Individual portfolio: John Burn-Murdoch

Individual portfolio: John Burn-Murdoch

Organisation: Financial Times (United Kingdom)

Publication Date: 04/10/2017

Applicant(s)

Size of team/newsroom:large

Description

2016/17 was an eventful year for data journalism, wth a number of major events either affirming or shattering our belief in numbers. As a senior data-visualisation journalist at the Financial Times I have used a variety of data-driven approaches on a range of projects over the last year: from analyses and visualisations of the forces that drove first Brexit and then Trump's victory, to a visual explanation the record-breaking 2016 El Niño, through visual histories of tennis and the summer Olympics, to a prediction model for the outcome of the English Premier League. In these projects I have put into action a wide range of journalistic, analytical and design skills: generating ideas, scraping, cleaning and analysing data, building robust prediction models, writing stories and producing a mix of static, animated and interactive visualisations. The following are ten examples of my work to be considered for the Individual Portfolio category. In each case I was responsible for all elements of the story unless otherwise stated: 1/ Detailed statistical analysis of the driving forces behind the Leave campaign's victory in the EU Referendum, carried out and published on the day of the results https://www.ft.com/content/1ce1a720-ce94-3c32-a689-8d2356388a1f 2/ Additional follow-up analysis of turnout in the EU Referendum, as well as modelling the different results that could have arisen from alternative turnout scenarios https://www.ft.com/content/2994afe4-43b0-3763-a80c-92351473496f 3/ A visual exploration of how the US electorate shifts back and forth from Democrat to Republican in swathes of blue and red every eight years https://www.ft.com/content/5ccc4c8c-aa5c-11e6-809d-c9f98a0cf216 4/ Detailed statistical analysis and theoretical discussion of the complex relationship between education levels, value systems and populist support in the US and UK https://www.ft.com/content/9fc71e40-b015-11e6-a37c-f4a01f1b0fa1 5/ A visual, research-led illustration of the mechanism that powers the El Niño weather pattern, and its impact on food production around the world https://www.ft.com/content/7f007de6-bba5-3ffd-ad04-52f3a51b31db 6/ An analysis of how the stars of the 2016 summer Olympians compare to the most decorated athletes of all time https://www.ft.com/content/5f353a1c-6928-11e6-a0b1-d87a9fea034f 7/ A data-driven animation of the remarkable feats of soccer goalkeeper Diego Alves https://twitter.com/jburnmurdoch/status/842681672797569024 8/ A Visual History of Women's Tennis, celebrating Serena Williams' unprecedented success https://ig.ft.com/sites/visual-history-of-womens-tennis/ 9/ A comparison of Leicester City's astonishing run to win England's Premier League title against similar efforts spanning 130 years https://ig.ft.com/sites/leicester-premier-league-champions/ 10/ Predicting the outcome of the 2016 Premier League table, using a robust and finely tuned statistical model https://www.ft.com/prem-predict

What makes this project innovative? What was its impact?

Working on the FT's same-day analysis of the EU referendum gave me a wonderful opportunity to live out the data journalist's motto of "social science done on deadline". I carried out a multiple regression analysis on more than 100 different demographic and socio-economic variables, which allowed the rapid identification of key narratives in the ensuing conversation "Brexit", such as the identification of education and age as key dividing lines among the electorate. This allowed us to publish striking and informative graphics early on the day as results were still coming in, leading to some of our most high-impact social media posts ever at that time [https://twitter.com/ft/status/746275255354818561?lang=en]. As a direct result, I was invited onto the UK's Sky News television channel to be interviewed about our analysis. Subsequent work with regression models and other moderate-to-advanced statistics allowed the FT to remain at the forefront of the conversation on the common themes -- and differences -- between the Leave campaign's victory and Donald Trump's win in the US [https://www.ft.com/content/9fc71e40-b015-11e6-a37c-f4a01f1b0fa1], and my work developing best-practices for regression in the FT's newsroom continues to underpin much of our coverage of the European elections taking place in 2017 [https://www.ft.com/dutchvoting]. Regression, alongside monte carlo simulations, is also at the heart of my model predicting the outcome of the 2016-17 English Premier League soccer season [https://www.ft.com/prem-predict]. This was the first Premier League model to be created in-house and published by a newsroom.

Technologies used for this project:

I use the R programming language for obtaining, cleaning and analysing data -- including all regression work -- and for producing some visualisations of the results. d3 is then my primary tool for creating the final published graphics. To create the animation of the formation of El Niño [https://twitter.com/jburnmurdoch/status/722678306588504064], I wrote a custom script that would parse and transform more than 1GB of fiddly netCDF satellite data files and then render hundreds of frames of an animation onto canvas in d3.
Follow this project
Wait
Comment

Comments (0)

You have to be connected to contribute

You have to be connected to follow

Leave this project and no longer be informed about this project

By joining this project, you will be informed by email when an update or a new contribution is posted on the website.

Thank you for your active participation !
Best,

The GEN Community Team