4:08Watch on OFTV
Machine Learning Versus March Madness
Ari Bets@ari-bets · video 89 of 97
Lifecycle
compared with the average for videos published since tracking started (227 so far)- Published on OFTV
- Picked up by OF Candy
- First featured on the OFTV homepagenot yet
- Creator shown on OnlyFans Discovernot seen on Discover since this video came out
- Last seen featured—
- OnlyFans profile likes during the featurenot enough profile history around the feature yet · now 3,326 likes on @aribets702
Description
Machine learning, March Madness, sports analytics, AI, and predictive modeling all come together in this breakdown of overfitting, generalization, data splitting, validation, testing, and covariate shift. March Madness is one of the clearest real-world examples of how a machine learning model can perform well on training data and still fail on unseen data. Regular-season college basketball provides the training set, including KenPom ratings, tempo-free efficiency metrics, turnover percentages, offensive rebounding rates, conference trends, and matchup data.
Where OFTV showed it
every homepage, hero and genre spotNot seen in a featured spot since tracking started.
Homepage position per day
best spot that day, lower is betterBuilds up as the hourly collector logs the homepage.
OFTV does not publish views or likes per video. Likes are tracked per OnlyFans profile instead.
