First, a bit of good cheer. Election forecasts often get it wrong. On 8 November 2016, the day of the last US presidential election, the Princeton Election Consortium ('A first draft of electoral history. Since 2004') gave the probability of Hillary Clinton's winning – that is, winning the electoral college, not merely the popular vote – as 93 per cent. And the rest is history, or at least the alternative version of it in which Andrew Jackson nearly prevented the civil war, and Frederick Douglass lives on into his third century, doing an ‘amazing job’. Meanwhile, Princeton's first draft went through the shredder. So, when betting websites have the Conservatives at 1/33 on to win the most seats in the British general election next month – an implied probability of over 97 per cent – wizened heads can nod indulgently and note that whether there be prophecies, they shall fail.

Now, a lot of bad cheer. Nate Silver notes the prevalence of error in UK polls, but misinfers that Theresa May’s punt in calling the election is ‘riskier than it seems’. Opinion surveys do err, but in Britain they usually err on the side of understating the Tory performance. Polls responsible for big prognostic mufferoos – the elections of 1992 and 2015 – both had Labour doing better and the Conservatives doing worse than the hungover, ashes-on-the-tongue reality. And even in the elections that Labour won easily, under Blair, the polls consistently overrated the level of Labour support: by 6 per cent in 2001 and 2005, and 5.6 per cent in 1997.

The polls may persistently flatter Labour because of ‘shy Tories’; or it may be that pollsters, who often face very high non-co-operation rates, can't get a sufficiently representative sample and somehow contrive a less Tory-friendly demographic from whoever’s bored enough to pick up the phone. Thus caution is advisable over the factoid that Labour has ‘cut the Conservative lead by eight points’. Even taken at face value, the poll still has the party trailing the Tories by 16 per cent.

Labour’s draft manifesto, leaked yesterday, contains a lot of good policies, which the hard-right – that is, ‘mainstream’ – press will brand as ‘far left’ or ‘back to the 1970s’. I remember the 1970s, which were not that bad. You could go to college without having to rack up a debt burden that you’d carry into middle age. People still had the prospect of being able to buy a home, in London, say, without facing prices that had been bid up far beyond affordability by absentee plutocrats. Unions defended workers’ rights. Public assets were publicly owned. Labour proposes to take steps back in this direction – in other words, towards social democracy. The draft manifesto’s title, ‘Creating an economy that works for all’, even references Theresa May’s mood-muzak burble outside Number 10 when she took over last summer.

Unfortunately none of this will be enough to swing the election. It’s not that most people are likely to reject Jeremy Corbyn because they don't like Labour policies. They are likely to reject Labour policies because they don't like Corbyn – or at least, have no confidence in him. In part this is because of the demonisation directed at him since he was elected leader in 2015 – today’s Telegraphhas Corbyn’s car running over a BBC cameraman’s foot. In part it is down to Blairite revanchisme. But it is also down to honest political incompetence on Corbyn’s part.