comprehensive player performance analysis.
Virtual Reality in Talent Identification
Ideally, when identifying sporting talent, two criteria should be met. Firstly, that athletes are being measured on the same task, i.e., 40m sprint in the same conditions (within reason). Secondly, what is being measured should be measured. In an open-skill, team-based sport, the former point is often challenging when measuring an in-game action. This is where virtual reality (VR) can help more immersive player evaluations as athletes are put through the exact same task.
A recent study evaluating athletes’ performance found compelling proof that well-crafted virtual reality tests can yield outcomes that are on par with conventional in-person evaluations. The three virtual reality tests evaluated reaction time (drop-bar test), jump ability (jump and reach test, which includes lower extremity strength ability), and accuracy (hits with the objects) for a variety of complex movement patterns (parkour test). Findings indicated that the VR tests were highly repeatable, especially for reaction time and leap height. There was a less than 1 cm discrepancy between the VR leap test and real-world results.
Inexperience with VR navigation meant participants took longer to finish the VR parkour course than the actual one. Nevertheless, the performance difference was reduced with the addition of a virtual opponent. This study shows that VR is and will have an ever-increasing role in talent identification. VR will help obtain standardised skill assessments, providing coaches with trustworthy metrics to spot talent early. Thus, a wider spectrum of athletes from underserved communities who might otherwise go unnoticed.
In Formula 1, virtual racing is now seen by leading Formula 1 teams such as Mercedes, McLaren, Red Bull and Ferrari as a key component for drawing in and identifying future stars. There are countless instances of drivers who have successfully transitioned from the virtual track to the real circuit, demonstrating that skill developed online can, in fact, lead to success in the real world of motorsport.
Before competing on the international scene, drivers like Max Verstappen and Lando Norris have refined their racing abilities in the virtual world. Lucas Ordóñez and Jann Mardenborough, who starred in the Gran Turismo film, are two examples of GT Academy graduates who have demonstrated how virtual racing may uncover latent abilities and offer a feasible route to professional racing careers.
You just don’t get this anywhere else.
Massive congrats to sim racer Lucas Blakeley on defeating 4-time F1 champ Vettel. Surely the biggest moment in his life. #ROCSweden pic.twitter.com/yXJ8fzQane
— Race Of Champions (@RaceOfChampions) February 5, 2022
Lucas Blakeley, the Open Qualifier (available to anybody with a gaming PC and an internet connection), defeated four-time Formula One World Champion Sebastian Vettel in their head-to-head race in the Race of Nations Cup. Lucas currently competes in a worldwide virtual Grand Prix as a racer for McLaren’s esports team. He started go-karting at the age of seven and soon realized he wanted to be an F1 driver. Lucas claims that he was unable to enter the sport because it was too costly. “For many racing families out there, that’s the harsh reality,” he states. Codemasters’ F1 game is used for the qualification rounds, and representatives from the major teams attend scouting sessions with the top players.
However, a significant barrier deterring athletes from incorporating such mnemonic devices into their formal training regimen is the absence of tactile input which is absent in VR training. The impact of the ball against the shoes, the feel of the grass beneath the foot, and the physical touch that occurs during in-person training with other players are all currently not replicated by virtual reality.
The first Footworn Player Development System
AI-driven technologies are already being used by organisations like PlayerMaker to evaluate and find talent in remote locations, giving young athletes a platform to display their abilities and gain access to specialized training programs. AI has the potential to reveal latent talent and provide a more inclusive approach to athlete development by facilitating earlier interventions and customized training.
Team sports fundamentally have gatekeepers, especially professional teams. These gatekeepers are often scouts. AI is literally pushing the gate aside for potential talent. In 2023, PlayerMaker boasted of the first Footworn Player Development System for football in the world, providing practical insights to improve mental, tactical, technical, and physical performance. Six-axis smart sensors measure ground impact, ball contact, and foot rotation precisely by sampling movement 1000 times per second.
In football, widely used tests such as a 5-10-5 agility test, a Yo-Yo endurance (or Yo-Yo intermittent) test, a vertical or broad jump power test, and a 10m/40m speed test. Technical assessments are subjective and circumstance dependant because they are based solely on observations, but the physical data from these tests is objective (remember, what is measured doesn’t mean it should be). Lastly, because players born in the first quarter of the year are usually chosen for their physical advantages, placing too much emphasis on physical skills may result in a common mistake.
In its white paper, three technical and two physical talents make up the first five PRS (Proprietary Rating System) skills. These consist of dribbling, speed, agility, first touch, and two feet. All are seen to be essential for success in international football. It is crucial to remember that these scores are determined during games, and supplemented by other metrics that are determined during preparation and competition.
