When it comes to scouting an athlete, various factors can be at play some of which may unintentionally reflect bias. But what if you could strip away everything except raw performance?
That's exactly what researchers at Toronto Metropolitan University (TMU) have explored in a first-of-its-kind Canadian study.
The study, Blind scouting: using artificial intelligence to alleviate bias in selection, was led by professors Louis-Étienne Dubois of the Creative Industries program and Laurel Walzak of RTA Sport Media atThe Creative School.
Their findings highlight how AI can help scouts focus more on a player's performance rather than physical characteristics that may unintentionally influence decisions.
Dubois and Walzak partnered with a professional North American soccer team to put this idea to the test.
Blind scouting
Using a "blind scouting" approach, scouts were shown anonymized game footage videos in which players' visual identifiers (such as race, gender or height) were digitally removed. As they watched, scouts were asked to think out loud, narrating their evaluations in real time a method called "think-aloud" analysis.
The result? A shift in focus from physical traits to tactical decision-making and in-game intelligence, leading to clearer, more objective evaluations.
Dubois and Walzak say this study shows the value of AI-based tools in scouting particularly when assessing a player's ability and suggests the technology can augment traditional approaches to support more objective, performance-focused evaluations, also noting that the study demonstrates how AI can help improve processes without removing the need for human experts.
Role of race
Bias in sports scouting is far from hypothetical it's a topic that continues to surface in real-world decisions.
For instance, in a recent CBC Sports piece, journalist Morgan Campbell questioned whether factors like race played a role in quarterback Shedeur Sanders's surprising drop in the NFL draft.
The article raised concerns that long-standing biases may still affect how athletes are perceived, regardless of performance metrics.
"It reminds us that there are many elements that factor into teams' decisions that may not be reliable and valid predictors of future performance," said Dubois. "Biases and discrimination are still widely problematic. Our study provides one avenue to focus the scouts' attention to what truly matters: Can the person play?"
Study shows how AI and humans can team up
The research, published in a leading peer-reviewed journal, challenges the notion that AI will one day replace human experts. Instead, it shows how the two can work together, with AI helping to surface talent that might otherwise be overlooked.
"This isn't just about saving time or money it's about making better decisions in high-stakes environments," said Walzak. "Our findings suggest AI can be a powerful tool, giving teams a competitive advantage, particularly in talent evaluation, where small biases can have big consequences."
And the implications go beyond the field. As one of the first major studies on AI in sports talent management, it opens the door for similar strategies in hiring, computer visioning and HR across industries.
With sports organizations increasingly adopting advanced technologies, TMU is leading the way in exploring how these tools can be used ethically and effectively. It's part of the university's broader commitment to driving innovation where creativity and technology intersect.
