An apparatus for video decoding includes receiving circuitry and processing circuitry. In some examples, processing circuitry determines a parameter based on prediction information decoded from a coded video bitstream. The parameter is in a range that depends on a flag indicative of an enabler disable status of subblock based temporal motion vector prediction. Then, the processing circuitry calculates the maximum number of candidates in the subblock based merge candidate lists based on the parameter, and reconstructs, m response to a current block in a subblock based prediction mode, samples of the current block based on a candidate selection from a constructed subblock based merge candidate list of the current block. The constructed subblock based merge candidate list of the current block is constrained by the maximum number.