Welcome to the second part of Eurovision Bets & Pieces’ resolutions!
How have I performed with my predictions and analyses of the Eurovision Grand Finals? What have I learned throughout the years, and what am I willing to change ahead of the Eurovision 2027 season?
It's time to look back at the numbers, revisit some of my biggest successes and failures, and set some new targets for the season ahead.
Winners: 3/7 — Keeping the Bigger Picture in Mind
My winner predictions on the day of each Grand Final were:
| Year | My predicted winner |
|---|---|
| 2019 | Netherlands |
| 2021 | Italy |
| 2022 | Italy |
| 2023 | Sweden |
| 2024 | Croatia |
| 2025 | France |
| 2026 | Australia |
That's 3/7 correct calls. Not a particularly impressive record on its own.
France 2025 was by far my biggest miss. The same could be said about Italy in 2022, not because Italy finished sixth, but because I failed to identify Ukraine as the winner.
In my defence, Ukraine in 2022 and Switzerland in 2024 were both second in my predictions and were mentioned as potential winners. The same can be said about Bulgaria last year, which ranked third in my projections. I considered Bulgaria a genuine contender and tipped them to win as well.
That leaves Austria in 2025, which I had ranked fourth. It remains the lowest-ranked country in my predictions to actually win the contest.
This was also my most controversial call, and I still believe Austria managed to escape with the victory because a series of unfortunate events weakened the other favourites and front-runners, while Austria was the only country that did not suffer the same fate.
My prediction for the winner's path was not wrong: a jury winner scoring below 300 points, followed by a public vote in which the entry would collect mostly 4s, 6s and 8s from the majority of participating countries.
Five blanks in the jury vote and zero sets of 12 points in the public vote proved my point.
For additional context, Bulgaria last year received 54 fewer jury points than Austria, despite having two fewer countries voting, but was blanked by fewer countries than Austria.
Sometimes you play well and lose. Other times, you play badly and win. For me, that explains Austria and Eurovision 2025 in general.
Another thing you may have noticed is that I have generally avoided making the market favourite my predicted winner. I had Sweden second in 2025 and Finland fourth last year, despite both trading at odds of 1.65–1.90 on Grand Final day.
Although this article is not about betting odds, they always have an anchoring effect on predictions made on the day of the contest.
To put things into perspective, someone placing 1,000 units on Finland on Grand Final day would have received 1,800 units in return. To get the same return with Bulgaria, you needed to invest just 56 units that same morning.
My incorrect winner predictions finished as follows:
- Italy 2022 — 6th
- Croatia 2024 — 2nd
- France 2025 — 7th
- Australia 2026 — 4th
The lesson for Eurovision 2027: When predicting the winner, I need to keep the bigger picture in mind rather than becoming too focused on individual indicators or the market narrative.
Top 3: My Achilles' Heel
My record in this category was remarkably stable for years, correctly predicting two of the three podium finishers in every edition from 2019 to 2024.
Then came 2025, when I failed to predict a single country in the Top 3. Last year, Bulgaria was my only correct call.
Here are my predictions through the years:
| Year | 1st | 2nd | 3rd |
|---|---|---|---|
| 2019 | Netherlands | Switzerland | Italy |
| 2021 | Italy | Malta | Switzerland |
| 2022 | Italy | Ukraine | Spain |
| 2023 | Sweden | Israel | Spain |
| 2024 | Croatia | Switzerland | Israel |
| 2025 | France | Sweden | Netherlands |
| 2026 | Australia | Greece | Bulgaria |
Correct calls: 11/21 — 52.4%
Seven of my missed predictions still finished relatively close to the podium:
- Three finished 4th: Switzerland 2019, Sweden 2025 and Australia 2026.
- One finished 5th: Israel 2024.
- One finished 6th: Italy 2022.
- Two finished 7th: Malta 2021 and France 2025.
Then there are three predictions that went completely off track: Greece 2026, which finished 10th; the Netherlands 2025, which finished 12th; and Spain 2023, which finished 17th.
The devil is in the details. With the exception of Russia in 2019, which I had ranked seventh, most of the other actual Top-3 finishers were ranked fourth or fifth in my projections.
And here's another interesting pattern: for three consecutive years, and four times in the last five editions, the country I ranked fifth went on to finish in the Top 3.
Those countries were the UK in 2022, Ukraine in 2024, Estonia in 2025 and Israel in 2026.
The goal for Eurovision 2027 is to get back on track and achieve the first 3/3!
Top 5: Hitting the Bullseye
Now we're talking business!
I correctly predicted three of the Top 5 finishers in three editions, and four of the Top 5 in three other editions. In 2024 I predicted all five of the Top 5. My record in this category is one of the strongest parts of my Grand Final analysis.
This is also where most of the corrections happen. In 2025, for example, my 0/3 Top-3 record turned into three correct Top-5 predictions, thanks to having Austria fourth and Estonia fifth, while Sweden finished slightly lower than I had expected.
The same story repeated itself last year, when one correct Top-3 prediction turned into three correct Top-5 calls.
Here are my predictions for fourth and fifth place:
| Year | 4th | 5th |
|---|---|---|
| 2019 | Azerbaijan | Australia |
| 2021 | Ukraine | France |
| 2022 | Sweden | UK |
| 2023 | Finland | Norway |
| 2024 | France | Ukraine |
| 2025 | Austria | Estonia |
| 2026 | Finland | Israel |
Correct calls: 26/35 — 74.3%
There are some interesting findings when looking at both positions.
