By the time you read this, the Australian Open will be over for another year, both victors and vanquished away with their spoils. Obviously I don't know who won, because there's still a few days left as I write this. This lag between what you know of the tennis and what I know is, of course, courtesy of the dynamic medium of print.
These days tennis coverage is much more up-to-the-moment and interactive. IBM has a network set up underneath the tennis centre in Melbourne (which I got to tour at the beginning of the tournament). At one end is the nerve centre, at which the match statistics are packaged, repackaged and re-purposed for everything from the graphics on TV to kiosks around Melbourne, with the Web site, Java applets and WAP thrown in.
And it's all hooked together in a server room with kilometre after kilometre of cable headed in all directions. Those cables lead out to every one of the couple-dozen courts, where they're plugged into ThinkPads operated by volunteers who enter the result of each point.
All these masses of expensive technology, and it all comes down to some volunteer with a ThinkPad. The response time between a point being finalised and the result appearing on TV screens is three seconds - which means there's much less time than that for the human volunteer to decide how the point was won.
I asked how much training went into helping these people judge the difference, say, between a forced error and an unforced one. The answer came back that, if the ball hits the racket and goes backwards, it's a forced error. If it hits the racket and goes forwards, it's unforced.
Imagine for a moment that you're Pete Sampras. You lay down perhaps the most perfect serve ever in the history of tennis. By rights it's an ace. It's unplayable, but the spoilsport at the other end manages to touch a racket to it. What's more, by some accident of physics, he touches his racket to it at such an angle that it bounces straight into the ground, and the spin carries it back towards the net. Your beautiful ace just became his "unforced" error.
After I'd done the tour, I watched a match on TV, making note of how I thought the points were won. I changed my mind several times on seeing the replay. The human on court has no replay. The human on court has less than three seconds. My stats at the end of the match were quite different to theirs.
IBM is hopefully working on some alternative to humans. Until then, at least you can rely on slow, old print.
Matthew JC. Powell likes people, really. Commune with firstname.lastname@example.org