Why Data Journalism Can’t Always Be Trusted
Data can tell us two stories – the one its creators want to tell us and the one they’d prefer to ignore. So when a politician boasts that, for instance, 70 per cent of business leaders say the economy is better than it was 10 years ago when the last government was in power, it ignores the fact that 73 per cent say that it’s worse off than it was last year.
I’m making it up of course. Yet that is sometimes what data allows us to do. It points not to one answer but many. People – typically economists and politicians – wield it as if it was a sword of truth when in fact it is a double-edged weapon. Proper analysis reveals several, sometimes contradictory, ‘truths’.
Sometimes the most irresponsible interpreters of data are journalists. I speak from experience. They like to pounce on a figure and spin it for maximum impact, ignoring anything that might get in the way of a good story.
Thus today those who like to bash the BBC at every opportunity – principally The Times and Daily Mail – and those who seek to redress every perceived gender imbalance – principally The Guardian – have latched on to an absurd figure that suggests the broadcaster prefers men’s tennis to women’s.
I say absurd not because the data-based allegation is untrue but because it doesn’t tell the whole story. So during the first week of Wimbledon, a keen-eyed tennis fan kept a tally of the BBC’s coverage and found that 75 per cent of it was devoted to the best-of-five set men’s tennis game. On one day, 93 per cent of coverage was of the male game rather than the best-of-three set women’s matches. The BBC covered only 48 per cent of Serena Williams’s first two matches, compared with 100 per cent of Andy Murray, Novak Djokovic, Roger Federer and Stan Wawrinka’s first two matches.
Gender bias, scream the journalists! The BBC meanwhile responded: ‘Whilst we value all feedback, all decisions around which games will be shown are based on editorial judgement.’
And so they should be, because this is what the data also shows. In the 64 first-round women’s matches, almost three quarters (45 of them) were straight sets victories and 29 matches (almost a half) were walkovers in that the opponent failed to win six games.
In the men’s game, only half of the 64 first-round matches were straight sets victories and only four matches saw the opponent win less than nine games.
So by my extremely rudimentary and wholly unscientific analysis of the data (I’m a journalist by trade so I’m good at that), most of the women’s matches were one-sided and lacking in competitive entertainment. One sixteenth of the equivalent men’s matches were walkovers.
Serena’s first two matches lasted less than an hour and were both 6-4 6-1 victories. Novak won slightly more tense games 6-4 6-4 6-4 and 6-4 6-2 6-3.
And here’s the point. When it comes to sport on television, a broadcaster should be able to show matches that are deemed exciting and competitive. That’s what viewers want and, unfortunately, many women’s first and second round matches are simply not very competitive. They are undoubtedly brilliant players and are paid just as much as the men for their efforts, despite playing far fewer games. First round losers picked up £29,000 last year and will take home an extra £1,000 in 2016.
But broadcasting decisions should not be guided by arguments about gender bias and certainly not by disgruntled couch potatoes armed only with a stopwatch and a gripe.