Eurovision 2025 grand final prediction model — interactive scorer with adjustable weights
The fastest way to lose your money
Betting the odds-on favourite at 1/2 or shorter gives you tiny returns for real risk — a single spectacular staging failure wipes you out. Spreading too wide on "value" bets (8+ countries) means your winnings never cover the losers. The classic punter's trap: recency bias after a barnstorming semi-final performance — semi juries and televoters are not the same people as the grand final ones. Also: bloc voting can completely flip a "sure thing". If you must gamble: pick one or two, accept the loss, enjoy the show.
Model weights — adjust how much each factor matters
15
20
18
10
20
12
8
5
Your personal adjustment — override any country
Slide each country's modifier to reflect your own view of the songs. Positive = you think it'll outperform the model; negative = you're sceptical.
Model methodology notes
Odds: capped logarithmically so a favourite can't dominate — reflects ~50% predictive accuracy historically. Jury history: based on 10-year average finishing positions of each country with national juries. Televote history: same, for public vote. Running order: songs in positions 18–24 get a bonus (data shows ~40% of winners come from positions 20–25 of 26). Song quality: composite of Spotify/YouTube pre-contest streaming ranks, fan poll aggregates (OGAE, Wiwibloggs), and ESC community rankings. Semi-final signal: whether the act topped or strongly won their semi with the televote — jury data not available until after the final. Bloc voting: Scandinavian, Balkan, ex-Soviet and Balkan blocs modelled as bonus vote pools. Diaspora/size: larger populations with diaspora communities (France, Italy, Spain, Ukraine, Israel) get a small structural boost. All factors normalised to 0–100 before weighting. Personal adjustment: ±25 raw points.