Notice of interruption to service

When this blog reached its five-year anniversary in 2018, I wrote a summary of the conclusions it had reached so far, and I vaguely thought at the time that, should it run for another five years, that might be a good time for a pause. Now we’ve reached that point – Ten Minutes Past Deadline is 10 years old this week! – and it does seem to be the moment to take stock.

Looking back at the five-year anniversary post, I discover that the blog still essentially agrees with itself in its attitude to the importance of editing, the complexities of online news as it expands into the anglosphere, and the nuanced importance of the role of formal English. The standard of mathematics in newsrooms has not improved over the past five years, and the corrections columns remain as embarrassing to people in our profession as ever. In fact, this is the problem: the blog has settled, as blogs tend to do, into a series of themes, and for a while now has been incrementally exploring them, rather than breaking new ground. It is, perhaps, getting a little repetitive.

So it’s time for some major mental engineering work: tracks of thought will need to be pulled up, sleepers will need to be replaced and some much-needed intellectual ballast laid down. Ten Minutes Past Deadline is not closing – blogs never really close – but the pace of updates will be slower, and motivated more by new thoughts, when they come along, rather than the rehearsal of old ones. Hopefully the work will result, like the replacement escalators at South Kensington, in a less juddery experience for customers, and hopefully will not take as long as that project seemed to.

And, as was the case five years ago, I remain always grateful for the blog’s readers. The visits, the engagement, the comments and the retweets are what make blogging worthwhile, and the content here has always been greatly enhanced by the contributions of others. I hope that we will be back soon with more. Until then, tickets remain valid via all reasonable routes and we would like to apologise for any inconvenience this may cause to your journey.

Zeroes and ones, part 6C (43F)

Journalism and maths – the adventure continues:

Although in fact, here the issue is not really numeracy: it’s a more abstract one of conceptualisation – confusing a temperature of 2.5C with a difference in temperature of 2.5C. A sum has been calculated correctly; it’s just not the appropriate sum. (If global temperatures rise by 36F, even air conditioning isn’t going to help.)

And I’m afraid, at the Tribune, this has happened quite often:

It’s not clear why we struggle with temperature like this when we navigate other conversions successfully and, as has been said elsewhere, sub-editors are capable of making much finer distinctions than this when it comes to language. I think it may have something to do with the fact that the zeroes on the two scales are so far apart and signify different things (whereas, for example, 0 mph and 0 kph signify the same thing, and the scales diverge after starting at a common point).

The classic formula for converting an actual temperature in celsius to fahrenheit is

(<temp in C> x 1.8) + 32 = <temp in F>

and it’s that addition of the constant, the 32, that causes the trouble when you are trying to calculate a difference in temperature. If you subtract 32 from all the erring totals (where given) in the corrections above, you get the right answer (or close to it, given some of the original fahrenheit totals have been rounded). So the correct C to F conversion for a difference in temperature is simply

<temp difference in C> x 1.8 = <temp difference in F>

and similarly the other way, for a difference in temperature:

<temp difference in F> x 0.5555 = <temp difference in C>

In such a critical decade for climate change policy, we may find ourselves needing to do these sums more and more often.

With thanks to the Tribune’s chief revise sub for spotting this one – a man who has seen too many improbable-looking pound-to-yen conversions (really, that many zeroes?) to let any figure in parentheses pass unscrutinised.

Zeroes and ones, part five

This week on Journalists’ Adventures in Maths: percentage changes are not reversible, or why a 75% rise from 4 to 7 is not a 75% fall from 7 to 4.

Dutifully checking the numbers* in the copy as they come up, I first get a result at variance with the reporter’s:

and then, by swapping the numbers around, get one that agrees:

However, the minus sign at the start of the second answer is the clue: the numbers in that sum are declining from over 400,000 to less than 300,000. But the copy talks about a rise.

The same thing happens with the second pair of numbers: the percentage rise is calculated as though it were a percentage fall.

Why is it not the same? The difference between 8,276 and 12,092 is always constant: 3,816. But in a percentage, of course, you can relate that constant difference to different comparators, and 3,816 is a much larger proportion of 8,276 than it is of 12,092.

This will hardly come as news to people who can do maths. But for arts-heavy newsrooms, this is slightly more perilous territory than the answer just being wrong – it is wrong, but it seems right if you do the sum the wrong way round. You need the strength of mind to remember which number you’re starting with and stick with it. It seems somewhat analogous to the evergreen error of mistaking ancestors for descendants, or confusing “overestimating” and “underestimating”. It’s not just the relationship between the two things that’s important, but the direction of travel too.

*Use of a percentage calculator is highly recommended, of course; I like this one, with its clear, question-based approach.

Box-office figures

Thanks, Google. Actually, I was just checking titles and release dates of films for a piece about how the sex scene is dying out in cinemas. But thanks.

I’m quite impressed that it even saw a sum it could calculate in that search. It’s not like I was looking for the French arthouse romance 5×2. Thank goodness I didn’t need to check From Here To Eternity.