My predictions become more reliable when it comes to this range of the scoreboard. The worst result for a country I predicted to finish fourth was Azerbaijan in 2019, which ended up eighth. There is also a winner in this category: Austria in 2025.
The most impressive statistic, however, comes from my predicted fifth-place finishers.
Three of them actually finished second: France in 2021, the UK in 2022 and Israel in 2026.
Two more finished third: Ukraine in 2024 and Estonia in 2025.
The lowest finish among my predicted fifth-place entries was Australia's ninth place in 2019.
In other words, my predictions around the fourth-to-fifth-place range have repeatedly identified countries with considerably higher finishing potential.
Top 10: Reaching for the Stars
Let's start with the lows.
I correctly predicted seven of the Top 10 finishers in three editions: 2019, 2023 and 2025. I got eight correct in 2022 and 2024, while my personal best came in 2021 and 2026, with nine correct calls.
Had I swapped France, which I projected 10th, with Denmark, projected 11th, last year, I would have achieved the ultimate 10/10!
Correct calls: 53/70 — 75.7%
There are a couple of interesting findings here.
Five of the seven countries I ranked 11th actually finished in the Top 10. The same happened to three of the seven countries I ranked 12th.
That means eight additional countries were just outside my predicted Top 10, despite eventually making it into the actual Top 10. These near misses show how much room there is for improvement in identifying the final qualifiers for that group.
Out of the 17 missed calls, only eight countries finished on the right-hand side of the scoreboard, in 14th place or lower.
Six of those eight misses came in just two editions:
- 2019: Serbia, Belarus and Greece.
- 2023: Spain, France and Armenia.
Looking at the bigger picture, my Top-10 predictions have generally been heading in the right direction.
There are only four cases in which a country finished 10th or lower despite being ranked in my Top 6: the Netherlands in 2022, Spain in 2023, the Netherlands in 2025 and Greece in 2026.
That's just 4/42.
The challenge for Eurovision 2027 is to turn those near misses into correct calls and improve the accuracy of my final Top-10 predictions.
Average Place Deviation per Country
Another way to measure the accuracy of my predictions is to calculate how far each country's predicted position was from its actual Grand Final result.
Here are the results for the seven editions so far:
| Year | Average place deviation |
|---|---|
| 2019 | 4.61 |
| 2021 | 3.15 |
| 2022 | 2.88 |
| 2023 | 4.15 |
| 2024 | 3.44 |
| 2025 | 3.61 |
| 2026 | 2.56 |
Eurovision 2026 was my most accurate year by far.
The biggest surprise, though, came in 2025. I expected my overall deviation to be considerably worse than in 2019 and 2023, but the final numbers told a different story.
My goal for Eurovision 2027 is to set a new personal record and bring the average place deviation below 2.50.
Eurovision Bets & Pieces vs The Model: Who Did It Better?
Do you need a model to predict Eurovision results? Can an algorithm solve the puzzle?
The results suggest that the answer is not quite that simple.
I have compared my predictions with those of the Model for the last three editions in which it has been running. Here are the findings.
Predicted Winners
| Correct | |
|---|---|
| Eurovision Bets & Pieces | 0/3 |
| The Model | 1/3 |
My predictions were Croatia 2024, France 2025 and Australia 2026.
The Model predicted Switzerland 2024, Sweden 2025 and Finland 2026.
Top 3 Predictions
| Correct | |
|---|---|
| Eurovision Bets & Pieces | 3/9 |
| The Model | 4/9 |
Top 5 Predictions
| Correct | |
|---|---|
| Eurovision Bets & Pieces | 11/15 |
| The Model | 8/15 |
Top 10 Predictions
| Correct | |
|---|---|
| Eurovision Bets & Pieces | 24/30 |
| The Model | 24/30 |
Average Place Deviation per Country
| Year | Eurovision Bets & Pieces | The Model |
|---|---|---|
| 2024 | 3.44 | 4.96 |
| 2025 | 3.61 | 5.19 |
| 2026 | 2.56 | 3.04 |
The Model identified more correct winners and Top-3 finishers, while we finished level in the Top-10 category.
On the other hand, Eurovision Bets & Pieces achieved more correct Top-5 predictions and a lower average place deviation in all three editions.
So, does the algorithm solve the Eurovision puzzle? Not entirely.
The results show that the Model can identify potential winners and podium finishers, while my own analysis has been more accurate when it comes to the broader Top 5 and the overall scoreboard.
The Man vs The Algorithm battle is far from over.
Eurovision 2027: What Comes Next?
Seven Grand Finals provide a useful dataset, but they also offer a reminder of how unpredictable Eurovision can be.
My priorities for the new season are clear:
- Keep the bigger picture in mind when identifying potential winners.
- Improve my Top-3 predictions and aim for the first 3/3.
- Turn near misses into correct Top-10 calls.
- Set a new personal record for average place deviation, below 2.50.
- Continue testing my analysis against the Model.
The numbers show where my predictions have been most consistent and where there is still room for improvement. The challenge now is to turn those lessons into a more accurate and reliable Eurovision 2027 prediction system.
And, of course, to keep looking for the signals that the market might be missing.
The next season is just around the corner. Let's see what the numbers have in store for us!
Eurovision Bets & Pieces — The Man vs The Algorithm, continued in 2027.
