Win Your Race Goal: A Generalized Approach to Prediction of Running Performance

Abstract

We introduce a novel approach for predicting running performance, designed to apply across a wide range of race distances (from marathons to ultras), elevation gains, and runner types (front-pack to back of the pack). To achieve this, the entire running logs of 15 runners, encompassing a total of 15,686 runs, were analyzed using two approaches: (1) regression and (2) time series regression (TSR). First, the prediction accuracy of a long short-term memory (LSTM) network was compared using both approaches. The regression approach demonstrated superior performance, achieving an accuracy of 89.13% in contrast, the TSR approach reached an accuracy of 85.21%. Both methods were evaluated using a test dataset that included the last 15 runs from each running log. Secondly, the performance of the LSTM model was compared against two benchmark models: Riegel formula and UltraSignup formula for a total of 60 races. The Riegel formula achieves an accuracy of 80%, UltraSignup 87.5%, and the LSTM model exhibits 90.4% accuracy. This work holds potential for integration into popular running apps and wearables, offering runners data-driven insights during their race preparations.

Keywords

running performance prediction - deep learning - running logs - ultramarathon - marathon

Publikationsverlauf

Eingereicht: 20. Mai 2024
Eingereicht: 30. Juli 2024

Angenommen: 31. Juli 2024

Artikel online veröffentlicht:
09. Oktober 2024

• References

• 1 Berger NJA, Best R, Best AW. et al. Limits of ultra: towards an interdisciplinary understanding of ultra-endurance running performance. Sports Med 2024; 54: 73-93
• 2 UTMB World Series. [Online]. Available from: https://uta.utmb.world/overview Accessed: 28-Jun-2024.
• 3 Coates M, Berard J, King TJ. et al. Physiological determinants of ultramarathon trail-running performance. Int J Sports Physiol Perform 2021; 16: 1454-1461
• 4 Denadai BS, Greco CC. Could middle- and long-distance running performance of well-trained athletes be best predicted by the same aerobic parameters?. Curr Res Physiol 2022; 5: 265-269
• 5 Balasekaran G, Loh MK, Boey P, Ng YC. Running energy reserve index (RERI) as a new model for assessment and prediction of world, elite, sub-elite, and collegiate running performances. Sci Rep 2023; 13: 7416
• 6 Nicot F, Sabater-Pastor F, Samozino P. et al. Effect of ground technicity on cardio-respiratory and biomechanical parameters in uphill trail running. Eur J Sport Sci 2022; 22: 1836-1846
• 7 Riegel PS. Athletic records and human endurance: a time-vs-distance equation describing world-record performances may be used to compare the relative endurance capabilities of various groups of people. Am Sci 1981; 69: 285-290
• 8 Vickers AJ, Vertosick EA. An empirical study of race times in recreational endurance runners. BMC Sports Sci Med Rehabil 2016; 8: 26
• 9 Blythe DAJ, Király FJ. Prediction and quantification of individual athletic performance of runners. PLoS One 2016; 11: e0157257
• 10 Przednowek K, Iskra J, Maszczyk A, Nawrocka M. Regression shrinkage and neural models in predicting the results of 400-metres hurdles races. Biol Sport 2016; 33: 415-421
• 11 Admetlla AR. Software tools for self-quantifier runners. [M.S. thesis]. Yverdon-les-Bains. Switzerland: Institute for Information and Communication Technologies; ; 2015
• 12 Smyth B, Cunningham P. Running with cases: a CBR approach to running your best marathon. in Case-Based Reasoning Research and Development 2017; 10339: 360-374
• 13 Smyth B, Cunningham P. Marathon race planning: a case-based reasoning approach. In: Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence (IJCAI’18). 2018 Jul 13–19; Stockholm, Sweden. Washington, DC: AAAI Press; 2018: 5364-5368
