Wednesday, July 9, 2014

We've got brackets!

I've just been speaking to Bazza on the phone. He hasn't changed. We hadn't spoken in almost a frigging year, and he just launches straight into the men's Wimbledon final. Heck, I like tennis and have blogged about it here a lot, but a how-are-you or a fancy-you-ringing-me-it's-been-ages would have been nice, if a huge surprise. I plan to see him on Sunday week when I'm up in Auckland.

When I was a kid, at about quarter to five on a Saturday, you'd see the football results trickling through on the TV, one letter or number at a time, via something called the vidiprinter. If there was a big score, they'd also put it in words in brackets, just so you'd know it wasn't a typo, as in "Oxford 1 Birmingham 7 (seven)". I'd get a bit excited whenever I saw brackets. Well this morning we had brackets all right.

Seriously, who’d have thought it? I actually wanted the Germans to win (yes, I know, English and all that) but with two-thirds of the game still to play, Germany had a scarcely believable five-goal lead! Someone at work managed to get a live stream, and at half-time I saw a replay of the goals. They were a result of abject defending more than attacking brilliance, and it seemed that for all the talk about Neymar, Brazil missed Silva (their best defender) even more. But 7-1! And it really should have been 8-0. The look on the Germans' faces when they gave up that last-minute goal showed what a determined bunch they are.

I wondered last night what the legacy of this World Cup would be, and barring something truly ridiculous in the other semi or the final, we’ve got our answer. Even if Germany lose in the final. As shock results in football go, this one is a magnitude 9.5. Kids with parents wealthy enough to get them tickets to the game, will relive it 60 or 70 years from now. In some ways I’m glad Brazil were beaten. If they’d lifted the trophy, it would have glossed over the obscene amounts of public money spent on the tournament.

Nate Silver and his mates at FiveThirtyEight.com have come in for some criticism for their World Cup prediction model following this crazy result. Some of it is deserved. If you’re going to give Brazil nearly a 50% chance in a four-horse race, in a sport as unpredictable as football, they need to be pretty damn good. Better than they’d been in their previous five games, which I don’t think were weighted heavily enough in the model. But all this talk about all statistics in football being meaningless is pure guff. Of course you can model football results, just like you can model election results. You just need to get your inputs right.
Edit: I now see that the FiveThirtyEight model gave a Brazil a 45% chance of lifting the trophy before the tournament started. Yikes. That was way off base, surely. This isn't rugby where one side (like the All Blacks) can dominate. That they even came up with a probability that high and didn't seriously question it makes me wonder.
 

Here’s Joe Bennett’s take on World Cups (he wrote it before today’s match). Very accurate I think.

No comments:

Post a Comment