The 2014 World Cup is the hottest World Cup ever. Not just because of the soaring temperatures in Brazil that send players begging for water breaks, but also because of the high activity on social networks. Forget Sochi - the Brazil vs. Chile game alone broke Twitter’s record for most tweets at a live event with more than 16.4 million tweets during the game.
Curious to take an in-depth look at what happens on Twitter during a game, we collected World Cup tweets during the Australia-Netherlands match on June 18, 2014. The data was collected via DataSift with the following hashtags: fifa2014, fifaworldcup, brazil2014, brasil2014, worldcup2014, and worldcup. It was then aggregated using Xplenty’s data integration on the cloud.
Twitter Game Timeline
Here’s a graph showing number of tweets over time. We started collecting the data before the game and finished after the game. Can you tell when it started, ended, came to half time, and how many goals were scored by looking at the graph?
The correct answers: the first half started at 16:00 UTC and ended at 16:48, while the second half started at 17:03, and ended at 17:51. Altogether, five goals were scored, but they are not represented by the five peaks in the graph - no goals were scored at the end of the match. As it turns out, the first huge peak, which is more or less double as high than the other major peaks, represents two consecutive goals by the opposing teams, so people may have been doubly as excited to tweet about them.
Here’s the same graph with annotations:
We can see several things here:
The maximum number of tweets came one minute after Australia scored their first goal (8,100), which is also two minutes after the Netherlands scored.
Users tweeted almost double as much on the last minute of the match (3,000) than the first minute (1,700). The game must have exceeded all expectations.
There were slightly more tweets than usual during fouls, yellow cards, and free kicks, but not nearly as many as there were for goals.
Coming in at the end of the graph, the first minute of the Spain-Chile match got about 33% more tweets than the first minute of the Australia-Netherlands match (2,310 vs. 1,740). Perhaps it was a more anticipated game, or maybe the Spanish and Chileans like to tweet more than the Oranges and the Ozzies.
Tweets per Event
Clearly, we can see that people tweeted the most during goals. But how much more? To find out the answers, we added play by play data from FIFA to understand which events happened during every minute of the game and categorized them. In the case where several events occurred on a certain minute, we chose the most prominent one. Altogether, there were 157,500 tweets from game start to finish. This is how they were distributed:
There were a lot of tweets when nothing in particular happened, because, well, most of the time nothing in particular happened. To get a better picture which events were tweeted the most, we calculated the average number of tweets per minute by event type.
Average Tweets per Minute by Event
Now we can see the picture in HD:
Goals scored the most average tweets per minute - about 4,500 TPM.
Surprisingly, hand balls came in second with about 2,500 average TPM.
Are tweeters less excited by shots on target? On average, there were 10% more tweets per minute during shots off target (1,650 average TPM) than shots on target (1,510 average TPM).
There wasn’t that much hassle about fouls (1,500 average TPM) and yellow cards (960 average TPM), not to mention offsides which came in last (600 average TPM). Perhaps they weren’t controversial enough.
As mentioned above, ‘blank’ didn’t come in last because it also includes minutes following major events.
World Cup tweets during the Australia-Netherlands game nearly mirrored the actual match. Twitter users screamed with joy along with the commentators during goals and went on about them for a minute or two. Hand balls seemed to provoke more talk on Twitter than fouls or free kicks. And ultimately, as football fans know, most of the time was spent in anticipation, probably glancing at the Twitter feed every now and then. Would you also like to process Twitter and data from social networks? Get a free Xplenty account