Eurovision Data Analysis: Fan Predictions Accurate, Language and Slot Number Key

Scavenger AI has systematically analyzed 68 years of Eurovision Song Contest data, encompassing 1,734 entries from 54 countries. The analysis reviewed 51,354 voting lines and 1,722 song lyrics to provide insights into the characteristics of winning songs and the reliability of predictions.

The analysis indicates that Eurovision fan predictions are surprisingly accurate. Between 2016 and 2024, predictions from the MyESC Community (over 10,000 users) placed all eight winners in at least seventh place. In contrast, predictions from the smaller OGAE fan club (under 900 members) often proved less reliable.

Performance order significantly impacts success; between 2016 and 2024, no country starting in the first five positions won the Grand Final. Winners tend to emerge from performance slots 11–15 and 21–26. Song language also plays a strategic role. After language restrictions were lifted in 1999, the proportion of English songs increased, but Portugal's 2017 victory with Salvador Sobral's Portuguese song has reversed this trend. An increasing number of winning songs are now performed in the native language.

In lyrical analysis, "forever" appears most frequently in winning songs, while "La La La" is found more often in last-place entries. Winning lyrics are generally more positive, although "love" is the most common word and "war" appears more often than "peace." The gender distribution also appears balanced, with four wins for male solo artists, three for female solo artists, and one for a group between 2016 and 2024. Solo artists tend to perform better than groups.

Based on Scavenger AI's findings, the optimal winning combination would involve a solo artist performing in slots 11–15 or 21–26, singing in their native language with lyrics about love. Ultimately, however, the performance itself, the music, and the moment on stage will decide the competition's outcome.