The first protocol of this type was provided by Blumenstein and Basmajian in the 1980s. It is based on a large set of experimental measures that indicate, for each underlying muscle, the exact area on the skin that allows for signal detection with minimum crosstalk (interference) from adjacent muscles. The protocol had been verified by Professor Yves Blanc in Geneva over the subsequent 20 years. A free description of the protocol for the lower limb is available here http://benthamopen.com/contents/pdf/TOREHJ/TOREHJ-3-110.pdf. This protocol for electrode placement is recommended when sEMG is used in the clinical field because the presence of crosstalk may alter the clinical interpretation of data.
A second protocol was developed in 1999, as the result of a European research project called SENIAM. The positions for electrode placement were obtained from the analysis of available literature and then filtered by inclusion/exclusion criteria. Thanks to both the authors’ effort to disseminate the protocol and its wide availability on the internet, it has become a standard among published papers. Unfortunately, validity of the proposed placement was never verified by experimentation. In my opinion, this protocol is a good choice for research studies without direct clinical impact. SENIAM recommendations for sensor locations are available here http://seniam.org/sensor_location.htm.
A third protocol has been proposed by researchers who use electrode arrays. Papers by Merletti and Rainoldi are the reference in this case. Their idea is to place the electrode pair over the portion of muscle where amplitude is less affected by anatomical factors; i.e. between the tendon and neuromuscular junction.