A seismic moment like this used to rest on the decision of one man alone, but not anymore. The decision is referred to VAR, whose evidence concludes that the goal should stand. The verdict is that a German player touched the ball to the South Korean striker. Would the referee have spotted it without the assistance of technology? Not likely. The touch was so sudden and slight as to be barely perceptible to the naked eye. But VAR spots it and confers with the referee, who awards the goal. The system works, and Germany crash out.
It’s not a stretch to see VAR as an analogy for the way automated decision-making can assist business operations. VAR represents a machine assisting a human to make the best decision possible, rather than relying solely on heat-of-the-moment, sometimes erratic refereeing instincts.
Transfer this VAR analogy to a bank that applies automated decision-making, and it starts to make a lot more sense. Substitute the referee in this analogy for a live agent, and VAR for an automated decision-making platform that can make informed recommendations about customer cases by tracing account history. It could flag up a transaction where an elderly customer is withdrawing their life savings to pay a fraudster, stopping a withdrawal that appears above board, and that a live agent could easily miss due to any number of reasons – time restraints, fatigue, or inefficient access to client history, for instance.
This is a perfect example of what Harvard Business Review termed ‘noise’ in their piece on the inherent inconsistencies of human decision-making. Their research reveals that, while company executives estimate skilled decision-making to have a variance across the workforce of just 5-10 percent, the typical discrepancy in judgements from one day to the next is a staggering 40-60 percent. This number differs only slightly across most jobs and sectors, so it more or less applies to any situation in which people are making decisions in pressured environments.
There are few more pressured environments than a World Cup game, where the amount of ‘noise’ potentially affecting each decision – not least from 80,000 or so screaming supporters – is almost unrivalled. One poor decision can have a knock-on effect on the rest of the tournament, just as the effects of one employee error can be felt across a business. The team that benefits from an incorrect call in the World Cup group stage could end up winning the competition, just as a missed email could result in an even bigger mistake further down the factory line. For every bad decision, a ‘butterfly effect’ could be set in motion.
“Is VAR destroying football?” ask the papers, the pundits, the fans. The debate has seen a reappearance of many of the same phrases that were being used when goal-line technology was introduced. Then, as now, resistance to change was manifested in platitudes about how technology “takes the charm out of the game” or “ruins the flow of football”. And yet looking back, who would now argue the case against goal-line technology? Who, Germany fans aside, would want a rerun of Frank Lampard’s “ghost-goal” at the 2010 World Cup?
This arc of reasoning will surely apply to VAR. A few years down the line, when refereeing decisions achieve unprecedented accuracy, few will argue against its benefit to the game. Even naysayers will concede that when technology and humans work in tandem, the best results are achieved.
So, to direct the analogy back to business: to avoid decisions being left solely up to erratic human tendencies, and to give people the best chance of getting decisions consistently right, they need to be backed up by the right technology. A recent survey of over 300 C-suite executives suggested that decision-making in many businesses could be fundamentally improved – 42% said their decisioning was slow, and only 32% rated their organisations highly effective in delivering timely and relevant information to decision-makers. Without the right support, people can only make so many correct decisions before one slips through the net. We’re only human, after all.