The ballots have been counted, the results have been verified and despite our best efforts…the bookies won.
In the end our OscarBot correctly predicted 9 out of 24 wins, compared to Lily’s 13 and the bookies getting 16. No-one predicted the two big shocks of the night – Olivia Coleman winning Best Actress and Green Book winning Best Picture.
Our film buff Lily had this to say about the surprise Best Film winner:
“In many ways, the Oscar preference voting system for Best Picture means that the films that no one really hates tend to win out over divisive but, in my opinion, superior films.”
Our Data Scientist James had this to say about OscarBot’s performance:
“I would say the bot did pretty well with very limited access to any good data, especially twitter data. Given that, it’s still pretty amazing it got 9 right, and it was very on par with a lot of other similar predictive models that had a lot more training.”
Sadly, OscarBot was not available for comment.
Perhaps the lesson here is that without the right data then AI will never win. Or maybe something so political and subjective as The Oscars is just simply impossible for anyone to get 100% right.
The final results are as follows:
The post The results: Machine vs humans – predicting the Oscars with AI. appeared first on FOUND.