Collegiate cross country teams often build their season schedules on intuition rather than evidence, partly because large-scale performance datasets are not publicly accessible. To address this limitation, we introduce the National Running Club Database (NRCD), the first openly available dataset to aggregate 23,725 race results from 7,594 collegiate club athletes across the 2023-2025 seasons. Unlike existing resources, NRCD includes detailed course metadata, allowing us to develop two standardized performance metrics: Converted Only (distance correction) and Standardized (distance, weather, and elevation adjusted). Using these standardized measures, we find that athletes with slower initial performances exhibit the greatest improvement within a season, and that race frequency is the strongest predictor of improvement. Using six machine learning models, random forest achieves the highest accuracy (r squared equals 0.92), revealing that athletes who race more frequently progress significantly faster than those who do not. At the team level, programs whose athletes race at least four times during the regular season have substantially higher odds of placing in the top 15 at nationals (chi-squared less than 0.01). These results challenge common coaching practices that favor minimal racing before championship meets. Our findings demonstrate that a data-informed scheduling strategy improves both individual development and team competitiveness. The NRCD provides a new foundation for evidence-based decision-making in collegiate cross country and opens opportunities for further research on standardized, longitudinal athlete performance modeling.