top of page

Sideways Looks #7: Invite to a Data Party, also Space

Updated: Aug 1, 2021

Hello you,

Sideways Looks #7 here – including the long-trailed 🎉 🖥️ Interactive Election Manifesto Dashboard 🖥️ 🍾. Though don’t get too excited just yet – read on to find out why. Also a great fact about humans in outer space. Last week I accidentally experimented with sending Sideways Look on a Sunday. Result: it got more engagement than usual, but spread out over a longer period of time. So now I’m trying Saturday. For those interested, the best-performing of these newsletters is still the first one. Maybe it’s because it was the first; maybe it’s because it mentioned Bears. Here’s the full list with % of recipients who opened each one:

I hope you all have good weekends. Yesterday the EU confirmed the UK’s ‘data adequacy’ decision, so congrats to my old team in DCMS. I hope everyone reading this manages a bit of data-related celebration. However your chosen European team is doing at football. Oliver Thought for the Week: Making Jokes, Telling Stories, Playing with Data I was first hired into government as a “Data Journalist”. I think I was the only person with that job title in the Civil Service; it sometimes caused panic when people thought an actual journalist was in their meeting. My role was to bridge the ‘technical data people’ and the ‘communications people’. I originally expected this to be a delivery job, ferrying messages between different sides. Actually, it was more about being part of a team that looked at data and created a story together. I find the oft-drawn distinctions between ‘technical’ and ‘creative’ people fascinating. Many people talk of technical prowess as completely incomprehensible. But I’d say that applies to both sides of this apparent divide – great poets and comedians make you think ‘how on earth did they think of that??’ much like genius mathematicians or coders. For the onlooker it all looks a bit like magic. But I'm less interested in these extremes. I'm more interested in how we all weave creative skill into everyday life. We tell stories without being novelists. We make jokes without being comedians. We play roles in conversations without being actors, and imagine scenes without being playwrights. I think this is more exciting than singular shows of skill; conversations where only one person talks usually become unengaging, even if the person is a great storyteller. Humans together can be creative for hours without being conscious we’re doing so. Which is amazing. This is where ‘data skills’ can start to feel a bit left out. Apart from dropping occasional stats we’ve read in a newspaper, it's rare to see people sharing ‘amateur data analysis’ the way we collaboratively create stories or jokes. Some might say most people don’t have the ability; but again, I'd say jokes are hard too. I wonder if it’s more an education and confidence thing. Worth noting that not too long ago, assuming most people in a conversation could read and write would have been very optimistic. All this is a preamble to a big reveal: that interactive dashboard of UK general election manifesto topics that I promised in Sideways Looks #5. More detail in that newsletter, but to summarise – a computer has read manifestos, and tried to cluster together related words into ‘topics’. The lists at the bottom show the topics, the bar charts show how strongly each topic features in the Labour and Conservative manifestos for each year. Here it is. Have a play around. Click some bars, or one of the wordlists, or the little tick-boxes, and see what happens. (To undo, press ‘esc’ or click item a second time). You could, for example, filter to just the 1980s and the 2010s to really bring out the differences over time; or compare manifestos that are considered more politically extreme (e.g. Labour 1983 & 2019, Conservative 2005) with those deemed more moderate (Labour 1997, Conservative 2010). Or whatever. Have fun. (Manifesto text was taken from the excellent Manifesto Project database; 2010 is currently missing, but I may try and rectify that). The dashboard is not finished; nor is it even good. There's two reasons. The first, boringly, is that I’m still learning to use the tools. But I strongly believe that, in general, you shouldn’t hold things back until you think they’re good. Otherwise you can spend ages creating something that you think others will find good, without ever confirming your thoughts. The second reason why it’s not (yet) good is more interesting. Sometimes you want to do something creative, but – to borrow a sociology-of-science phrase – the world pushes back*. Sometimes your jokes fall flat and you don’t know why. Sometimes you just can’t think of anything interesting to say. And sometimes your topic modelling programme spits out topics which just look like jumbles of words:

That’s just one example, but basically all the topics in the dashboard are weird hodgepodges of general political words. I’ve played around with the inputs to see if I could get nice clear clusters around ‘education’, ‘health’, ‘economy’ and the like. It just gave me more jumbles. Maybe if you play around hard enough you’ll spot some things. For instance, by comparing to 1980s to the 2010s, I noticed that maybe the topics reflect changes in general political language – e.g. maybe nowadays both parties say “enterprise” less and "business" more, or mention "inflation" less than in the ‘80s. It’s an intriguing thought, though a bit of a leap from the available data. But to go back to our original point – this isn’t supposed to be rigorous, professional-level data analysis. It’s like a joke that I wouldn’t put in a standup routine, or a story I wouldn’t make into a novel. It’s supposed to be playful and open and engaging. It's limited right now, like a conversation with a limited list of words. But over time I’ll keep adding more things, so that playing around can produce more things to talk about. It's a data party that I hope you’ll all want to join. * Short explainer of 'why sociologists of science use the phrase 'the world pushes back'. Science isn’t just neutral observation. Turning observations into accepted facts involves all sorts of human interpretations and social factors. Advocates of this view – known as ‘social constructionists’ – have been accused of saying anything can be a fact if society says it is. But that’s a caricature; many social constructionists also point out it’s hard to make something a fact if 'the world pushes back'. An example - for centuries various societies told fanciful stories about humans flying, but gravity stopped that being an actual fact. Stories didn't switch off gravity. Eventually, society pushed back hard enough and invented the plane. Now ‘humans can fly’ is a fact – but the world didn’t make it easy for us. Interesting, if quite academic and expletive-laden, article on that here. Fun Fact about: Humans In Space Short but great fact: there has been at least one human on the International Space Station since 1 November 2020. To put it another way – the last time all humans were together on Earth was over 20 years ago. Recommendations The CogX festival claims to “gather the brightest minds in business, government and technology to celebrate innovation”. I don’t like people being described as “the brightest”– intellectual ability is extremely diverse, not a simple scale. And even if it was, the idea the "brightest people" would all be attending your one festival seems unlikely. BUT: I saw that many folks I respect were attending; and it was virtual and free so I logged in for the final day. And I’m very glad I did; the sessions mixed perspectives from academics, businesspeople, technologists and many others into some really wide-ranging discussions about technology and society. Recordings can be found here. With all the politics around nowadays, my first port of call has become the New Statesman Podcast. The hosts have the political insider knowledge common to a lot of journalists; but they also really engage with evidence (whether that's polling data, history, or testimonies from people on the receiving end of e.g. Universal Credit) to critically examine received Westminster wisdom. Also one of the hosts - Stephen Bush - has a meme page dedicated to him. Finally, job opportunities. Helpful Digital is a consultancy for a range of digital skills, including crisis simulations. The managing director is Tim Lloyd, who delivered some really interesting sessions as part of the Accelerate course for Civil Servants that I helped design (and is an all-round lovely guy). Anecdotally, I feel we’ve reached a stage of unlocking where people have started looking for new job opportunities – so if you know someone with good digital communication chops, please do pass these two roles on.

19 views0 comments

Recent Posts

See All


bottom of page