Predicting Talent
In sports, predictive performance analysis is the data-driven process of forecasting future events, including player performance, injury risk, and game outcomes, utilizing both historical and current data.
Long-term Potential
An AI-powered tool evaluates athletes’ speed, agility, technique, and other performance parameters by analysing video of performances over time after recognizing gifted individuals. The trajectory of exceptional talent can then be assessed using this data by contrasting it with benchmarks and historical data. This allows AI to monitor an athlete’s progress over time and provide real-time improvement recommendations. This proactive approach increases the likelihood of success by ensuring that talented athletes are identified and nurtured at an early age. Though, it must be noted that historical benchmarking can work against ‘new’ talent, i.e., athletes that break the mould.
These days, teams use complex analytics to forecast how a player’s abilities will transition to more competitive levels. For example, in baseball, predictive models may use elements such as these to assess a minor league player’s potential:
where every element is a complicated function of quantifiable metrics in and of itself.
The NBA’s Golden State Warriors are perhaps the best club to demonstrate the potential of predictive analytics. A championship dynasty founded on knowledge gained by studying shot effectiveness in space. According to their models, most players’ projected value from 15 feet was lower than Stephen Curry’s predicted points per shot from 30 feet. Over a season, the variation in three-point shooting percentages even out, making it a viable tactic. For other high-percentage shots, three-point shooting produced the ideal space. The Warriors’ adoption of these ideas resulted in three titles in four years and fundamentally altered basketball strategy across the board.
Adapting to Different Playing Styles
Performance analysis has historically placed a strong emphasis on using discrete performance indicators that give team coaching staffs useful information. Research on team playing styles has been around for a while. On the other hand, studies on individual players’ playing styles have only lately surfaced, with earlier years largely concentrating on assessing and ranking players according to performance metrics. Soccer players’ performance, however, is multifaceted and influenced by a variety of contextual factors, including match status and location, as well as individual tactics, psychology, personality, technique, and physical attributes. Therefore, a football player’s unique profile-style may be identified by considering all of these criteria.
A systematic review of 12 football studies identified 26 players styles (33 while considering comparable styles for both sides of the field) which can be seen in the image below. Thus, AI’s predictive solutions are able to inform scouts/teams of players’ playing styles as well as evaluate how well a ‘talented’ player would fit into the new squad or a particular system.
Measuring Talent using AI
Even when talent has been identified, it needs to be measured accurately so the talent is ordered. In other words, once the talent has been found, it needs to be ranked. Thus, sports analytics is key.
Sports Analytics
Arguably the most important job of AI, is managing vast volumes of data with the highest level of accuracy in the shortest amount of time. Since information is typically obtained by an individual’s personal observation, traditional scouting frequently yields relatively little information. AI is able to compile and examine data from a range of sources, such as player statistics, game footage, physical attributes, injury records, etc.
Algorithms use this data to evaluate, and uncover details or information that a scout might not capture. When it comes to match judgments, AI might, for example, highlight players who consistently step up or identify talent in lower divisions and non-traditional soccer-playing countries. AI is revolutionising how talent is measured and is just the latest in the evolution of sports analytics in helping create more robust, objectional and relevant stats.
More Robust, Meaningful, Relevant Stats
The techniques for assessing player performance in team sports have changed throughout time, moving from basic measurements like box-scores (goals, assists, and hits) to more sophisticated metrics like offensive rating, wins above replacement, and expected goals. These criteria were generally developed to help teams identify players with market value and talent.
There are a number of metrics that evaluate a player’s overall contribution to the success of their team. For example, in American football (e.g., Quarterback Rating), Australian football (e.g., AFL Player Rating), baseball (e.g., Runs Created), basketball (e.g., True Shooting Percentage), and ice hockey (e.g., Expected Goals) and Wins Above Replacement (WAR). WAR is one of the more widely used advanced metrics in baseball, determining how many more wins a player can bring to his team than a player who is regarded as “league-average”. The WAR is regarded as a crucial metric to guide hiring decisions and contract extensions in order to sustain and improve team performance.
In football (soccer), Hudl Statsbomb have created a metric called On-Ball-Value (video explaining OBV). Football is a low-scoring game, goals naturally draw a lot of attention. After monitoring shots, sports analysts began using Expected Goals to gauge the quality of those shots. However, as shots make up less than 1% of all football-related actions, there is a lot of football in the middle that is not being tracked. Finally, analytics is able to measure what happens between the two boxes. Models of Possession State Value (PSV) emerged.
