The champion's audience seated in grandstands around a circuit, fourteen interest groups The club's audience seated in a stadium, nineteen country stands around the pitch The UK group selected: its links to every other country light up across the pitch
The 2026 explorer, captured from the rebuilt app with the real 2019 data: the circuit, the stadium, and the UK group lighting its links.

Sport · Audience networks

The home crowd came third.

Two audience maps from 2019, a Bundesliga club’s and a Formula 1 world champion’s: 10,856 Twitter accounts and 306,501 links, drawn so a brand team could see who it was really talking to.

The club and the champion are not named here. Every count is as in the 2019 files; the names on the screens are the maps’ largest public accounts.

Story by Javier AlcívarNetwork analysis and visualisation · Raven Path, 2019 · recovered and rebuilt, 2026

Client
Two audiences at Raven Path: a Bundesliga club and a Formula 1 world champion
The problem
A follower count says how many, not who, or who they listen to
My role
Network analysis and visualisation; in 2026, the recovery and the new explorer
The outcome
Two maps, 10,856 accounts and 306,501 links, rebuilt in 2026 as a live explorer

A follower count answers one question: how many. The questions a brand team actually has come after it. Who are these people? Where do they talk from? What else do they care about, and who do they listen to?

A network map answers those. Every account is a dot and every link between two accounts a line. Accounts that share links pull together, clusters form, and each cluster gets a name. In 2019 I made two of these maps at Raven Path, for two very different audiences in sport: a Bundesliga club, read by country, and a Formula 1 world champion, read by interest.

The standsThe home crowd came third.

The club’s map holds 6,891 accounts and 143,911 links in 19 country groups. You would expect Germany to lead. It comes third. The UK is the biggest group with 1,354 accounts, a fifth of the map, and Spain is close behind with 1,292. Germany has 827.

Who is in the stands?

Accounts in each group, to scale within each map. The club’s home country and the champion’s own sport are highlighted.

The club · 19 country groups

The champion · 14 interest groups

the 2019 Gephi exports (data.json), recovered October 2026. Group names are the maps’ own, spelling included.

The bridgesSpain holds the map together.

Size is only half of it. The lines between groups show who carries a conversation from one country to the next. The strongest bridge in the club’s map runs between the UK and Spain, with 6,395 links. Germany’s strongest, to Spain, has 1,878.

Spain sits in seven of the eight strongest bridges. And Nigeria, only 282 accounts, is in the second and the third: a small group that links the two biggest. That is the kind of account a brand team wants to know about, and the kind a follower count hides.

A follower count says how many. A map says who, and who they listen to.

The paddockFour in five came for something else.

The champion’s map is smaller, 3,965 accounts, but denser: 162,590 links in 14 interest groups. Motorsport is the biggest group and still only 21% of the audience. The rest follow world news, British television, the car industry and sustainable technology, among others.

The two audiences behave differently, too. In the club’s map 71% of links stay inside their own country group. In the champion’s, 45% of links cross from one interest to another. A club’s fans stay home; a champion’s audience travels. The strongest bridge is the obvious one, motorsport and the auto industry with 14,171 links, but the next two lead to world news and to British television.

Seated where they belong

In 2026 the maps were only files in an archive. I rebuilt them as an explorer: each audience fills its own venue, then opens group by group. Recorded from the explorer, so the page plays video instead of running 20 MB of graph.

The stadiumThe club’s audience fills the stands country by country, biggest accounts in the front rows.
The circuitThe champion’s audience sits in grandstands; a car laps the track and lights each one as it passes.
The ripplePick a group, here the UK, and its first, second and third-degree links light in turn; groups it barely touches fold away.

The rebuildFrom a blotch of colour to a map you can read.

The 2019 sites drew every line at once, 162,590 of them on the champion’s map, and the result was mostly colour. The new explorer runs on WebGL and draws only what the eye can use. At the farthest zoom it shows the links among the most important accounts and hides dots that would overlap; zoom in and the detail returns.

Selecting a group keeps its relations lit and lets the rest go grey: any group that gets under 3.5% of the selected group’s links, and is not among its four closest partners, folds into 100 to 200 representative circles. Every layout is the same data rearranged: the original map, a solar system with the biggest group as the sun, groups in a line by size, and 3D.

The champion's groups as a solar system
Solar system: the biggest group at the centre, the rest in slow orbit
The dashboard's map of where the club's audience lives
The dashboard: each country a circle sized by its accounts, with the 30 strongest ties
The motorsport group selected on the circuit
Motorsport selected: 53% of its links stay inside, and the auto industry is its closest partner

Captured from the rebuilt explorer with the 2019 data. The client names on screen are replaced with neutral labels.

The record

What the files held.

Both sites were still in Raven Path’s archive in October 2026, as Gephi’s Sigma.js export: a page, a settings file and the full graph. I recovered them as they were and kept untouched copies before rebuilding anything.

  1. The club’s map6,891 accounts, 143,911 links, 19 country groups.data.json, 2019
  2. The champion’s map3,965 accounts, 162,590 links, 14 interest groups.data.json, 2019
  3. Profile picturesMany 2019 links are gone; the ones still online were saved, the rest show initials.Recovery notes, October 2026
  4. One gap left as foundThe original “insight per group” panel held placeholder text, so the insights in the new dashboard are worked out from the data.Recovery notes, October 2026

Why it is hereThe same eye, seven years apart.

The tools changed between 2019 and 2026; the job did not. Find the shape hiding in the numbers, then build the thing that lets other people see it for themselves. A stadium for a club, a circuit for a driver, and one surprise each: the home crowd came third, and four in five came for something other than racing.

The explorer

Every account.
Every link.

The explorer holds both full maps, about 20 MB of graph each, so it is shown in a walkthrough rather than on this page. Ask for one on LinkedIn.

Ask for a walkthrough ↗

Is there a story hiding in your data? Write to me on LinkedIn ↗