Fabian Forsberg
@fabianforsberg.bsky.social
Engineering student at KTH
That could also work as a reasonable explanation! 😄
May 20, 2025 at 5:17 AM
That could also work as a reasonable explanation! 😄
If you would like to check my work, or contribute to the project, I have uploaded everything to my github!
github.com/kth-fabfor/E...
I have also posted my assumptions in making the analysis there!
If you have aby questions or thoughts, don't hesitate to DM me.
#EBU #ESC #Eurovision #Data
github.com/kth-fabfor/E...
I have also posted my assumptions in making the analysis there!
If you have aby questions or thoughts, don't hesitate to DM me.
#EBU #ESC #Eurovision #Data
GitHub - kth-fabfor/Eurovision-Analysis
Contribute to kth-fabfor/Eurovision-Analysis development by creating an account on GitHub.
github.com
May 18, 2025 at 7:11 PM
If you would like to check my work, or contribute to the project, I have uploaded everything to my github!
github.com/kth-fabfor/E...
I have also posted my assumptions in making the analysis there!
If you have aby questions or thoughts, don't hesitate to DM me.
#EBU #ESC #Eurovision #Data
github.com/kth-fabfor/E...
I have also posted my assumptions in making the analysis there!
If you have aby questions or thoughts, don't hesitate to DM me.
#EBU #ESC #Eurovision #Data
In conclusion:
Yes, the televote was REALLY weird, you are not crazy.
The most likely explanation for it seems to have been a lack of engagement from traditional Eurovision fans, most likely due to a distrust for the EBU and an effort to boycott them due to platforming Israel in the competition.
Yes, the televote was REALLY weird, you are not crazy.
The most likely explanation for it seems to have been a lack of engagement from traditional Eurovision fans, most likely due to a distrust for the EBU and an effort to boycott them due to platforming Israel in the competition.
May 18, 2025 at 7:11 PM
In conclusion:
Yes, the televote was REALLY weird, you are not crazy.
The most likely explanation for it seems to have been a lack of engagement from traditional Eurovision fans, most likely due to a distrust for the EBU and an effort to boycott them due to platforming Israel in the competition.
Yes, the televote was REALLY weird, you are not crazy.
The most likely explanation for it seems to have been a lack of engagement from traditional Eurovision fans, most likely due to a distrust for the EBU and an effort to boycott them due to platforming Israel in the competition.
In addition, Israel sending overtly political songs (as they did in 2024 and 2025) while also making rule-breaking ads encouraging people to vote for them may be contributing further to the disillusionment among "fans". This is difficult to definitely determine, but could be a contributing factor.
May 18, 2025 at 7:11 PM
In addition, Israel sending overtly political songs (as they did in 2024 and 2025) while also making rule-breaking ads encouraging people to vote for them may be contributing further to the disillusionment among "fans". This is difficult to definitely determine, but could be a contributing factor.
But why did we see a reduced turnout among traditional Eurovision fans in 2025? The most likely explanation is simply the EBU
Fans are in touch with the competition throughout the year, and often against Israeli participation due to the war in Gaza. This is the group that would boycott Eurovision.
Fans are in touch with the competition throughout the year, and often against Israeli participation due to the war in Gaza. This is the group that would boycott Eurovision.
May 18, 2025 at 7:11 PM
But why did we see a reduced turnout among traditional Eurovision fans in 2025? The most likely explanation is simply the EBU
Fans are in touch with the competition throughout the year, and often against Israeli participation due to the war in Gaza. This is the group that would boycott Eurovision.
Fans are in touch with the competition throughout the year, and often against Israeli participation due to the war in Gaza. This is the group that would boycott Eurovision.
I assume that the absolute number of votes for Israel thus remained approximately constant, (as 2024 was closer temporally to Oct 7th) leading to the natural conclusion that the total number of votes in the competition did not increase, but was reduced instead. This was especially true among "fans".
May 18, 2025 at 7:11 PM
I assume that the absolute number of votes for Israel thus remained approximately constant, (as 2024 was closer temporally to Oct 7th) leading to the natural conclusion that the total number of votes in the competition did not increase, but was reduced instead. This was especially true among "fans".
2. Israel retained basically the same vote/point share they had in 2024. It is a pointer to the fact that the absolute number of votes in the competition did not increase this year, as Israel's song in 2025 was significantly less popular than the one they performed in 2024.
May 18, 2025 at 7:11 PM
2. Israel retained basically the same vote/point share they had in 2024. It is a pointer to the fact that the absolute number of votes in the competition did not increase this year, as Israel's song in 2025 was significantly less popular than the one they performed in 2024.
I will also add that voting for Estonia can act as a protest vote against the EBU, for fans. Regardless, this would be a clear change in behavior among traditional "fans", and it seems that many of those who favored Sweden's song chose not to vote at all.
