Modeling the Prediction of the Session Rating of Perceived Exertion in Soccer: Unraveling the Puzzle of Predictive Indicators

Prediction of Perceived Exertion Ratings in National Level Soccer Players Using Wearable Sensor Data and Machine Learning Techniques

This study aimed to identify relationships between external and internal load parameters with subjective ratings of perceived exertion (RPE). Consecutively, these relationships shall be used to evaluate different machine learning models and design a deep learning architecture to predict RPE in highly trained/national level soccer players. From a dataset comprising 5402 training sessions and 732 match observations, we gathered data on 174 distinct parameters, encompassing heart rate, GPS, accelerometer data and RPE (Borg's 0-10 scale) of 26 professional male professional soccer players. Nine machine learning algorithms and one deep learning architecture was employed. Rigorous preprocessing protocols were employed to ensure dataset equilibrium and minimize bias. The efficacy and generalizability of these models were evaluated through a systematic 5-fold cross-validation approach. The deep learning model exhibited highest predictive power for RPE (Mean Absolute Error: 1.08 ± 0.07). Tree-based machine learning models demonstrated high-quality predictions (Mean Absolute Error: 1.15 ± 0.03) and a higher robustness against outliers. The strongest contribution to reducing the uncertainty of RPE with the tree-based machine learning models was maximal heart rate (determining 1.81% of RPE), followed by maximal acceleration (determining 1.48%) and total distance covered in speed zone 10-13 km/h (determining 1.44%). A multitude of external and internal parameters rather than a single variable are relevant for RPE prediction in highly trained/national level soccer players, with maximum heart rate having the strongest influence on RPE. The ExtraTree Machine Learning model exhibits the lowest error rates for RPE predictions, demonstrates applicability to players not specifically considered in this investigation, and can be run on nearly any modern computer platform.

Relationship between external and internal load indicators and injury using machine learning in professional soccer: a systematic review and meta-analysis

This study verified the relationship between internal load (IL) and external load (EL) and their association on injury risk (IR) prediction considering machine learning (ML) approaches. Studies were included if: (1) participants were male professional soccer players; (2) carried out for at least 2 sessions, exercises, or competitions; (3) correlated training load (TL) with non-contact injuries; (4) applied ML approaches to predict TL and non-contact injuries. TL included: IL indicators (Rating of Perceived Exertion, RPE; Session-RPE, Heart Rate, HR) and EL indicators (Global Positioning System, GPS variables); the relationship between EL and IL through index, ratio, formula; ML indicators included performance measures, predictive performance of ML methods, measure of feature importance, relevant predictors, outcome variable, predictor variable, data pre-processing, features selection, ML methods. Twenty-five studies were included. Eleven addressed the relationship between EL and IL. Five used EL/IL indexes. Five studies predicted IL indicators. Three studies investigated the association between EL and IL with IR. One study predicted IR using ML. Significant positive correlations were found between S-RPE and total distance (TD) (r = 0.73; 95% CI (0.64 to 0.82)) as well as between S-RPE and player load (PL) (r = 0.76; 95% CI (0.68 to 0.84)). Association between IL and EL and their relationship with injuries were found. RPE, S-RPE, and HR were associated with different EL indicators. A positive relationship between EL and IL indicators and IR was also observed. Moreover, new indexes or ratios (integrating EL and IL) to improve knowledge regarding TL and fitness status were also applied. ML can predict IL indicators (HR and RPE), and IR. The present systematic review was registered in PROSPERO (CRD42021245312).

Feature importance for estimating rating of perceived exertion from cardiorespiratory signals using machine learning

Introduction

The purpose of this study is to investigate the importance of respiratory features, relative to heart rate (HR), when estimating rating of perceived exertion (RPE) using machine learning models.

Methods

A total of 20 participants aged 18 to 43 were recruited to carry out Yo-Yo level-1 intermittent recovery tests, while wearing a COSMED K5 portable metabolic machine. RPE information was collected throughout the Yo-Yo test for each participant. Three regression models (linear, random forest, and a multi-layer perceptron) were tested with 8 training features (HR, minute ventilation (VE), respiratory frequency (Rf), volume of oxygen consumed (VO2), age, gender, weight, and height).

Results

Using a leave-one-subject-out cross validation, the random forest model was found to be the most accurate, with a root mean square error of 1.849, and a mean absolute error of 1.461 ± 1.133. Feature importance was estimated via permutation feature importance, and VE was found to be the most important for all three models followed by HR.

Discussion

Future works that aim to estimate RPE using wearable sensors should therefore consider using a combination of cardiovascular and respiratory data.

