Predicción de resultados en fórmula 1 mediante técnicas de aprendizaje automático y modelos de ranking
Loading...
Official URL
Full text at PDC
Publication date
2025
Authors
Advisors (or tutors)
Editors
Journal Title
Journal ISSN
Volume Title
Publisher
Citation
Abstract
Este trabajo evalúa la capacidad de distintos modelos estadísticos para predecir los resultados de las carreras de Fórmula 1 en la temporada 2025. Se basa en datos históricos desde la implementación del reglamento técnico de 2022, un periodo marcado por mayor fiabilidad mecánica y menor variabilidad en los resultados. Los modelos se entrenan exclusivamente con información disponible antes de cada carrera, como estadísticas de pilotos, equipos, posiciones de clasificación previas y variables derivadas. La motivación del estudio surge del interés en el deporte y del creciente debate sobre su previsibilidad. Además, sería posible explorar el uso de estos modelos en apuestas deportivas.
Abstract: This study evaluates the ability of various statistical models to predict the results of Formula 1 races during the 2025 season. It is based on historical data collected since the implementation of the 2022 technical regulations, a period characterized by greater mechanical reliability and reduced variability in race outcomes. The models are trained exclusively on information available before each race, such as driver and team statistics, previous qualifying positions, and derived variables. The motivation for the study stems from an interest in the sport and the growing debate around its predictability. Additionally, the use of these models could potentially be explored in the context of sports betting.
Abstract: This study evaluates the ability of various statistical models to predict the results of Formula 1 races during the 2025 season. It is based on historical data collected since the implementation of the 2022 technical regulations, a period characterized by greater mechanical reliability and reduced variability in race outcomes. The models are trained exclusively on information available before each race, such as driver and team statistics, previous qualifying positions, and derived variables. The motivation for the study stems from an interest in the sport and the growing debate around its predictability. Additionally, the use of these models could potentially be explored in the context of sports betting.







