Stat Definitions

---

Understanding the performance of an A.I. handicap or score prediction model requires more than just looking at the final numbers. Below are key statistical terms—explained in plain English—that help measure how well the model works and why it matters for your game. 

Precision

•  What it means: The consistency of the model’s predictions—how tightly grouped they are around the actual result, even if not perfectly centered.
• Why it matters: A model can be precise without being accurate, but when it’s both, it provides stable, repeatable results. Precision ensures that the prediction doesn't swing wildly from round to round without reason. 

Accuracy

•  What it means: How close the predicted hole score is to the actual hole score.
•  Why it matters: A highly accurate model gives golfers a realistic expectation of their performance. The closer the prediction is to your real result, the more useful and trustworthy it becomes. 

Bias

•  What it means: A consistent error or skew in predictions—either too high or too low.
•  Why it matters: Reducing bias is key to fairness. For example, if the model consistently underestimates scores for high-handicap golfers, that bias can lead to unfair competition. The best models minimize bias across different player types and course difficulties. 

Average Net Par

•  What it means: The average number of strokes a golfer scores above net par, where net par is par adjusted for handicap.
•  Why it matters: It’s a quick way to measure overall performance trends—whether a golfer typically plays better or worse than their expected score. This is especially useful in formats like quota games or leagues. 

Standard Deviation

•  What it means: A measure of how much golfer hole scores vary from their average hole score.
•  Why it matters: High standard deviation means hole scores are more inconsistent; low standard deviation means they’re more steady. It helps A.I. models understand volatility in performance and set appropriate expectations. 

Absolute Deviation

•  What it means: The average distance between predicted and actual hole scores, regardless of whether the prediction was too high or too low.
•  Why it matters: Unlike standard deviation (which squares differences), absolute deviation gives a simple, direct measure of prediction error. It's a key way to evaluate how well a model performs across all players and courses.