Pre-race interviews are tricky things. Though regularly scheduled at most road races and track meets, I’ve always found them to be of limited use. Maybe I’m just a poor interrogator, but I’ve always likened the experience to a trip to the dentist, with me being the dentist and the athletes in the role of reluctant patients as I try to pull teeth, or in their case information out. Inevitably, nobody says jack to give their condition away, which is totally understandable, though on rare occasions you can elicit a telling story to share with readers or a TV audience.
In general, however, a pre-race interview is a ritual with both sides accepting the other as readily as two eighth-graders at a fortnightly dance. And yet it remains one of the only means of reading an athlete’s tea-leaves to determine their readiness and frame of mind beyond traveling to their training camp and observing their workouts first hand.
But in the not too distant future, the pre-race TV interview may come to be one of the most important tools in the sport, the means of screening for drug cheats.
A new artificial-intelligence program has been developed by a team of computer scientists at the University of Rochester in New York that has compiled the largest public data resource of facial expressions that indicate when people are lying. It’s like a new polygraph/lie-detector.
The program has developed a machine-learning algorithm that scanned over 1.3 million frames of one-to-one interactions with some 300 people taken during questioning and discovered the facial ticks that indicate when someone is not telling the truth. Seems the most distinctive facial giveaways of lying show up around the mouth and eyes involving what’s known as the Duchenne smile, an involuntary cheek muscle that we can’t control.
Imagine. Here the sport is spending millions of dollars sending unannounced testers to athletes homes all over the world at odd hours to draw blood and watch them pee trying to discover who the real dicks in the sport are. And yet soon they may only need to plop these folks down in front of a camera and ask, “So, you taking something there, Gingah? You on the juice? Don’t turn away! Look into that lens and tell me straight!”
All sports, but especially athletics, have long sought a fool-proof method of screening for drug use, as PEDs have contributed to both the corruption of the game and the decline of public interest in the integrity of the sport. (If only that were the only problem the sport had.)
If this new facial-recognition technology is successfully proven and perhaps accepted in courtrooms – an acceptance that the current lie-detector machines never achieved – this could be the closest the sport might come to a fail-safe mechanism.
Then, the only remaining line of defense for cheaters against the new system would be the George Costanza method, which he reminded Jerry of in one Seinfeld episode when Jerry was about to take a polygraph test. “Remember, Jerry, it’s not a lie if you believe it.”