• 14 Smyth B, Cunningham P. An analysis of case representations for marathon race prediction and planning. In: Cox M, Funk P, Begum S (eds). Case-Based Reasoning Research and Development. Cham: Springer International Publishing; 2018: 369-384
• 15 Feely C, Caulfield B, Lawlor A, Smyth B. Modelling the training practices of recreational marathon runners to make personalized training recommendations. In: Proceedings of the 31st ACM Conference on User Modeling, Adaptation and Personalization. 2023 Jun 26–29; Limassol, Cyprus. 2023, 183–193
• 16 Lerebourg L, Saboul D, Clémençon M, Coquart JB. Prediction of marathon performance using artificial intelligence. Int J Sports Med 2023; 44: 352-360
• 17 Maurer J. Race time prediction on individual historical training data for hilly and non-hilly courses. [M.S. thesis] Saarbrücken, Germany: Faculty of Natural Sciences and Technology I, Department of Computer Science,. Saarland University. 2018
• 18 Dracopoulos DC. A better predictor of marathon race times based on neural networks. In: Bramer M, Petridis M (eds) Artificial Intelligence XXXIV. Lecture Notes in Computer Science. Cham: Springer International Publishing; 2017: 293-299
• 19 Feely C, Smyth B, Caulfield B, Lawlor A. Estimating the cost of training disruptions on marathon performance. Front Sports Act Living 2023; 4: 1096124
• 20 Lövdal SS, Den Hartigh JR, Azzopardi G. Injury prediction in competitive runners with machine learning. Int J Sports Physiol Perform 2021; 16: 1522-1531
• 21 Fogliato R, Oliveira NL, Yurko R. TRAP: a predictive framework for the assessment of performance in trail running. J Quant Anal Sports 2021; 17: 129-143
• 22 McConnell C, Smyth B. Going further with cases: using case-based reasoning to recommend pacing strategies for ultra-marathon runners. In: Bach K., Marling C (eds). Case-Based Reasoning Research and Development. 2019. 11680. 358-372
• 23 Coquart JB. Prediction of performance in a 100-km run from a simple equation. PLoS One 2023; 18: e0279662
• 24 Gilligan M. UltraSignup – Frequently Asked Questions. Available from: https://ultrasignup.com/faqs.aspx Accessed: Mar. 18, 2024
• 25 Hill T, O’Connor M, Remus W. Neural network models for time series forecasts. Manage Sci 1996; 42: 1082-1092
• 26 Filipas L, Bonato M, Gallo G, Codella R. Effects of 16 weeks of pyramidal and polarized training intensity distributions in well-trained endurance runners. Scand J Med Sci Sports 2022; 32: 498-511
• 27 Wang Z, Tai Wang Y, Gao W, Zhong Y. Effects of tapering on performance in endurance athletes: A systematic review and meta-analysis. PLoS One 2023; 18: e0282838
• 28 González-Ravé JM, González-Mohino F, Rodrigo-Carranza V, Pyne DB. Reverse periodization for improving sports performance: a systematic review. Sports Med Open 2022; 8: 56
• 29 Campos Y, Casado A, Vieira JG. et al. Training-intensity distribution on middle- and long-distance runners: a systematic review. Int J Sports Med 2022; 43: 305-316
• 30 Kenneally M, Casado A, Santos-Concejero J. The effect of periodization and training intensity distribution on middle- and long-distance running performance: a systematic review. Int J Sports Physiol Perform 2018; 13: 1114-1121
• 31 Gauthier DJ, Bollt E, Griffith A, Barbosa WAS. Next generation reservoir computing. Nat Commun 2021; 12: 5564
• 32 Borovykh A, Bohte S, Oosterlee CW. Dilated convolutions neural networks for time series forecasting. J Comput Finance 2019; 22: 73-101
• 33 Lindemann B, Müller T, Vietz H, Jazdi N, Weyrich M. A survey on long short-term memory networks for time series prediction. Procedia CIRP 2021; 99: 650-655
• 34 Martınez-Llop PG, Sanz Bobi JD, Ortega MO. Time consideration in machine learning models for train comfort prediction using LSTM networks. Eng Appl of Artif Intell 2023; 123: 106303