Artificial intelligence applications in the football codes: A systematic review

Artificial Intelligence (AI) is increasingly being adopted across many domains such as transport, healthcare, defence and sport, with football codes no exception. Though there is a range of potential benefits of AI, concern has also been expressed regarding potential risks. An important first step in ensuring that AI applications in football are usable, beneficial, safe and ethical is to understand the current range of applications, the AI models adopted and their proposed functions. This systematic review aimed to identify different applications of AI across football codes to synthesise current knowledge and determine whether potential risks are being considered. The systematic review included 190 peer-reviewed articles. Nine areas of application were found ranging from athlete evaluation and event detection to match outcome prediction and injury detection and prediction. In total, 27 different AI models were identified, with artificial neural networks the most frequently applied. Five AI assessment metrics were identified including specificity, recall, precision, accuracy and F1-score. Four potential risks were identified, concerning data security, usability, data biases and inappropriate athlete load management. It is concluded that, though a wide range of AI applications currently exist, further work is required to develop AI for football and identify and manage potential risks.

A random forest approach to explore how situational variables affect perceived exertion of elite youth soccer players

The aim of this study was to explore how situational variables affect youth soccer players' perceived exertion (RPE) after official matches. Thirty-five elite youth male players (14.33 ± 0.86 years; 173.49 ± 6.16 cm; 63.44 ± 5.98 kg) who belonged to two different teams of a professional club participated in this study. Data collection was conducted during two seasons (2016-2017, 2017-2018) and included 60 official matches (30 official matches per team). Ten minutes after each match players rated their RPE and using a modified Borg CR-10 scale. A Random Forest Regression was used to quantify the importance of match-related situational variables in RPE. Afterwards, a linear mixed model analysis was applied to identify the variability in RPE among the situational variables. The game-playing time, the player status (starter or substitute) and the player identity were the strongest predictors of RPE. Moreover, the match outcome and the final scoreline showed significant effects on both starter and substitute players but the main effect of the quality of the opponent was only identified in starter players (p < 0.05). These results allow practitioners to know how situational variables interact and modulate RPE after official matches and help them to prescribe and adapt the players' training content and load before and after matches.

Relationships between External, Wearable Sensor-Based, and Internal Parameters: A Systematic Review

Micro electro-mechanical systems (MEMS) are used to record training and match play of intermittent team sport athletes. Paired with estimates of internal responses or adaptations to exercise, practitioners gain insight into players' dose-response relationship which facilitates the prescription of the training stimuli to optimize performance, prevent injuries, and to guide rehabilitation processes. A systematic review on the relationship between external, wearable-based, and internal parameters in team sport athletes, compliant with the PRISMA guidelines, was conducted. The literature research was performed from earliest record to 1 September 2020 using the databases PubMed, Web of Science, CINAHL, and SportDISCUS. A total of 66 full-text articles were reviewed encompassing 1541 athletes. About 109 different relationships between variables have been reviewed. The most investigated relationship across sports was found between (session) rating of perceived exertion ((session-)RPE) and PlayerLoad™ (PL) with, predominantly, moderate to strong associations (r = 0.49-0.84). Relationships between internal parameters and highly dynamic, anaerobic movements were heterogenous. Relationships between average heart rate (HR), Edward's and Banister's training impulse (TRIMP) seem to be reflected in parameters of overall activity such as PL and TD for running-intensive team sports. PL may further be suitable to estimate the overall subjective perception. To identify high fine-structured loading-relative to a certain type of sport-more specific measures and devices are needed. Individualization of parameters could be helpful to enhance practicality.

The contributing external load factors to internal load during small-sided games in professional rugby union players

Introduction

This study aimed to investigate which external load variables were associated with internal load during three small-sided games (SSG) in professional rugby union players.

Methods

Forty professional rugby union players (22 forwards, 18 backs) competing in the English Gallagher Premiership were recruited. Three different SSGs were designed: one for backs, one for forwards, and one for both backs and forwards. General linear mixed-effects models were implemented with internal load as dependent variable quantified using Stagno's training impulse, and external load as independent variables quantified using total distance, high-speed (>61% top speed) running distance, average acceleration-deceleration, PlayerLoad™, PlayerLoad™ slow (<2 m·s^-1), number of get-ups, number of first-man-to-ruck.

Results

Internal load was associated with different external load variables dependent on SSG design. When backs and forwards were included in the same SSG, internal load differed between positional groups (MLE = -121.94, SE = 29.03, t = -4.20).

Discussion

Based on the SSGs investigated, practitioners should manipulate different constraints to elicit a certain internal load in their players based on the specific SSG design. Furthermore, the potential effect of playing position on internal load should be taken into account in the process of SSG design when both backs and forwards are included.