May 18, 2025 at 7:11 PM
I will also add that voting for Estonia can act as a protest vote against the EBU, for fans. Regardless, this would be a clear change in behavior among traditional "fans", and it seems that many of those who favored Sweden's song chose not to vote at all.
1. Estonia's meme-song heavily outperforming Sweden's meme-song, despite the latter being highly favored by "fans" in polling. Estonia's song is a lot more whimsical and would probably appeal more to casual viewers only watching the event during the evening.
May 18, 2025 at 7:11 PM
1. Estonia's meme-song heavily outperforming Sweden's meme-song, despite the latter being highly favored by "fans" in polling. Estonia's song is a lot more whimsical and would probably appeal more to casual viewers only watching the event during the evening.
In previous years, "fans" have been very representative for the televote results, and this was not the case in 2025.
There are 2 signs of this being due to reduced turnout among "fans":
There are 2 signs of this being due to reduced turnout among "fans":
May 18, 2025 at 7:11 PM
In previous years, "fans" have been very representative for the televote results, and this was not the case in 2025.
There are 2 signs of this being due to reduced turnout among "fans":
There are 2 signs of this being due to reduced turnout among "fans":
First, I will define what I mean with "Eurovision fans":
"Fans that would interact with the competition outside of the competition night, for instance through knowing and listening to the songs in advance. Assumed to be the same group that would engage with pre-ESC polling."
i.e. me and you lol 😅
"Fans that would interact with the competition outside of the competition night, for instance through knowing and listening to the songs in advance. Assumed to be the same group that would engage with pre-ESC polling."
i.e. me and you lol 😅
May 18, 2025 at 7:11 PM
First, I will define what I mean with "Eurovision fans":
"Fans that would interact with the competition outside of the competition night, for instance through knowing and listening to the songs in advance. Assumed to be the same group that would engage with pre-ESC polling."
i.e. me and you lol 😅
"Fans that would interact with the competition outside of the competition night, for instance through knowing and listening to the songs in advance. Assumed to be the same group that would engage with pre-ESC polling."
i.e. me and you lol 😅
It points towards a reasonable qualitative explanation for the differing voting behavior between 2025 and previous years: Reduced turnout among Eurovision fans.
Let me explain! 🤓☝️
Let me explain! 🤓☝️
May 18, 2025 at 7:11 PM
It points towards a reasonable qualitative explanation for the differing voting behavior between 2025 and previous years: Reduced turnout among Eurovision fans.
Let me explain! 🤓☝️
Let me explain! 🤓☝️
However, even when excluding these outliers, the correlation factor remains the same: 0.59. This would point to the vote shares projected to go to Sweden and Austria being evenly distributed among the other finalists, not just going to Israel and Estonia.
Why is this interesting? 👀
Why is this interesting? 👀
May 18, 2025 at 7:11 PM
However, even when excluding these outliers, the correlation factor remains the same: 0.59. This would point to the vote shares projected to go to Sweden and Austria being evenly distributed among the other finalists, not just going to Israel and Estonia.
Why is this interesting? 👀
Why is this interesting? 👀
Israel and Sweden had a |Z-value| of > 2, making them clear outliers.
It should be noted that the televote numbers are the percentage of points awarded, as the vote counts are not released by the EBU, making this analysis significantly more difficult. 🥲
It should be noted that the televote numbers are the percentage of points awarded, as the vote counts are not released by the EBU, making this analysis significantly more difficult. 🥲
May 18, 2025 at 7:11 PM
Israel and Sweden had a |Z-value| of > 2, making them clear outliers.
It should be noted that the televote numbers are the percentage of points awarded, as the vote counts are not released by the EBU, making this analysis significantly more difficult. 🥲
It should be noted that the televote numbers are the percentage of points awarded, as the vote counts are not released by the EBU, making this analysis significantly more difficult. 🥲
I was able to detect 2 clear outliers (|Z-value| > 2) and 2 potential outliers with a Z-value of ~1.9. These were the following:
Israel - Polling: 4.7%, Televote: 13.5%.
Estonia - Polling: 5.4%, Televote: 11.7%.
Sweden - Polling: 16.1%, Televote: 8.8%.
Austria - Polling: 14.3%, Televote: 8.1%.
Israel - Polling: 4.7%, Televote: 13.5%.
Estonia - Polling: 5.4%, Televote: 11.7%.
Sweden - Polling: 16.1%, Televote: 8.8%.
Austria - Polling: 14.3%, Televote: 8.1%.
May 18, 2025 at 7:11 PM
I was able to detect 2 clear outliers (|Z-value| > 2) and 2 potential outliers with a Z-value of ~1.9. These were the following:
Israel - Polling: 4.7%, Televote: 13.5%.
