Objective of Hoops Forecast

The main goal of Hoops Forecast is to project how players will perform in the future, several years down the line. We use data visualization to make these projections more understandable, especially when showing the range of possible outcomes. While averages serve as a useful metric, they are insufficient on their own. To fully appreciate the projections, it is important to also understand concepts like variance.

Constantly Improving

This is v1.0, and while we believe the interface and projections are valuable, there are identifiable areas for improvement. We want to emphasize the interactive tools as much as the projections themselves.

Scope and Limitations

The projections should not be interpreted as definitive; they are intended as a foundation for further discussion and debate. Here are a few of our model's blind spots and areas where we look to improve in a potential v2.0:

• Input Data: Currently, the model is restricted to using publicly available historical statistics, biographical information, and a select set of amateur and professional data.
• Individualized Projections: Our projections are tailored to each player based exclusively on their past playing time as well as NBA-standard aging curves, not taking into account any team-specific situations. If a player switched teams this offseason, our model makes no adjustments until new data comes in.

Methodology of Projections

We employ sophisticated machine learning algorithms to forecast various skills for individual players. Using these projected skills, we can extrapolate a player's statistics for any number of future years. It's worth noting that the farther the projection extends into the future, the broader the range of potential outcomes. As clarified in the limitations, our projections are predominantly influenced by a player's true talent and historical playing time, along with traditional NBA aging curves.

Feedback and Iteration

This is our first time doing this, so we're open to hearing suggestions and feedback. If there's enough interest and support, we'll plan on improving what we're doing.

Contact Information

Should you encounter any anomalies or have any questions, please feel free to reach out to us either via email or direct message on Twitter at @HoopsForecast.