PERSIST: A Multimodal Dataset for the Prediction of Perceived Exertion during Resistance Training

Measuring and adjusting the training load is essential in resistance training, as training overload can increase the risk of injuries. At the same time, too little load does not deliver the desired training effects. Usually, external load is quantified using objective measurements, such as lifted weight distributed across sets and repetitions per exercise. Internal training load is usually assessed using questionnaires or ratings of perceived exertion (RPE). A standard RPE scale is the Borg scale, which ranges from 6 (no exertion) to 20 (the highest exertion ever experienced). Researchers have investigated predicting RPE for different sports using sensor modalities and machine learning methods, such as Support Vector Regression or Random Forests. This paper presents PERSIST, a novel dataset for predicting PERceived exertion during reSIStance Training. We recorded multiple sensor modalities simultaneously, including inertial measurement units (IMU), electrocardiography (ECG), and motion capture (MoCap). The MoCap data has been synchronized to the IMU and ECG data. We also provide heart rate variability (HRV) parameters obtained from the ECG signal. Our dataset contains data from twelve young and healthy male participants with at least one year of resistance training experience. Subjects performed twelve sets of squats on a Flywheel platform with twelve repetitions per set. After each set, subjects reported their current RPE. We chose the squat exercise as it involves the largest muscle group. This paper demonstrates how to access the dataset. We further present an exploratory data analysis and show how researchers can use IMU and ECG data to predict perceived exertion.

An Exploratory Data Analysis on the Influence of Role Rotation in a Small-Sided Game on Young Soccer Players

The aim of the present study was to analyze the behavior of players in a standard small-sided game (SSG) according to the role played (offensive (OF), defensive (DF), and wildcard (W)) and its relationship with physical demands (PHYD), technical performance (TP), and internal load (RPE). A total of 24 young highly trained male soccer players (under 16: n = 12; under 19: n = 12) participated. During the SSG, the players alternated the three roles (OF, DF, and W). The duration of each repetition was 4 min with a passive rest of 3 min between them. Furthermore, it emphasized the high demand in all defensive parameters. In addition, DF roles showed higher values in PHYD and RPE, followed by the OF roles, and finally by the W roles. A complementary, positive moderate correlation was found between PHYD and RPE in the U16 dataset (r = 0.45, p < 0.006). Very large positive correlations were also found between PHYD and RPE in the U19 and merged dataset (r = 0.78, p < 0.001 and r = 0.46, p < 0.63, respectively). This information could be useful for coaches in order to structure the roles in SSGs and control training load.

Machine Learning-Based Identification of the Strongest Predictive Variables of Winning and Losing in Belgian Professional Soccer

This study aimed to identify the strongest predictive variables of winning and losing in the highest Belgian soccer division. A predictive machine learning model based on a broad range of variables (n = 100) was constructed, using a dataset consisting of 576 games. To avoid multicollinearity and reduce dimensionality, Variance Inflation Factor (threshold of 5) and BorutaShap were respectively applied. A total of 13 variables remained and were used to predict winning or losing using Extreme Gradient Boosting. TreeExplainer was applied to determine feature importance on a global and local level. The model showed an accuracy of 89.6% ± 3.1% (precision: 88.9%; recall: 90.1%, f1-score: 89.5%), correctly classifying 516 out of 576 games. Shots on target from the attacking penalty box showed to be the best predictor. Several physical indicators are amongst the best predictors, as well as contextual variables such as ELO -ratings, added transfers value of the benched players and match location. The results show the added value of the inclusion of a broad spectrum of variables when predicting and evaluating game outcomes. Similar modelling approaches can be used by clubs to identify the strongest predictive variables for their leagues, and evaluate and improve their current quantitative analyses.

Load Measures in Training/Match Monitoring in Soccer: A Systematic Review

In soccer, the assessment of the load imposed by training and a match is recognized as a fundamental task at any competitive level. The objective of this study is to carry out a systematic review on internal and external load monitoring during training and/or a match, identifying the measures used. In addition, we wish to make recommendations that make it possible to standardize the classification and use of the different measures. The systematic review was carried out according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. The search was conducted through the electronic database Web of Science, using the keywords "soccer" and "football", each one with the terms "internal load", "external load", and "workload". Of the 1223 studies initially identified, 82 were thoroughly analyzed and are part of this systematic review. Of these, 25 articles only report internal load data, 20 report only external load data, and 37 studies report both internal and external load measures. There is a huge number of load measures, which requires that soccer coaches select and focus their attention on the most useful and specific measures. Standardizing the classification of the different measures is vital in the organization of this task, as well as when it is intended to compare the results obtained in different investigations.