Implementation of Data Mining for Predicting Formula 1 Team Performance Using the Trend Moment Method Based on Historical Data

Abstract

Formula 1 is an international racing competition that generates large-scale performance data of constructors in the form of time series, particularly total points accumulated each season. Such data can be utilized for predictive analysis using data mining techniques. This study aims to implement the Trend Moment method to predict the performance of five Formula 1 constructor teams for the 2026 season based on historical standings points data from 2019 to 2025. The data used in this study is secondary data obtained from the official FIA and Formula1.com websites. The research method applies time series forecasting using a simple linear regression model Y = a + bX. Model evaluation and validation are conducted using Mean Absolute Percentage Error (MAPE) to measure the level of prediction accuracy. The results show that McLaren Mastercard F1 Team is predicted to achieve 800 points in the 2026 season, followed by Oracle Red Bull Racing with 700 points, Scuderia Ferrari HP with 539 points, Mercedes-AMG Petronas Formula One Team with 366 points, and Atlassian Williams F1 Team with 94 points. The evaluation results indicate MAPE values ranging from 9.38% to 71.89%, suggesting that the model performs well on stable data patterns but is less effective on data with high volatility. The novelty of this study lies in the application of the Trend Moment method to Formula 1 constructor performance data based on official historical records, combined with MAPE evaluation to provide a simple, measurable, and easily interpretable predictive model that can be applied to broader professional sports analytics.