Saptashwa
← All projects
Data and NLP analysis, published as a videoOct 2024

Champions League format analysis

What fans, clubs and players made of the new format: a comment graph, like-weighted sentiment and wage data.

My partDid all of it by hand: collecting the comments, the graph model, the analysis, the interviews and the video.
Hand-built
Watch the video
A network graph of comment keywords on UEFA's league-phase draw video, with the top phrases labelled
Top key phrases in comments on UEFA's draw video, as a graph.
5,575 commentsfrom 12 YouTube videos on 10 channels, weighted by likes

What it does

  • Collected YouTube comments on the new format, extracted keywords and modelled them as a graph in Neo4j.
  • Weighted sentiment by likes, so opinions others agreed with counted more than single loud comments.
  • Compared reactions to the new format with reactions to the Super League.
  • Compared each club's wage bill with the prize money for the round it was expected to reach.
  • Interviewed fans, and put it all together as a 26-minute video.

What weighting by likes changed

Comment sentiment across all twelve videos, counted once per comment and weighted by likes.

PositiveNeutralNegative
By number of commentsBy number of comments: 26.4% positive26% positiveBy number of comments: 45.8% neutral46% neutralBy number of comments: 27.8% negative28% negativeWeighted by likesWeighted by likes: 40.7% positive41% positiveWeighted by likes: 34.9% neutral35% neutralWeighted by likes: 24.4% negative24% negative

The comments people agreed with were more positive than a plain count suggests.

Prize money against wages

Prize money for the round each club was expected to reach, as a share of its yearly wage bill. 32 of the 36 clubs had wage data.

0%50%100%150%200%Sturm GrazSturm Graz: prize money 191.4% of yearly wages191.4%Young BoysYoung Boys: prize money 145.5% of yearly wages145.5%SalzburgSalzburg: prize money 99.1% of yearly wages99.1%ShakhtarShakhtar: prize money 90.3% of yearly wages90.3%BrestBrest: prize money 79.1% of yearly wages79.1%CelticCeltic: prize money 72.5% of yearly wages72.5%Club BruggeClub Brugge: prize money 70.8% of yearly wages70.8%Sporting CPSporting CP: prize money 68.6% of yearly wages68.6%FeyenoordFeyenoord: prize money 62.9% of yearly wages62.9%PSVPSV: prize money 60.5% of yearly wages60.5%LilleLille: prize money 57.6% of yearly wages57.6%BolognaBologna: prize money 53.1% of yearly wages53.1%GironaGirona: prize money 50.3% of yearly wages50.3%StuttgartStuttgart: prize money 49.4% of yearly wages49.4%BenficaBenfica: prize money 37.1% of yearly wages37.1%MonacoMonaco: prize money 36.4% of yearly wages36.4%AtalantaAtalanta: prize money 33.1% of yearly wages33.1%LeverkusenLeverkusen: prize money 26.5% of yearly wages26.5%InterInter: prize money 21.6% of yearly wages21.6%ArsenalArsenal: prize money 20.9% of yearly wages20.9%MilanMilan: prize money 20.3% of yearly wages20.3%LiverpoolLiverpool: prize money 20% of yearly wages20.0%LeipzigLeipzig: prize money 19.5% of yearly wages19.5%B. DortmundB. Dortmund: prize money 18.1% of yearly wages18.1%Man CityMan City: prize money 17.8% of yearly wages17.8%JuventusJuventus: prize money 16.4% of yearly wages16.4%Aston VillaAston Villa: prize money 16% of yearly wages16.0%Real MadridReal Madrid: prize money 15.9% of yearly wages15.9%PSGPSG: prize money 14.7% of yearly wages14.7%Atletico MadridAtletico Madrid: prize money 14.4% of yearly wages14.4%Bayern MünchenBayern München: prize money 11.2% of yearly wages11.2%BarcelonaBarcelona: prize money 9.6% of yearly wages9.6%

The line marks 100%: prize money equal to the whole wage bill.

Show as a table
ClubPrize money as % of wages
Sturm Graz191.4%
Young Boys145.5%
Salzburg99.1%
Shakhtar90.3%
Brest79.1%
Celtic72.5%
Club Brugge70.8%
Sporting CP68.6%
Feyenoord62.9%
PSV60.5%
Lille57.6%
Bologna53.1%
Girona50.3%
Stuttgart49.4%
Benfica37.1%
Monaco36.4%
Atalanta33.1%
Leverkusen26.5%
Inter21.6%
Arsenal20.9%
Milan20.3%
Liverpool20.0%
Leipzig19.5%
B. Dortmund18.1%
Man City17.8%
Juventus16.4%
Aston Villa16.0%
Real Madrid15.9%
PSG14.7%
Atletico Madrid14.4%
Bayern München11.2%
Barcelona9.6%

Screens

A hand-drawn notebook sketch of weighting a comment's sentiment by its likes
The method, sketched: a comment's sentiment counts in proportion to its likes.
A network graph of keywords across all twelve videos combined
Top key phrases across all twelve videos.

Shares a stack with

PythonRoleRadarLLM job-posting scorer