Ich habe dies gestern gepostet und viel Feedback zu Verbesserungsmöglichkeiten erhalten. Hier ist eine aktualisierte Version (der gestrige Beitrag ist hier: https://www.reddit.com/r/dataisbeautiful/comments/1u36v98/oc_which_stadiums_got_the_best_world_cup_group/ )
Von m0nkeybl1tz
8 Kommentare
Ticket price data came from [https://www.ticketdata.com/performer/world-cup-soccer](https://www.ticketdata.com/performer/world-cup-soccer) and FIFA rankings came from [https://inside.fifa.com/fifa-world-ranking/men](https://inside.fifa.com/fifa-world-ranking/men) . Right now it all exists in a giant Excel spreadsheet and I annotated the graph slightly in Figma.
From the other post:
I rated games on 3 metrics:
1. The **quality** of the teams (based on FIFA rankings)
2. The **closeness** of quality of the two teams (everyone wants to see Spain, not everyone wants to see Spain play Cabo Verde)
3. The **popularity** of teams (US and Canada games are going to be more popular, despite not being the best teams)
Metric 3 required some additional calculation, which came down to running a decomposition on each team’s 3 games — basically was their game against Team X more expensive or less expensive than Team X’s average. I then ran a linear regression against ticket prices to weigh each metric and combined them to generate a final „game score“.
I know it’s a meme by now but SF got truly fucked
Got anymore of them pixels?
*me, zoomed in:*
“Wow, Guadalajara only got three matches…… oh nevermind”
realise it’s based on price and closeness etc but wild seeing USA vs AUS ranked higher for quality than like England vs Croatia
Interesting analysis but beautiful this ain’t. Any chance of making those team names a bit blurrier?
Canada vs Bosnia is higher than France vs Norway?
ENG vs CRO and FRA vs SEN have to the top 2 no?
I know this is using a system but I think the system might be working unintended