The idea behind PSV models is to quantify and objectively assess the worth of every event that taking place on the field. This is accomplished by evaluating the shift in a team’s chances of scoring and giving up as a direct consequence of an incident. The OBV allows players (talent) to be measured objectively more accurately, thus determining which talent is performing the best. In other words, ranking the talent.
In the 2024/25 season, the Team of the Season for the Bundesliga, Premier League and LaLiga is shown above. It shows by each position which player had the greatest impact. Naturally, attacking players do score highest, after all, you can still win matches with poor defenders, but cannot win matches if you do not score. And whilst fans, managers and clubs will always have their preferred style (taste) to watch, the OBV metric (and others similar) will hopefully reduce why some players in the best teams have been massively under- or over-valued.
Fairer, Performance-Based Contracts
In 2021, before he renegotiated his contract with Manchester City, Kevin De Bruyne and his legal team hired Analytics FC to do a research to examine every facet of his value to the squad. This marked for the first time a player seeking this type of study to strengthen contract negotiations.
“I always make my own decisions. I like to use data and analytics to help me negotiate contracts, and I don’t think you need an agent for that. I think if you have the right information and the right people around you, you can do it yourself.” Analytics FC used objective data from the study showing that De Bruyne was one of the Premier League’s finest players, especially when it came to creating chances. Using their own Goal Difference Added (GDA) performance model, which considers the impact that each player’s touch has on a game (both positively and negatively), De Bruyne’s production really distinguished him as the greatest in Europe.
The report also includes salary comparisons with other players, highlighting De Bruyne’s value in comparison to some of Europe’s top attacking players who earned higher pay than the Belgian but substantially lower GDA. Consequently, De Bruyne secured a new four-year, £104 million contract with Manchester City, which included a 30% wage raise. Since De Bruyne used sports analytics to help renegotiate his contract, many players have followed suit. However, the primary concern is the enormous expense. The service from Analytics FC is likely to be beyond the budgets of many players, with even one of their directors stating, “It costs a lot.”
Ranking players more objectively / More objective awards
Should individual sporting awards (that do come with great financial rewards) be wholly objective, based on a single metric alone? It most circumstances, yes!
Currently, there are many individual awards that are solely handed out based on statistics. These tend me be very basic, counting statistics, such as Golden Boot Award in football (the most goals scored in a competition). More advanced, is golf’s prestigious Vardon Trophy for the golfer with the lowest adjusted scoring average on the PGA Tour. (Note: The Bryon Nelson Award is similar but a lower minimum round qualification.) Instead of focusing solely on a golfer’s raw score, this adjusted average is a formula that accounts for the difficulty of the courses played and the aggregate scoring performance of the whole field in each tournament.
When awards are determined by objective statistics, there is little to no argument about the true winner being recognised. It becomes a problem when human opinions are used to determine which athlete ‘deserves’ to win. Typically, this arises for the ‘best’ player awards. In most sports, especially football, these awards are decided by a select panel of ‘experts’ or journalists simple voting. One of the most prestigious awards in sport is the Ballon d’Or (with the Femine Ballon d’Or in its early phases). The Ballon d’Or is an annual football award that is presented to the world’s best footballer.
Every year there is uproar between fans on who ‘deserved’ the award (often arguing for their favourite player). A debate exacerbated when the award was cancelled in 2019 due to COVID with fans in uproar protesting that their player was the deserving winner (read The Ballon d’Or That Never Was). Furthermore, what makes the Ballon d’Or award lose further credibility is that team success plays a large role in determining the winner. In other words, the top individual prize is based on their teammates and opinions of journalists.
The Great Leveller or the Great Divider?
In the past, scouting relied heavily on intuition and being in the right place and the right time. No more. AI-powered systems now examine every move made on the field, revealing hidden patterns and abilities that conventional scouting techniques frequently overlook.
The Great Leveller
There are three main advantages using AI to identify sporting talent. First, it widens the talent pool. While conventional scouting typically concentrated on prestigious universities and events, it also lacked the ‘man’ power. Humans are limited by time and resources, and there just aren’t enough scouts to evaluate everyone. By spotting outstanding players from remote locations, AI changes the game.
Second, the use of AI allows for greater objectivity. Computer vision is able to discover patterns not visible to the human eye, thus forming deeper, more robust statistics that measure athletic performance more accurately. Liverpool FC, for example, employs AI to assess “football IQ” traits like positioning and response time.
Third, AI helps removes bias (not completely – see below). AI’s analysis is more impartial since it is not influenced by human preferences or reputations. Researchers have shown in a first-of-its-kind study in Canada how artificial intelligence (AI) may be used to lessen bias in professional sports scouting while maintaining the vital function of human expertise. The study titled “Blind scouting: using artificial intelligence to alleviate bias in selection”.
