A data-driven dive into the terrorist nature of African warfare

This story is part of my newsletter. Click here to subscribe.

RZESZÓW, Poland — According to the United Nations, the crisis in Cabo Delgado (Northern Mozambique) could lead to 1,000,000 people being displaced. Such an estimation could sound realistic if we consider that between the beginning of the year and April 3 there were more than 90 engagements in the country. Yet there is something that does not completely fit. Let us look at the situation on the ground.

Between January 1 and April 3 (the timespan of the map) there had been more than 90 armed clashes. Some of…


A data-driven dive into the *mythical* relationship between democracy and development.

This story is part of my newsletter DaNumbers. Click here to know more (and to subscribe — for free).

BRUSSELS — For more the 60 years, we lived under the assumption that democracy was the magic wand capable of bringing welfare and stability to the countries where it flourished. This assumption is increasingly hard to prove, particularly in 2021, in the last phases of a pandemic and where a large part of the world had been experiencing democratic declines for more than a decade.

Is it time, already, to sing the Requiem mass for democracy or we are asking ourselves…


Data suggest that Europe is pushing migrants southwards where they just can’t stop dying.

This story is part of my newsletter DaNumbers. Click here to know more (and to subscribe — for free).

BRUSSELS — Before the April 23 tragedy, in the Mediterranean 421 people had already lost their life since January 1, 2021. That number now is now above 551, given that last week approximately 130 people lost their lives in one of those tragedies that have been commonplace since 2011. The Missing Migrant Project by the International Organization of Migration collects data since 2014. The following map shows how things went in 2021, right before the last shipwreck.

Lampedusa, in the centre…


As Italy reaches 100,000 deaths, here a handy tutorial on how the NYT’s data team could have made its standard-setting front page data visualization.

The New York Times set a new standard (again) in data visualization a few weeks ago. They practically invented a new chart (which from now on I will call ‘static noise chart’) to visualize victims from the pandemic. They decomposed fatalities data in points where each point represents a dead person and scattered those data points more or less randomly on a horizontal axis, whereas, on the vertical axis, we have dates, with the most recent on the top. Given that on March 8, 2021, Italy passed the 100,000 COVID-related deaths, I decided to try to replicate the work by…


A Games Theory application says no.

Will the Republican Party split in two during the Biden administration? No, no and (again) no. The problem is very simple: there is a huge divide between Republican and Democrats but, more importantly, the party is too loyal to Donald Trump. More importantly, according to a cluster analysis of FiveThirtyEight’s loyalty index, loyalty was the single biggest ideological cleavage during the 116th Congress with the Democrats strongly antagonizing the soon-to-be-ex president and the Republicans divided in seven different scents of Trumpism.

The above chart shows very well the proceedings. Here I show how the clustering works. I used a very…


Notes from a partially successful experiment

In 2019, the RayShader R package hit CRAN with the promise of being a gamechanger. The promise was kept, in facts, it is a gamechanger. But the problem is that it just might be ahead of its time. I have just finished an experiment. It proved that a low-budget augmented reality data visualization initiative is technically feasible, but it needs still further technical (on my side), and technological advancements.

This chart represents the level of democracy in the main presidential regime in the Americas. Originally, it was meant to be a heatmap. I decided to turn it into a more…


We know restrictions are effective but we just don’t know how much

Three days after Boris Johnson’s announcement of the new English harsh lockdown, it is increasingly harder to policymakers to justify such measures. In facts, they are useful, but unfortunately, they are not the only way to prevent the spread of the pandemic. They are a patch on extremely worn clothing which is the European healthcare system, which is under unprecedented pressure right now.

The severity of lockdowns has had some impact on the rate of virus transmission, but it was not the definitive answer to the problem. If we study the problem under probabilistic terms, we see that, yes, the…


An R exercise offers mixed results.

Although the answer to the question ‘What do migrants run away from?’ might look obvious, the answer, in strict data terms, might not be so. That’s why I ran this exercise. So, what do migrants run away from? In this post, I assume that the answer is war. We have plenty of data on war and migration. The International Organization of Migration collects data on missing migrants, the Armed Conflict Location & Event Data Project (ACLED, one of the best conflict studies think tank in the world) collects data on conflict fatalities. Do conflict fatalities push migration fatalities, somehow?

To…


Don’t let the third wave get you unprepared

Italy’s multilevel governance system is complex. We have national, regional, provincial, and communal levels. The most granular COVID data available in Italy are at the provincial level. After some reflection on how to better visualize COVID in Italian provinces, I came up with a solution. Ladies and gentlemen, please welcome the: log-new daily cases with loess smoother.

This ggplot2 chart uses the y-axis with a logarithmic (base 10) transformation. This is different from most of the things you will see around. First, the logarithmic scales are (in theory) more suited for cumulative data. Yet, they have huge advantages in this…


538’s prediction on Nov. 4 2020 (11.30, CET)

A former data journalist's battle is finally vindicated.

I do not trust public opinion polls. I spent some of my best energies figuring out how to predict elections, not relying on public opinion polls. I pitched Ph.D. proposals openly criticizing global and well-established polls-based datasets (hello, World Values Survey!) and failed miserably, but what is apparent from last night’s results is that public opinion polls just do not work. Here are the problems, in my humble opinion.

People do not answer the damn phone.

My friend Juraj Medzihorský shows it very well on his Facebook profile: there has been a strong decline in telephone interview responses in the last 20 years. The reason is simple…

Francesco Piccinelli

Journalist and long time blogger

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store