November 6, 2025
Beef on the dancefloor
I analyzed the lineups at the most popular nightclubs
Data nerd maps club lineups—fans cheer, locals cry “RA is messy”
TLDR: A coder compared 2019 club lineups and found most venues barely share DJs—about 1% overlap. The crowd loved the visuals but blasted messy data, missing context (locations, LGBTQ), and argued residents matter more than touring stars, turning it into spreadsheets vs vibes.
A data geek scraped Resident Advisor listings for 2019 and charted which clubs book similar DJs, using a “how much do we overlap?” score. The punchline: across 131 clubs and 8,502 events, overlap averaged just 1%. Cue dance floor drama. Some cheered the slick charts (built with D3, a fancy web graph tool) and the polite scraping, but the comments quickly went full nightclub bouncer.
The hottest takes? “Your data is messy,” cried one, pointing out artists often appear in descriptions but aren’t properly linked, so bookings get missed. Another dragged the venue list, noting Croatia’s “nightclubs” included an actual beach and arguing spots should be grouped, not split. Others demanded basics the charts skipped—like where clubs are and who they serve—because LGBTQ+ nights and straight scenes don’t book the same vibes. Meanwhile, purists said this isn’t about “popular DJs” at all: top clubs often thrive on residents and local heroes (cue the “people go to Berghain for Berghain” meme).
So we got two camps: spreadsheet stans worshiping the similarity math, and club rats shouting that vibes and context beat stats. Bonus jokes: “Pandas in the DJ booth,” “D3 dropped bangers,” and pandemic-era “Streamland” got roasted as “Dreamland.”
Key Points
- •Data from Resident Advisor was scraped with a Python/Beautiful Soup tool, throttled via robots.txt and cached locally.
- •The dataset covered the 20 most popular RA regions (excluding “Streamland”) and included all 2019 events for popular clubs.
- •Cleaned and normalized data was consolidated in Pandas; club similarity was measured using the Jaccard index.
- •A weighted club graph was built with NetworkX and clustered via community detection to identify similar booking profiles.
- •2019 results: 131 clubs, 8,502 events, 9,405 unique artists, 30,482 bookings; 3,716 of 8,515 club pairs overlapped with ~1% average overlap.