What the VAR controversy tells us about Technology

How video assistants could actually help soccer without the current downsides

I’m a huge soccer (and to a degree sports) nerd. I know that on a rational level being a fan and watching sports on TV doesn’t make sense. Neither does it advance humankind. It’s just one of my guilty pleasures.

I’m also a fan of equitableness. (If you think well isn’t everybody? then you haven’t heard enough American VCs.)

In sports, one can be treated unfairly by fortune (in some disciplines more than in others) or by injuries. That’s luck, and we can’t do anything about it. What we can try to bring down is the amount of false calls by referees.

So for years I was arguing with friends and fellow fans that soccer referees should get help from technology — a pretty normal thing in other sports. The basic counter-argument was often the removal of emotions. And yes, that’s right, false calls triggers emotions, but only the bad kind: Schadenfreude or a feeling of undeservedness in the winner, a feeling of betrayal or even hatred against the referee (and the other team who isn’t at fault) in the looser.

In 2018 the International Football Association Board (IFAB) finally introduced the video assistant referee (VAR).

It’s a disaster.

As Gary Lineker said: “If VAR was a manager, he’d have been sacked weeks ago”.

It’s not that too much results are being changed — actually this is a sign that a system like this was in need.

It’s that:

• Too many decisions are still questionable
• The VAR is often getting applied differently
• Goal celebrations are often premature because the scene is getting reviewed (sometimes up to 10 minutes)
• It’s not clear which decisions on the field can be overruled and which can’t

The public opinion sees the technology as the enemy. Of course, discussions with friends turned to a I told you so.

No, technology is not to blame. Technology is a tool. Tools are important, they are what distinguishes us from animals. But Tools have no agenda, you can use them for good and for bad.

Having said that, especially computer technologies are good when there is a binary decision to make: 0 or 1, yes or no.

Applied to sports, the guiding principle should be obvious: Let the black and white decisions to the machine, keep the “Fingerspitzengefühl” to the human.

Here’s how I would design a tech-aided system:

• If the ball crosses the goal-line its a goal — that is already the case and nobody — I repeat nobody — has any issues with that. It’s instant, no discussions, no one feels betrayed.
• The same should be applied to every time the ball crosses a line: think throw-ins or corner kicks.
• By adding sensors like the one in the ball to every players shoes, we are able to apply that to offsides as well.
• Because of the automation of these things it’s pretty easy to make stoppage time computational. Instead of arbitrary adding minutes (they have rules but everybody interprets them differently), we get a countdown to the end like in basketball. Basketball can look boring for the amateur at first. But where it definitely shines are the last minutes when heroes are made. There is the term buzzer beater for it. Because of computer time Tracy McGrady could score 13 point in 33 seconds. Or the Raptors could catch up on a 30 point lead. I don’t see how such moments could be bad for any sport.

It’s far less ideal in case the boundaries are blurry. There are decisions that aren’t easy — even after the 5th camera perspective and slow motion.¹

• Humans (referees) decide what is e.g. a foul and what isn’t — on the fly just like in old times. Nobody outside the field is allowed to intervene.
• Borrowing from other sports, this time from American Football: Ruling “on the field” is the truth, until someone throws a flag on the field. In soccer I would get every coach one flag per half. The scene is question then gets to be reviewed with different angles and multiple referees. If the challenge is seen as correct, the challenging team gets the flag back and the ruling on the field is overruled. If not the opportunity to intervene is gone. This adds a whole tactical layer to the game: Do I dare to risk my flag or am I not sure?

I think this would be a much better implementation over the current one.

Soon everything will be tech-aided in some way or another. It’s important to keep in mind: Tech is neutral — what we do with it isn’t.

¹ Fair enough, that’s where AI comes into play. But you’ll see, AI systems are based on a training set. Someone has to tell the machine what the rules are, so that the machine can learn based on that. And that someone is naturally a human, who else. All humans are biased. When you tell me you are not, it’s the clearest sign that you are. So every bias will get replicated through the system.