Estonia - Polling: 5.4%, Televote: 11.7%.
Sweden - Polling: 16.1%, Televote: 8.8%.
Austria - Polling: 14.3%, Televote: 8.1%.
Israel - Polling: 4.7%, Televote: 13.5%.
Estonia - Polling: 5.4%, Televote: 11.7%.
Sweden - Polling: 16.1%, Televote: 8.8%.
Austria - Polling: 14.3%, Televote: 8.1%.
The voting behavior in the 2025 televote was definitely different from previous years, and I think our null hypothesis can be safely rejected at this point.
The question that remains now is why it was different, and what is responsible for the reduced correlation.
The question that remains now is why it was different, and what is responsible for the reduced correlation.
May 18, 2025 at 7:11 PM
The voting behavior in the 2025 televote was definitely different from previous years, and I think our null hypothesis can be safely rejected at this point.
The question that remains now is why it was different, and what is responsible for the reduced correlation.
The question that remains now is why it was different, and what is responsible for the reduced correlation.
To illustrate the significance of this, when Russia invaded Ukraine just months before the 2022 ESC, the correlation factor was 0.70.
That year, Ukraine got 439 (of max 468) points from the public, despite only 9% of the pre-ESC polling favoring them as the winner.
2025 had a correlation of 0.59.
That year, Ukraine got 439 (of max 468) points from the public, despite only 9% of the pre-ESC polling favoring them as the winner.
2025 had a correlation of 0.59.
May 18, 2025 at 7:11 PM
To illustrate the significance of this, when Russia invaded Ukraine just months before the 2022 ESC, the correlation factor was 0.70.
That year, Ukraine got 439 (of max 468) points from the public, despite only 9% of the pre-ESC polling favoring them as the winner.
2025 had a correlation of 0.59.
That year, Ukraine got 439 (of max 468) points from the public, despite only 9% of the pre-ESC polling favoring them as the winner.
2025 had a correlation of 0.59.
Here’s the correlation between pre-contest polling and televote outcome:
2021: r = 0.87
2022: r = 0.70
2023: r = 0.78
2024: r = 0.81
2025: r = 0.59
While the pre-ESC polling is still (really) statistically significant, its relevance has been greatly reduced.
2021: r = 0.87
2022: r = 0.70
2023: r = 0.78
2024: r = 0.81
2025: r = 0.59
While the pre-ESC polling is still (really) statistically significant, its relevance has been greatly reduced.
May 18, 2025 at 7:11 PM
Here’s the correlation between pre-contest polling and televote outcome:
2021: r = 0.87
2022: r = 0.70
2023: r = 0.78
2024: r = 0.81
2025: r = 0.59
While the pre-ESC polling is still (really) statistically significant, its relevance has been greatly reduced.
2021: r = 0.87
2022: r = 0.70
2023: r = 0.78
2024: r = 0.81
2025: r = 0.59
While the pre-ESC polling is still (really) statistically significant, its relevance has been greatly reduced.
I started off with a question: Did the voting behavior for the 2025 ESC follow a similar pattern to previous post-covid ESC entries? (including (!) 2022)
This acted as my primary null hypothesis and I was able to safely reject it. This is why:
This acted as my primary null hypothesis and I was able to safely reject it. This is why:
May 18, 2025 at 7:11 PM
I started off with a question: Did the voting behavior for the 2025 ESC follow a similar pattern to previous post-covid ESC entries? (including (!) 2022)
This acted as my primary null hypothesis and I was able to safely reject it. This is why:
This acted as my primary null hypothesis and I was able to safely reject it. This is why:
I also made a fully pandas-ready json file that you can just import and use in your analysis!
Here it is!
drive.google.com/file/d/19ccL...
Here it is!
drive.google.com/file/d/19ccL...
ESC_points_pandas_ready.json
drive.google.com
May 18, 2025 at 10:44 AM
I also made a fully pandas-ready json file that you can just import and use in your analysis!
Here it is!
drive.google.com/file/d/19ccL...
Here it is!
drive.google.com/file/d/19ccL...
Unless you're trying to win an election 😅
December 15, 2024 at 4:32 PM
Unless you're trying to win an election 😅
They know that China won't pay for the tariffs. The problem is that the Trumpsters don't actually care about inflation, it was just a way of bashing the democrats, even though the inflation was mostly caused by Covid.
They will never complain about the inflation from the tariffs, trust me.
They will never complain about the inflation from the tariffs, trust me.
December 15, 2024 at 4:29 PM
They know that China won't pay for the tariffs. The problem is that the Trumpsters don't actually care about inflation, it was just a way of bashing the democrats, even though the inflation was mostly caused by Covid.
They will never complain about the inflation from the tariffs, trust me.
They will never complain about the inflation from the tariffs, trust me.