The world of technology in sports at the professional and collegiate level has undergone a series of changes over the past eight decades. 1936 was the first year people across the globe were able to watch sports events on the television, and since then technology has become a necessary aspect of sports.
However, one staple has shined on throughout history—human officials and referees.
Since 1896, when judges for the Olympic Games brought their own stopwatches to keep time, several professional sports have relied on technology or artificial intelligence (AI) to make critical calls. Wimbledon implemented computer technology in tennis in 1980 to determine whether a ball is out or on the line, and football turned to instant replay in 1986 to overturn missed calls.
Without instant replay, spectators in the crowd may have been confused as to how the Oregon Ducks scored a touchdown in the 2014 PAC-12 Championship after Kaelin Clay of the Utah Utes dropped the football at the 1-yard line in an early celebration.
Without AI in tennis, line judges would rely on human error to determine if a ball is out, even if it means getting it wrong within 3.6 millimeters.
Who could forget the moment Michael Phelps won the 2008 Olympic gold medal in the butterfly? Phelps beat his opponent by 1/100th of a second. Had by-the-millisecond touchpads not been invented yet, who’s to say who would have won?
Basketball referees are not exempt from error either. NBA officials have caused controversy over the years in several ways, most famously Tim Donaghy, who spent 15 months in prison for gambling and fixing professional basketball matches.
With sports gambling continually on the rise, questions have been raised as to whether or not switching to AI officiating in NBA basketball could have an impact on the betting scene. Would gamblers be more comfortable placing a bet on a game knowing there would be zero mistakes made by the officials?
NBA players have voiced their own opinions on whether or not AI should be implemented as referees. Former Golden State Warriors teammates Kevin Durant and Stephen Curry fall on opposite ends of opinion.
Durant, in an interview at annual technology convention TechCrunch Disrupt, said he prefers “human judgement” during games and joked he could get away with more lax or petty fouls that a robot would catch.
Curry, on the other hand, supports the idea, though claimed he would still argue bad calls despite a robot’s near-perfect accuracy.
Basketball referees can get into especially hot water when they miss critical calls in the final minutes of games. Mark Bartruff—a Division 6A high school referee with experience officiating college basketball games and NBA scrimmages—supported the idea that a referee could get fined if they blow a call that determines the outcome of a game.
However, Bartruff also noted that everybody makes mistakes, and the end of a game means the end of the game—period. “Somebody won and the other one lost…after review, I [saw] we missed that call. But, oh well,” Bartruff said.
When asked about the concept of involving robot referees in NBA officiating, Bartruff responded, “Robots are involved already. All the time.”
AI is used in basketball not only in slow motion replay, but for ball and player movement on the court. Since 2013, there have been SportVU cameras set up in every NBA arena to track players and ball movement in order to record stats and gameplay for teams across the league.
As the sports world experiences the fallout of COVID-19, conversations are taking place regarding changes to longstanding practices that wouldn’t have otherwise occurred.
Specifically as it relates to COVID-19, the implementation of AI officiating could play a role in limiting human contact and keeping players and coaching staff safer during games. NBA commissioner Adam Silver said the NBA season could continue without fans present at games, saying “It was a larger decision than just the NBA.”
Hypothetical conversations—like the possibility of robot referees—can be considered more seriously as league officials contemplate the future of their sport in light of the disruption caused by COVID-19.
More frequently when this possibility has been considered, the discussion has focused around eliminating referee error. Basketball is regarded as a fast-paced game, which can make it easy for a referee to make mistakes. This was all too apparent in the February 7 NBA matchup between the Portland Trail Blazers and the Utah Jazz.
In the last seconds of the fourth quarter, Blazers point guard Damian Lillard drove to the basket and shot a layup, resulting in an illegal goaltend from Jazz center Rudy Gobert.
Except in this case, there was no call on the play, which made it impossible for Trailblazers coach Terry Stotts to use his Coach’s Challenge.
A Coach’s Challenge allows a coach from either team to stop gameplay at a specific moment during a game when they believe a call has been missed, forcing the referees to review the video cameras. This challenge may only be used once, and can only be implemented if a call is made. It does not account for a no-call like the play missed in the Portland versus Utah game.
At the time of the play, referee Josh Tiven claimed he had not seen the goaltend and maintained that stance despite the Trailblazers’ pleads for a replay of the video. The Blazers would go on to lose the game 117-114. Following the game, the NBA acknowledged the officials had missed the call.
Currently, there are no NBA rules in place to protect a team from being the victim of a no-call, which leaves a gray area for referees to make severe mishaps. Scenarios like this one bring to light the possibilities that could be opened by using AI referees in NBA games to eliminate costly missed calls.
Fair play is a critical element in sports. No team wants to feel they’ve been shorted or overrun by a referee who is not 100% accurate. The primary factor brought up in opposition of AI officiating is the impact on the pace of the game. Would robotic referees make NBA basketball too slow?
Any small infraction that can be picked up by a robot designed to catch fouls would cause more stoppages during games, and that’s not stimulating for the fans. Karen Levy, a professor of law and information science at Cornell University, claimed fans wouldn’t want every minor foul or offense to be called, stating that these would create “a more mechanical and pedantic game.”
Some believe AI referees in basketball would change the fan experience, slowing down the game significantly and unnecessarily. There is also the possibility that AI referees would increase the number of NBA players fouling out each season, since perfect accuracy would increase the number of foul calls that would otherwise go unnoticed.
Another important aspect when considering the implementation of AI referees in NBA basketball—let alone other sports—is the job losses so many will face. There are 70 staffed NBA officials, and a staggering 950 in D1 college basketball alone. That’s more than 1,000 people without jobs if they were to be replaced by robots.
According to a study by the Center for Business and Economic Research at Ball State University, over 85% of job loss from the years 2000–2010 was due to automation development replacing human production.
Few could argue sports would be the same without the technological advancements that track athlete’s abilities and statistics to help further their development. The question is, are things moving too quickly?
Sports were created for people to enjoy themselves in a competitive setting. If all of the control is placed into the hands of robots, will that dull the experience for fans?
Basketball, like every sport, evolves with time. Undoubtedly there are future technological advancements to come, but the human aspect of referees will always be at the heart of basketball.
As Bartruff said, nobody’s perfect. NBA players work on their game just as referees review tape, looking for ways they can improve themselves for next time.
The solution to increased foul calls is to adjust the foul sensitivity of the AI. We do this in video games already, wherein foul sensitivity can be adjusted with a slider. It can easily be done when AI is implemented. In fact, the sensitivity can be adjusted so specifically, that the NBA can control the pace of the game and decide how fluid the game can be. Touch fouls can be eliminated, flopping won’t be rewarded, and the rules as written can be enforced.
Yes, refs will lose their jobs. But people lose their jobs all the time due to automation. Let’s not stop progress in the name of joblessness. A better officiated game is a better product.