Reaction to a Moving Object Assessment Results in Combat Sports Athletes

This dataset contains the results of a reaction to a moving object (RMO) assessment conducted in 312 combat sports athletes aged 10-25 years. The assessment was performed using the specialized mobile application on iPadOS “Reaction RMO Pro”, designed to evaluate the accuracy, speed, stability, and direction of sensorimotor responses to a moving visual stimulus. During the test, a visual object moved across the screen toward a designated target. Participants were instructed to respond when they perceived that the moving object had reached the target position. Depending on the timing of the response, each attempt was classified as premature, delayed, or accurate. The assessment consisted of three consecutive stages, recorded in the dataset as RT1, RT2, and RT3. Each row represents the de-identified test result of one athlete. The dataset includes the following variables: ID — sequential de-identified participant identifier; RT1_mean, RT2_mean, and RT3_mean — mean reaction-time values for the three test stages, expressed in milliseconds; RT1_SD, RT2_SD, and RT3_SD — standard deviations of reaction-time values for the three stages, expressed in milliseconds; Premature_% — percentage of premature responses; Delayed_% — percentage of delayed responses; Accurate_% — percentage of accurate responses; Responses — dominant response direction: premature responses, delayed responses, or an equal proportion of premature and delayed responses; RMO_Level_Target — predefined classification target with three levels: Low, Medium, and High. The classification target was formed using the integral RMO_Index, which combines standardized values of the average reaction time (RTAvg) and reaction-time variability (RTSD): RMO_Index = -(zRTAvg + zRTSD), where the reference parameters were RTAvg: mean = 32.44 ms, SD = 10.29 ms; and RTSD: mean = 23.30 ms, SD = 7.16 ms. Higher RMO_Index values correspond to faster and more stable reactions. Based on the tertile distribution of the integral index, three levels were defined: High: RMO_Index > 1.08; Medium: -0.56 < RMO_Index ≤ 1.08; Low: RMO_Index ≤ -0.56. The resulting target classes are balanced, with 104 records in each category. These categories are relative to the investigated sample and should not be interpreted as clinical or population-based normative thresholds. The dataset contains 312 complete records and 12 variables. No missing values or duplicate participant IDs were identified. The percentages of premature, delayed, and accurate responses sum to 100% for every record. The released data contain only sequential study identifiers and do not include names, contact information, dates of birth, or other direct personal identifiers. The dataset is provided in machine-readable CSV format and as a documented XLSX workbook containing the original data, variable codebook, dataset description, descriptive statistics, and quality-control results. Additional files include a data dictionary, README, license information. The dataset may be used for research in sports science, psychophysiology, sensorimotor control, combat sports, and machine learning, including the development and validation of models for automated classification of reaction to a moving object performance. The dataset is distributed under the Creative Commons Attribution 4.0 International license (CC BY 4.0).