It’s time for pundits to put some skin in the prediction game

Ed Smith's "Left Field" column.

The US election was nowhere near as close as many pundits suggested. Photograph: Getty Images

You would expect a column called “Left Field” to challenge conventional wisdom. So let’s say goodbye to 2012 by recalling some moments – of wildly diverging importance, drawn from politics, sport and health – when we’ve been misled by the mainstream.

First, on the evidence of the US presidential election, it is time to ditch the pundits and follow the bookies. Right up until polling day, there was bizarre agreement across the mainstream media that the election was “too close to call”. Paddy Power, Europe’s largest bookmaker, knew better. It paid out on a win for Barack Obama two days before polling.

It would be easy to list all the Republican shock jocks whose electoral college projections missed the dartboard, let alone the bull’s eye. Yet in this instance, we might absolve right-wing ideologues from blame. At least they can blame woeful analysis on wishful thinking. When Karl Rove was holding out for a Mitt Romney win, the Fox News anchor Megyn Kelly asked him (with more honesty than she had perhaps intended), “Is this just math you do as a Republican to make yourself feel better or is it real?”

Crying wolf

It wasn’t just the Republicans. How can we explain the determination of the allegedly nonaligned mainstream media, in Britain as well as in the US, to pretend that the election was close, even in the last couple of days? Why were the experts, even those who didn’t like Romney, so reluctant to acknowledge what state-by-state polls had told us for weeks?

Partly, it was because pundits are reluctant to admit that opinions polls are now much more accurate and reliable than they were before. Mistakes such as what happened in the 1992 British general election, when polls predicted a Labour victory, are far rarer. Second, the narrative of electoral reporting relies on the elevation of inside sources. If a prediction, even a wildly inaccurate one, comes from “inside” the Obama or Romney camp, it gains specious respectability – the problem being that both camps had reasons to argue that the election was close. Romney’s people preferred the idea that it was close to the idea that it was over. The Obama camp, meanwhile, was worried that complacency might lead to a low turnout. So the “inside view”, from both sides, was credulously (and wrongly) passed on as revealed truth.

However, the strongest factor is the love of drama. It is hard to avoid the conclusion that the media pretended the election was close to make it more “interesting”. That is a risky strategy, to say the least. It is short-term attention-grabbing at the expense of long-term trust. It is a form of crying wolf. It leaves viewers and readers that little bit less likely to pay attention next time.

There is a further irony about the election. The accurate predictions of the statistician Nate Silver on his New York Times blog FiveThirtyEight were hailed as a victory for scientific rigour over gut instinct and hunches. It is true that Silver had a great election. Around polling day, “FiveThirtyEight” was the most searched-for term that led people to the New York Times website. I, too, was part of the crowd: all my visits to the site were to access Silver’s blog.

For all their sophistication, Silver’s projections merely reflected the odds widely on offer from bookmakers. The lesson to draw fromthe woeful analysis of the election is not that “science works”. It is that we shouldn’t listen to people without skin in the game. If they’d been forced to bet on their projections, how many pundits would still have insisted that the election was too close to call? Maybe a journalistic warning should appear alongside every “predictive” column or piece of television punditry: “Do not trust my words: I have nothing at stake.” Conversely, those who are prepared to stake, say, £500 on their predictions would be rewarded with a special asterisk. Perhaps this skin-in-the-game asterisk could replace the much discussed “Leveson Kitemark” as proof of respectable journalism.

Silver learned his trade analysing baseball statistics before transferring his skills into the political arena. Not all sports punditry is prepared to join the real world. London hosted a wonderful Olympics this summer but it is in danger of being interpreted into absurdity. According to the Guardian, the shortlist for the BBC’s Sports Personality of the Year, which inevitably had a strong Olympic flavour, showed “a depth of excellence that is emphasised by the meritocratic diversity on show”. Put differently: we ought to love the Olympics and hate football because it is tainted by big business and commercialism.

Fat chance

Being uncommercial does not make a sport meritocratic. Quite the reverse. Thirty-seven per cent of Team GB medallists were privately educated. If there was a country called “British independent schools”, it would have come 12th in the overall medal table.

You might prefer sailing to football. You may prefer the way less flashy athletes conduct themselves; you may celebrate their restraint and dignity. But don’t pretend 2012 was the year when elite sport was finally opened up to the masses. If you really like the idea of sporting meritocracy – a broad talent base narrowing to an open elite – then I have bad news for you: it’s time to celebrate the Premier League.

Finally, as you reach for another mince pie this Christmas, spare a thought for Denmark’s legislators. They have had to make a U-turn and scrap the tax on fatty foods that they introduced in 2011. The problem was that Danes, desperate for a sudden injection of lard, were going on fat cruises to Germany, just as English drinkers used to cross the Channel Tunnel on booze cruises to France.

There is, sadly, a simpler problem with taxing fat. Fat doesn’t make you fat. Sugar does. The word “fat” has a PR problem. It needs a new spin doctor. Unfortunately, fat (the food type) has to share a linguistic term with fat (the body shape). There is little correlation between the two. Sugar, on the other hand, certainly does make people fat, as well as causing diabetes and a daily cycle of energy booms and busts.

So leave the skin on the turkey but pass on the Quality Street. Happy Christmas.