Rank Position Forecasting in Car Racing

TL;DR

This paper introduces RankNet, a deep learning model that improves the accuracy and stability of predicting future race car positions by separately modeling rank sequences and pit stop events, incorporating uncertainty estimation.

Contribution

The paper proposes a novel cause-effect decomposition approach in deep forecasting models, specifically tailored for the irregular pit stop events in car racing.

Findings

RankNet outperforms baseline models with over 10% MAE improvement.

It demonstrates greater stability and adaptability to unseen data.

The approach offers promising tools for racing analysis and general forecasting challenges.

Abstract

Forecasting is challenging since uncertainty resulted from exogenous factors exists. This work investigates the rank position forecasting problem in car racing, which predicts the rank positions at the future laps for cars. Among the many factors that bring changes to the rank positions, pit stops are critical but irregular and rare. We found existing methods, including statistical models, machine learning regression models, and state-of-the-art deep forecasting model based on encoder-decoder architecture, all have limitations in the forecasting. By elaborative analysis of pit stops events, we propose a deep model, RankNet, with the cause effects decomposition that modeling the rank position sequence and pit stop events separately. It also incorporates probabilistic forecasting to model the uncertainty inside each sub-model. Through extensive experiments, RankNet demonstrates a strong performance improvement over the baselines, e.g., MAE improves more than 10% consistently, and is also more stable when adapting to unseen new data. Details of model optimization, performance profiling are presented. It is promising to provide useful forecasting tools for the car racing analysis and shine a light on solutions to similar challenging issues in general forecasting problems.