SQL Sequels

 

I applied for what looked like a dream job recently. Down in Dorset, about ten miles from a town I visit every couple of months, a place I know and love, where I have friends and people to say hello to in the street. I’ve been going there since school trips, back when we didn’t so much find dinosaur fossils as dodge them lumbering up out of the primeval ocean. Kidding – we just chatted with old Tom Hardy and gave him some plot ideas. Turned out he only ever used one of them, really.

Not Your Heart Away

Immediately after A Levels I’d bizarrely got into the habit of working and banged-out 82,000 words that became A Day For Pyjamas. Half our lives away I wrote a sequel to it, Not Your Heart Away, which got some nice reviews on Amazon and won a BBC award when I re-wrote it as a screenplay. Another one (Janni Schenck, a story about a nice orphan kid who loves Swing music and just happened to be in the Hitler Youth) was based on fact; that one was entered for Cannes with the Maison des Scenaristes, back in 2017.  I wrote some stand-up performance spoken word and performed them a bit, but got tired of the don’t-get-paid local stand-up circuit. So yes, stories. I can do that. Gizzajob, as we used to say.

I’d also spent two decades analysing business information, from going around the world listening to people to poring over data tables with my lucky ruler (metal, triangular, I think they’re really something to do with woodwork, but hey) to reading hundreds of pages of interview transcripts and spreadsheet print-outs and finding the tiny clues that open-up markets. Which I did more than once. So yes, I can do data. What do you need?

What the committee decided they needed, after a blistering presentation that I could feel in the soles of my feet had rocked everyone’s socks off, was someone who can do that and was a ‘data expert’ at SQL and Python and Tableau. I can do that too, but those are computer programmes. I’ve nothing against computer programmes. I’ve designed and got written two apps myself, one to track HTML 404 errors, the other to compare and assess casualty data. Is that tech enough?

Apparently not. But they’re still two different skills. One you can learn in a month. Sit me or let’s be honest, pretty much anyone down in front of a PC and an online learning course and you’ve got a competent machine-minder. They won’t necessarily know anything about the data they’re interrogating and most of them won’t worry their pretty little heads over the fact that using data this way is akin to regression modelling; it’s great if you know for certain the future is going to be exactly like the past. Which is a pretty massively flawed assumption in lots of areas of life, not least the national economy after Brexit.

You can’t learn to tell stories convincingly and well to an audience of one or a couple of hundred, online, on TV, at conferences or anywhere else, in a month.  Like any performance, it’s an iterative thing. You get better at it the more you do it. You learn from your mistakes. You might even integrate them into the performance to get a reaction to work back to from the people you’re telling the story to. It’s a two-way thing. It’s adaptive. It’s interpretive. And whatever you do it, however much you call it ‘science’ data crunching isn’t, any more than  power loom operators were weavers. It’s reductive, it’s literally codified; it’s not even about understanding numbers and their relationships. Just learning how to get a machine to tell you ‘how many.’ Never how, and don’t even ask why.

 

 

Share Button