The best laid schemes

The best laid schemes o’ Mice an’ Men Gang aft agley, An’ lea’e us…

So, How Will You Do It Better?

By 2353NMIn the next 12 to 18 months there are a number…

A personal view of a failed democracy

I sometimes wonder if it is in becoming a parent, responsible for…

Assange’s Fourteenth Day at the Old Bailey: Elections,…

September 25. Central Criminal Court, London.On this Friday, the Assange trial moved…

Reverse JobKeeper cuts and protect working people, say…

Two days after assailing the Morrison government for inaction on a jobs…

Dean Jones: Life of a Cricketing Entertainer

He was very much one of those cricketers who made the pulse…

Seeking the Post-Corona Sunshine: More Continuity for Post-Fitzgerald…

By Denis Bright  Even Nostradamus himself might be hesitant about predicting the final…

Jobs plan must be a way forward, says…

On the day where the Morrison government’s announced cuts to the JobKeeper…

«
»
Facebook

Accurate predictions for election result, Melbourne Cup and the stock market

Image by thevarguy.com

Image by thevarguy.com

More than 90% of everything is predictable; most predictions are wrong.

How do I reconcile those two things. Well, quite simply, people don’t bother to make predictions about the predictable things, and if they do, the prediction won’t get much oxygen. If I predict that Monday will be followed by Tuesday, then no-one’s really going to be interested. Nor, when I tell people that this years Melbourne Cup will be run on the first Tuesday in November and won be a jockey riding a horse, is anyone likely to contact me for my amazing prognostications.

To raise interest, predictions have to be outrageous and unexpected, which is why they’re so often wrong. If I predict that Turnbull will challenge Abbott for the leadership, then people will want to know when and what basis I have for my belief. Unless, of course, I say something like next Thursday week at 3pm. And happen to be right.

Philip Tetlock did a study on expert prediction, and, apart from finding that they were little better than a monkey with a dartboard, he concluded that he could divide them into “foxes” and “hedgehogs”. The “hedgehogs” were good at one thing, and they knew they were right. Their predictions tended to be specific and clear. (“The GFC is far from over – the market will crash again in 2011, 2012 at the latest.”) The “foxes” were able to consider a number of things and couched their predictions in generalities and qualifers. They could take into account a number of possible scenarios, using “if/then” phrases.

An example of a “hedgehog” would be someone like Andrew Bolt or Tim Flannery. (Some of Flannery’s statements are far too specific and don’t help in the attempt to educate the general public on the difference between weather and climate.) I can’t think of a good example of a “fox” because generally they don’t get much air time. They don’t make for good headlines, so who wants to talk to them?Might as well talk to me about my prediction for the Melbourne Cup. (“Don’t forget I told you that it would be on a Tuesday and won by a horse.”)

The only trouble is that the “foxes” are the ones who actually frame the discussion in terms of intelligent questions. And they have more success in their actual predictions than the “hedgehogs”. The interesting thing is that success doesn’t seem to matter. People who get things wrong over and over again are still asked for their thoughts in the media.

How is this possible? Well, it’s easy to explain away why you were wrong. You can say that your timing was out, but that what you predict will still happen. (“I know I said 2012 for the stock market crash, but because of the way Obama has propped up the economy, he’s delayed the inevitable.”) Or you can cling to the part of your prediction that was correct. (“I know that I said that the Cup would be won by an imported horse, well, the winner’s sire was imported, so I was on the right track.”)

So, I’m tempted to go out on a limb here and to be a “hedgehog” and say that Labor has this election in the bag now. Rudd’s return will throw Abbott out of stride, and the pressure will get to him, leading to some Liberals speculating privately about whether it’s too late to go back to Turnbull. Of course, I don’t actually believe that, but it’d sound more impressive than what I actually think will happen. I think Abbott may well be rattled. He’s been cruising to a victory, but the latest polls make it close. And, just like a sporting team that’s given up a large lead, they often try to hang on, change their strategy and end up choking, there’s a genuine possibility that Abbott will repeat his: “Of course, I read the report” fiasco.

But I’m more circumspect than that. I’m going to predict – with certainity – that what happens now is uncertain. There are so many variables going into the election that only a “hedgehog” would try to call it. Rudd has taken the wind out of the sails of the simplistic “Juliar” campaigners. No-one will accuse him of lying, in spite of his promise not to challenge. But he does come with his own baggage. And if you go on any social media, you’ll be able to find disaffected Gillard supporters who swear that they won’t vote Labor now. What happens when it becomes a choice between Rudd and Abbott, or when they actually consider voting for the Opposition candidate in their electorate is anyone’s guess. If the Liberals actually start trying to articulate their policies, will it turn voters off? If they try to attack Rudd, in the same way they attacked Gillard, will it just make them look negative? Will Katter’s party affect how the Liberals go in Queensland? Will Palmer have any effect? Could it be another hung Parliament with Abbott having to negotiate with Katter?

Like I said, only a “hedgehog” would be definite about the coming months.

As for the Melbourne Cup, that’s easy – take Bart’s horse and the French one with your grandmother’s tip for the trifecta!

 

Leave a Reply

Your email address will not be published. Required fields are marked *

Return to home page
Scroll Up
%d bloggers like this: