Court and Surface Effects: Distribution Applications

## Surface as a Probability Modifier Different playing surfaces produce systematically different probability distributions for key statistics. Identifying these surface effects and incorporating them into probability models produces measurable accuracy improvements. ## Football: Grass vs Artificial Turf Research shows artificial turf (used in some Scandinavian and lower-division leagues) produces: - Approximately 0.2–0.4 more goals per match - Higher passing completion rates - Different injury patterns (surface-specific conditioning) A model without a turf adjustment will systematically underestimate goals in turf league matches. ## Tennis: Clay vs Grass vs Hard Surface is the most significant predictor in tennis, often overriding raw player quality: - **Clay:** Long rallies, lower service dominance, higher break point percentage - **Grass:** Short points, high service dominance, fewer breaks - **Hard:** Intermediate — most variable depending on specific court speed Statistical profiles on each surface should be modelled separately. A clay specialist's win probability on grass may be 15–20 percentage points lower than their clay probability against the same opponent. ## Basketball: Home vs Away NBA home court advantage is more stable and larger than in most sports: approximately 3–4 points per game. Regular season vs playoff home advantage differs significantly (playoff away teams perform closer to home). ## Golf: Course Characteristics Specific course types (links, parkland, elevation) correlate with specific skill sets. Long hitters outperform on wide courses; accurate iron players outperform on tight courses. Course history for each player is among the most predictive factors in golf models. Building course/surface/venue-specific adjustments is a key differentiator for sport-specific models.