The researchers used AI-anonymised game footage to test an experimental “blind scouting” approach in partnership with a major North American soccer team. Scouts used a technique called “think-aloud” cognitive analysis, in which they were instructed to evaluate players while expressing their mental processes. Interestingly, the use of AI to eliminate distinguishable features improved the scouts’ emphasis on tactical performance and reduced their attention to physical characteristics that may cause bias.
The Great Divider
There are two main issues with using AI to find sporting talent. Firstly, AI is expensive. AI is costly because it requires a lot of data management, specialized hardware like GPUs, hefty development costs, and continuous maintenance. Whilst these significantly high costs could be reduced by teams/leagues of the same sport sharing the AI model, or that teams pay a lower subscription fee to a private company, nevertheless AI is still going to cost money.
A study found that the initial cost of VR training was more than that of traditional training, at $327.78 per participant as opposed to $229.79. However, it did highlight the cost of VR training decreased dramatically to $115.43 per person when these expenses were distributed over a three-year period with frequent use of the training package. Football’s first footworn technology PlayerMaker costs £20 p/m with 12 and 24-month subscriptions costing £199 and £299 per year, reducing the cost to about £16 and £12 p/m. The CITYPLAY version is slightly more expensive.
The cost of AI is linked to the second disadvantage, it creates further inequality. Families that can afford this technology for their children are able to showcase their child’s abilities to top professional clubs/scouts whereas youngsters from poor families do not get this opportunity. Furthermore, with the increase in VR being used for talent identification, those who have used the technology before have advantages, thus ‘talent’ might be talent + socioeconomic status.
Furthermore, the cost of AI is not only played out at the individual / athlete level, but at all levels, for example, club level. The best / largest teams are going to develop the best AI tools in recruiting talent and are likely to recruit the most talented players, thus widening the sporting competition between teams.
Challenges for AI in Talent Identification
The are three main challenges for using AI in finding sporting. First, the accuracy of the models. AI models are solely reliant on the data that is fed into them. In 2024, a systematic review of 44 studies in football codes (e.g., soccer (mainly), American football, Australian rules and rugby) discussing the use of AI to identify talented athletes or forecast how they will perform during matches, found that AI-based athlete evaluations may not take into account all pertinent and contextual elements. Therefore, practitioners should carefully analyse them before incorporating the results into decision-making processes.
Second, understanding AI models. A common issue with AI models is that they function as “black boxes,” which obscures their decision-making procedures. Decisions about identifying and developing talent are hampered by this lack of openness, which also undermines accountability. Building trust and guaranteeing justice need the development of explainable AI (XAI) strategies that provide an explanation for the judgments made by AI models. XAI can assist athletes and coaches make reliable, well-informed decisions by elucidating how particular parameters, such as biometric data or movement patterns, affect injury risk assessments or training recommendations.
AI systems that have been educated on skewed data have the potential to reinforce and magnify current disparities, known as Algorithmic Bias. Consider an artificial intelligence (AI) talent scouting tool for sports that favours particular physical characteristics or socioeconomic origins, possibly eliminating gifted people from underrepresented groups.
Third, ethical and personal data. AI’s ethical and privacy ramifications are particularly important when it comes to sensitive data, like biometric information, performance measures, and athlete health metrics. These databases are extremely useful for improving performance and training, but there is a serious risk of prejudice, abuse, and unauthorized access. Applications of AI may result in intrusive monitoring or data breaches if they are not properly regulated. Thankfully, contemporary privacy-preserving technology can now safely store, process, and aggregate such data (using techniques like homomorphic encryption and aggregated conversion modelling) without sharing or disseminating it broadly.
Conclusion: Talent Identification and AI
Artificial intelligence is proving to be revolutionary in the coaching field. It is changing the coaching profession from one that relies on talent, experience, and intuition to one that is more data-driven and dependent on outside resources.
AI should support the current scouting models rather than serving as a stand-alone scouting method. It has been shown that the best results are obtained when both exacting analytical techniques and the qualitative evaluations of observers are used.
The Stat Squabbler says:
- AI is enabling a more level playing field. The widening of the talent pool is allowing previously untapped talent a greater opportunity to be identified
- AI models are expensive. With great disparity of resources the top and bottom teams have in leagues, AI has the possibility of becoming an arms race and increasing competition inequality.
- Deeper, more robust AI-generated stats still need to be treated with caution. AI-based athlete evaluations may not take into account all pertinent and contextual elements
Do you agree / disagree with the Stat Squabble? Is AI making talent identification a more level playing field?
Comment below. 🙂
