There’s a middle-aged office worker named Caleb Davies who uses models to predict movie ratings and Spotify charts, then places bets on platforms like Kalshi; cumulatively, he has earned about $1.1 million.

He has long been scraping Rotten Tomatoes movie ratings and Spotify play data, updating his model every day. After platforms announce how many points a specific movie will ultimately get, or whether a particular song can climb onto the charts, he bets according to the results his model predicts.

Just from music and movie contracts, he has made more than $500k.

Davies has over ten years of trading experience, and has also worked for financial institutions and in IT.

His first bet was in 2015, for only $300.

In March 2024, he deposited $10k into his Kalshi account; by the end of the year, the account had already exceeded $200k, and he is now publicly ranked among Kalshi’s top traders.

The model can also run into data that humans have manipulated. This year, a trader bet on the niche song “Earrings” to suddenly surge; Davies placed a reverse order and lost $4,500.

Afterward, he found that the song’s play count showed extreme anomalies. Spotify later confirmed that there was manual volume boosting and removed more than 500,000 plays, but Kalshi had already settled based on the chart at the time of the boosted play count.

Davies is still doing prediction markets now; until the issue is resolved, he will no longer take new Spotify chart contracts.