Reporting reliability with precision and accuracy is of paramount importance in empirical data collections to ascertain whether data is trustworthy. Reliability is often quantified using the intraclass correlation coefficient (ICC), from which the standard error of measurement (SEM) and the minimal detectable change (MDC) can be calculated. However, the literature outlined limited validity of the ICC to account for systematic and random measurement errors stemming from learning or fatiguing effects or a lack of standardization or normal biological variability, respectively. Therefore, the Bland–Altman analysis was introduced to illustrate the systematic bias and quantify the random error via the limits of agreement, originally used to evaluate agreement between devices. Unfortunately, the literature presents common interpretation problems, including missing reference values or misunderstanding of the message transported by the upper and lower border of the Bland–Altman analysis. In this communication paper, we introduce a modified reporting of the mean absolute percentage error that can solve this interpretation problem as it provides a percentage reporting of the mean random error and orientate on the random error calculation performed in the Bland–Altman analysis.