The surface EMG signal in the action of dynamic contraction has more movement interference compared to sustained static contractions. In addition, the recruitment and de-recruitment of motor units causes a faster change in the surface EMG signal's proprieties. Therefore, more complex techniques are required to extract information from the surface EMG signal. The standardized protocol for surface myoelectric signal measurement in table tennis was a case study in this research area. The Autoregressive method based on the Akaike Information Criterion, the Wavelet method based on intensity analysis, and the Hilbert-Huang transform method were used to estimate the muscle fatigue and non-fatigue condition. The result was that the Hilbert-Huang transform method was shown to be more reliable and accurate for studying the biceps brachii muscle in both conditions. However, the Wavelet method based on intensity analysis is more reliable and accurate for the pectoralis major muscle, deltoideus anterior muscle and deltoideus medialis muscle. The results suggest that different time-frequency analysis techniques influence different muscle analyses based on surface EMG signals in fatigue and non-fatigue conditions.