While I've been heads down building out ndricks...

While I've been heads down building out ndricks Software's first product, SentiTrac, for the past couple of weeks, today I wanted to talk about the power of sentiment data for sports. While there are many use cases, let's start with the one I've already proven value for: fantasy projections. " "" "Traditionally, player projections are based on statistical data, but sentiment data offers a compelling, unexplored angle. In my Master's thesis research, I demonstrated that incorporating this data into our modeling can significantly increase projection accuracy." "" "For those unacquainted, sentiment data measures positive or negative perceptions about players drawn from various sources like social media and news articles. This data often reflects aspects not quantitatively measurable, such as a player's morale, team dynamics, or public perception, which can influence on-field performance." "" "My analysis involved developing a unique model that combined traditional NFL statistics with sentiment data from Twitter. The outcome? An impressive improvement in accuracy over NFL.com projections, using both regression and neural network models." "" "This upcoming football season, for the first time, I'll be making available -- admittedly for a subscription fee -- both the raw sentiment data and the fantasy projections leveraging that data. I am hard at work getting this ready for the upcoming NFL season, and I hope to have raw sentiment data available in the next couple of weeks. So stay tuned to see just how powerful sentiment data can be!

---

Original post on LinkedIn