Recruiting and NHL Drafting Under the Limelight

Predicting performance:

• Through skills and references
• Without bias
• Using analytics 
  

How do you accurately predict performance from experience and references? 

Just as predicting performance is difficult in employee recruitment, the same is true in the realm of sports. Scouts have their work cut out for them, as they must assess young players, even teenagers, and attempt to project their capabilities into the future. 

The primary method of assessing a hockey player’s performance has always been watching them play live. Scouts travel the world to evaluate players based on criteria provided to them by their team’s management and pass on feedback to the recruiters at their organization. Scouts also talk to coaches, managers, players, agents, and other scouts to help create a clearer picture of a player’s strengths and weaknesses. Compiling this information, organizations then create lists ranking players in order of interest to their team.  

Though scouts have many sources at their disposal to learn as much as possible about a player, the tricky part is taking that collected data and turning it into a prediction of the player’s future performance.  

Meanwhile, what a recruiter looks for in a candidate depends on the industry and specific organization or company’s needs. Across sectors, however, interviews and experience are not the solid indicators of performance recruiters once thought.  

A Florida State University study demonstrated, to the researchers’ surprise, that there is no significant correlation between an employee’s prior work experience and their performance in a new organization. Past behaviour predicts future behaviour, right? Well, the study shows not so much! The employee may not have succeeded in their previous position, but seeing it listed on a resume or CV gives recruiters no way of knowing this. So, if traditional interviews and experience don’t help recruiters accurately predict performance, what does?  

• Asking behavioural questions in interviews  
• Collecting work samples from portfolios  
• Using job-relevant tests to assess a candidate’s skills*

  *Be sure to find a balance between asking candidates to demonstrate skills in job simulations and asking for too much unpaid work. These tests need not be unpaid (an employer willing to compensate candidates for their prehire work may fare better in a candidate’s eyes), nor should they exceed the size and scope of the business (smaller companies shouldn’t expect candidates to complete several large tasks to have a shot at a job, or they may find themselves struggling to hire at all).

NHL scouts and recruiters both have their work cut out for them in assessing experience and recommendations. Hockey scouts have easier access to proof of experience than recruiters. They also have the leg up in terms of references. While scouts have numerous people ready and willing to provide feedback on players, recruiters only have access to candidate-selected references. However, not all recommendations are equal, leading us to our next difficulty in predicting performance: bias. 

How do you predict performance without bias? 

Overcoming biases in interviews is crucial in attempting to even the playing field in recruitment. According to Dr. Adam Grant, organizational psychologist and contributor for the New York Times, “Across industries and occupations, economists find that even when candidates’ résumé are identical, those with white-sounding names like Allison and Matthew get 50 percent more callbacks than those with African-American sounding names like Lakisha and Jamal.”  

Many other factors may influence and lead recruiters to make hiring decisions that reflect their preconceived notions of the “best” candidate, rather than who the top talent truly are. Recruiters may be more likely to choose a candidate who went to the same university as them, who shares interests with them, or who simply appears very confident.  

One of the ways recruiters and anyone else present in interviews can overcome these biases is by conducting structured interviews, in which every candidate is asked the same set of predetermined questions and their responses are graded on the same scale. To conduct your own structured interviews, identify which skills and qualifications are essential to the job and ask behavioural and situational questions that assess the candidate’s abilities in these areas. This interview style can double or triple accuracy in predicting performance on the job.  

Just as recruiters may carry biases with them into interviews, hockey scouts may go into a rink with preconceived notions about a player that affect their judgment or cause them to miss someone else. This is the one of the biggest mistakes scouts make and they should be careful of listening to outside influences. 

Whether it’s internal bias or external influences, both scouts and recruiters must work hard to combat bias in their work. With challenges such as these in the endeavor of predicting performance it's no wonder that data analytics has increased in importance in all fields of recruitment. 

The power of data analytics  

It’s only within the last 15 years or so that recruiters and scouts alike have truly embraced advanced data and analytics in their work. Now, every NHL team has an analytics department, whose job involves studying opponents and analyzing player tendencies in advance of the draft. This trend, compounded by the COVID-19 pandemic, has meant that scouts are increasingly using video playback and other forms of assessment to determine players’ potential value to a team. 

Some of the foundational metrics of hockey analytics include Corsi (the raw sum of all shot attempts), Fenwick (the raw sum of all shot attempts excluding blocked attempts), and expected goals (how likely a given shot is to become a goal based on historical success rate). There are also many micro-stats and other miscellaneous statistics a scout may use to measure a player’s performance.  

Recruiters also have a wealth of data available to them, from personality evaluations and scientific aptitude tests, to resume databases and applicant tracking systems. Perhaps recruiters should take a page out of an NHL scout’s playbook and utilize data even more than they currently do to bolster their chance of hiring the best candidates.  

Predictive analytics has been shown to save up to 23 hours a week in labour. By using data and algorithms to recommend the most fitting candidates for the role, predictive analytics allows recruiters and hiring managers to spend less time on shortlisting and pre-screening candidates and more time on their other duties. For example, applicant tracking systems gather inputs from CVs, cover letters, work samples, and other sources and use predictive modeling to determine whether a candidate is a good fit. By widening the hiring pool and removing some of the biases inherent in the traditional hiring process, data analytics helps companies and organizations make the best hire, forming part of a holistic hiring strategy.  

To incorporate predictive analytics into your hiring process, first identify your KPIs or the data you want to use, such as lead time, candidate success, or quality of hire. Then, choose your tools. Many HR platforms and data analytics tools are available, so it’s worth taking the time to research which best suits your organization’s needs.  

Using data analytics doesn’t take the human aspect out of the recruitment process; rather, it frees up time to focus more on the tasks that truly require a human to decide who to add to the lineup, on or off the ice. While the fruits of your recruitment labour may not be as widely anticipated, scrutinized, or celebrated as that of an NHL scout, the end goal is the same: find the next perfect addition to the team.